FNB v30n1 Supplement Izinc PDF

FOOD AND
NUTRITION
BULLETIN Volume 30, Number 1, March 2009
SUPPLEMENT
International Zinc Nutrition Consultative Group
Technical Document #2
Systematic Reviews of Zinc Intervention
Strategies
Kenneth H. Brown and Sonja Y. Hess, guest editors
Advances in zinc nutrition and health
Preventive zinc supplementation in children
Therapeutic zinc supplementation in children
Zinc supplementation during pregnancy and lactation
Zinc fortification
Dietary diversification or modification to enhance zinc intakes
Zinc intake through breastmilk
Improving zinc status through biofortification
Conclusions and mainstreaming zinc interventions
International Nutrition Foundation

for
United Nations
University Press
FOOD AND NUTRITION BULLETIN The Food and Nutrition Bulletin also serves as a principal
Published by the International Nutrition Foundation for the outlet for the publication of reports of working groups and
United Nations University other activities of the SCN. The SCN itself is a focal point
for coordinating activities of the Food and Agriculture
Editorial office: Organization of the United Nations (FAO), the World Health
Food and Nutrition Bulletin Organization (WHO), the International Atomic Energy Agency
Michelle Badash, Managing Editor (IAEA), the United Nations Children’s Fund (UNICEF), the
150 Harrison Ave. UNU, the World Bank, the World Food Program (WFP), and
Boston, MA, 02111, USA other bodies of the United Nations system that have an interest
Tel: 617-636-3778 in food and nutrition. The Bulletin also publishes reports of
Fax: 617-636-3727 working groups and task forces operating under the IUNS, a
E-mail: FNB@inffoundation.org collaborating institution.
Submissions. Unsolicited manuscripts of articles of the type
Administrative office: published in this and previous issues may be sent to the
International Nutrition Foundation Managing Editor at the address given above electronically or
150 Harrison Ave. by regular mail. Manuscripts must be submitted electronically
Boston, MA, 02111, USA as e-mail attachments sent directly to the Managing Editor.
Tel: 617-636-3771 (Please refer to the Information for Authors page in the back
Fax: 617-636-3727 of this issue for further details on submission requirements.)
E-mail: infoperations@inffoundation.org Any disciplinary or conceptual approach relevant to
problems of world hunger and malnutrition is welcome, and
Subscriptions: controversy over some of the articles is anticipated. Letters to
FNBsubscriptions@inffoundation.org the editor are encouraged and will be printed if judged to have
an adequate basis and to be of sufficient general interest.
The Food and Nutrition Bulletin is published quarterly Also of interest are “Short Communications,” defined as
by the International Nutrition Foundation (INF) for the papers that explore a topic with the customary scientific rigor
United Nations University (UNU), in collaboration with the but that do not need extensive analysis and discussion of the
United Nations system Standing Committee on Nutrition data or that have useful information that is not sufficient for a
(SCN) and the International Union of Nutritional Sciences full article. These would be identified as such at the beginning
(IUNS). All correspondence concerning the content of the of each article.
Bulletin, comments, news, and notices should be sent to the Peer review. The Bulletin is a peer-reviewed journal. Every
Managing Editor at the editorial office address given above. article is submitted first to editorial review and then sent to
All material in the Bulletin may be reproduced freely provided two reviewers for further evaluation. After review, the editor
acknowledgment is given and a copy of the publication may accept, reject, or return a paper to the authors with
containing the reproduction is sent to the editorial office. suggestions for improvement. In these cases, the authors are
given a chance to respond to the reviewers’ comments.
Editorial policy Disclaimer. It is expressly understood that articles published
The Food and Nutrition Bulletin is intended to make available in the Bulletin do not necessarily represent the views of the
policy analyses, state-of-the-art summaries, and original INF, UNU, SCN, or any United Nations organization. The
scientific articles relating to multidisciplinary efforts to alleviate views expressed and the accuracy of the information on which
the problems of hunger and malnutrition in the developing they are based are the responsibility of the authors. Some
world. It is not intended for the publication of scientific articles articles in the Bulletin are reports of various international
of principal interest only to individuals in a single discipline committees and working groups and do represent the
or within a single country or region. Reviews or notices of consensus of the individuals involved; whether or not they
relevant books and other publications will be published if they also represent the opinions or policies of the sponsoring
are received for review. The Bulletin is also a vehicle for notices organizations is expressly stated. Mention of the names of
of forthcoming international meetings that satisfy the above firms and commercial products does not imply endorsement
criteria and for summaries of such meetings. of the United Nations University.
The United Nations University (UNU) is an organ of the United Nations established by the General
Assembly in 1972 to be an international community of scholars engaged in research, advanced training,
and the dissemination of knowledge related to the pressing global problems of human survival,
development, and welfare. Its activities focus mainly on peace and conflict resolution, development in
a changing world, and science and technology in relation to human welfare. The University operates
through a worldwide network of research and postgraduate training centres, with its planning and
coordinating headquarters in Tokyo. The United Nations University Press, the publishing division of
the UNU, publishes scholarly books and periodicals in the social sciences, humanities, and pure and
applied natural sciences related to the University’s research.
Contents
International Zinc Nutrition Consultative Group Technical Document No. 2

Systematic reviews of zinc intervention strategies
Kenneth H. Brown and Sonja Y. Hess, guest editors
Acknowledgments . ..........................................................................................................................................................S3
Recent advances in knowledge of zinc nutrition and human health
—S. Y. Hess, B. Lönnerdal, C. Hotz, J. A. Rivera, and K. H. Brown...........................................................................S5
Preventive zinc supplementation among infants, preschoolers, and older prepubertal children
—K. H. Brown, J. M. Peerson, S. K. Baker, and S. Y. Hess .......................................................................................S12
The effect of therapeutic zinc supplementation among young children with selected infections:
A review of the evidence —B. A. Haider and Z. A. Bhutta ............................................................................S41
Effects of maternal zinc supplementation on pregnancy and lactation outcomes
—S. Y. Hess and J. C. King . ........................................................................................................................................S60
Impact of zinc fortification on zinc nutrition —S. Y. Hess and K. H. Brown...................................................S79
A review of interventions based on dietary diversification or modification strategies with the
potential to enhance intakes of total and absorbable zinc —R. S. Gibson and V. P. Anderson . ............ S108
Dietary intervention strategies to enhance zinc nutrition: Promotion and support of breastfeeding
for infants and young children —K. H. Brown, R. Engle-Stone, N. F. Krebs, and J. M. Peerson . .......... S144
The potential to improve zinc status through biofortification of staple food crops with zinc
—C. Hotz................................................................................................................................................................... S172
Galvanizing action: Conclusions and next steps for mainstreaming zinc interventions in public health
programs —K. H. Brown, S. K. Baker, and the IZiNCG Steering Committee.......................................... S179
List of contributors....................................................................................................................................................... S185
This publication does not necessarily represent the decisions or the stated policy of the World Health
Organization. The named authors of the different papers are solely responsible for the views expressed in
these papers.
Food and Nutrition Bulletin
Editor: Dr. Irwin H. Rosenberg, Friedman School of Nutrition Science

and Policy, Tufts University, Boston, Mass., USA
Senior Associate Editor: Dr. Nevin S. Scrimshaw
Associate Editor—Food Policy and Agriculture: Dr. Suresh Babu, International
Food Policy Research Institute (IFPRI), Washington, DC, USA
Associate Editor — Program Communication: Dr. Gary R. Gleason, Tufts
University, Gerald J. and Dorothy R. Friedman School of Nutrition Science
and Policy, Boston, MA USA
Associate Editor—Food Science and Technology: Dr. V. Prakash, Central Food
Technological Research Institute (CFTRI), Mysore, India
Statistical Advisor—Dr. William M. Rand, Tufts University School of
Medicine, Boston, Mass., USA
Managing Editor: Ms. Michelle Badash
Manuscripts Editor: Mr. Jonathan Harrington
Copyeditor: Ms. Ellen Duff
Editorial Assistant: Georgette Baghdady
Editorial Board:
Dr. Ricardo Bressani, Institute de Investigaciones, Universidad del Valle
de Guatemala, Guatemala City, Guatemala
Dr. Hernán Delgado, Director, Institute of Nutrition of Central America
and Panama (INCAP), Guatemala City, Guatemala
Dr. Cutberto Garza, Academic Vice President and Dean of Faculties, Boston
College, Chestnut Hill, Mass., USA
Dr. Joseph Hautvast, Secretary General, International Union of Nutritional
Sciences (IUNS), Department of Human Nutrition, Agricultural University,
Wageningen, Netherlands
Dr. Peter Pellett, Professor, Department of Food Science and Nutrition,
University of Massachusetts, Amherst, Mass., USA
Dr. Zewdie Wolde-Gabreil, Director, Ethiopian Nutrition Institute, Addis
Ababa, Ethiopia
Dr. Aree Valyasevi, Professor and Institute Consultant, Mahidol University,
Food and Nutrition Bulletin, vol. 30, no. 1, Supplement

© The United Nations University, 2009
United Nations University Press
Published by the International Nutrition Foundation for The United Nations University
150 Harrison Avenue, Boston, MA 02111 USA
Tel.: (617) 636-3778 Fax: (617) 636-3727
E-mail: FNB@inffoundation.org
ISSN 0379-5721
Design and production by Digital Design Group, Newton, MA USA
Printed on acid-free paper by Webcom, Toronto, ON Canada
Acknowledgments
This document was an initiative of the International The comments and suggestions by the IZiNCG
Zinc Nutrition Consultative Group (IZiNCG). Steering Committee were used to revise each of the
The initial drafts of each paper were reviewed by the papers, which were then submitted to the International
IZiNCG Steering Committee: Union of Nutritional Scientists (IUNS). IUNS and
Kenneth H. Brown, M.D. (Chair) the Food and Nutrition Bulletin jointly coordinated
Juan A. Rivera, Ph.D. (Co-chair) the external review by representatives of the United
Shawn K. Baker, M.P.H. Nations Children’s Fund (UNICEF) and the World
Zulfiqar A. Bhutta, M.D., Ph.D. Health Organization. Responses to the reviewers’ com-
Omar Dary, Ph.D. ments and suggestions were incorporated into the final
Rosalind S. Gibson, Ph.D. published version.
Christine Hotz, Ph.D. The work was carried out with the financial support
Janet C. King, Ph.D. of the International Zinc Association ((IZA) Brus-
Bo Lönnerdal, Ph.D. sels, Belgium) and the Micronutrient Initiative ((MI)
Marie T. Ruel, Ph.D. Ottawa, Canada). Support for publication was provided
Emorn Wasantwisut, Ph.D. by MI, UNICEF, IZA, and SIGHT AND LIFE (Basel,
Sonja Y. Hess, Ph.D. (Executive officer) Switzerland).
Food and Nutrition Bulletin, vol. 30, no. 1 (supplement) © 2009, The United Nations University. S3
S4 S. Y. Hess et al.
Recent advances in knowledge of zinc nutrition and
human health
Sonja Y. Hess, Bo Lönnerdal, Christine Hotz, Juan A. Rivera, and Kenneth H. Brown
Abstract Key words: Assessment, zinc, zinc deficiency, zinc

metabolism, zinc status
Zinc deficiency increases the risk and severity of a variety
of infections, restricts physical growth, and affects specific
outcomes of pregnancy. Global recognition of the impor- Background
tance of zinc nutrition in public health has expanded
dramatically in recent years, and more experience has In 2004, the International Zinc Nutrition Consulta-
accumulated on the design and implementation of zinc tive Group (IZiNCG) published a technical review
intervention programs. Therefore, the Steering Com- [1] that was designed to provide an overview of cur-
mittee of the International Zinc Nutrition Consultative rent knowledge regarding zinc nutrition in relation to
Group (IZiNCG) completed a second IZiNCG technical human health, to summarize the available information
document that reexamines the latest information on on assessing population zinc status, and to describe
the intervention strategies that have been developed to the range of programmatic options for controlling
enhance zinc nutrition and control zinc deficiency. In zinc deficiency. Since the publication of that docu-
particular, the document reviews the current evidence ment, recognition of the importance of zinc nutrition
regarding preventive zinc supplementation and the role for human health worldwide has expanded dramati-
of zinc as adjunctive therapy for selected infections, zinc cally, and more experience has been accumulated on
fortification, and dietary diversification or modification the design and implementation of zinc intervention
strategies, including the promotion and protection of programs. Moreover, during the workshop on zinc
breastfeeding and biofortification. supplementation and child mortality and morbidity
The purposes of this introductory paper are to summa- held by the World Health Organization (WHO) in
rize new guidelines on the assessment of population zinc September 2006, it was concluded that “in view of the
status, as recommended by the World Health Organi- results of all the trials examining the impact of zinc
zation (WHO), the United Nations Children’s Fund supplementation on mortality, morbidity and growth,
(UNICEF), the International Atomic Energy Agency a consensus was reached on the need to develop new
(IAEA), and IZiNCG, and to provide an overview on feasible approaches to improve the intake of zinc and
several new advances in zinc metabolism. The follow- its bioavailability in young children, in order to achieve
ing papers will then review the intervention strategies adequate population coverage” [2]. Hence, the IZiNCG
individually. Steering Committee concluded that this would be an
opportune time to reexamine the latest information
Sonja Y. Hess, Bo Lönnerdal, and Kenneth H. Brown are on strategies to control zinc deficiency and to reassess
affiliated with the Department of Nutrition and the Program the state of knowledge concerning interventions to
in International and Community Nutrition, University of enhance zinc nutrition.
California, Davis, California, USA; Christine Hotz is affiliated Adequate zinc nutrition is essential for human health
with HarvestPlus and the International Food Policy Research
Institute, Washington, DC, USA; Juan A. Rivera is affiliated because of zinc’s critical structural and functional roles
with the National Institute of Public Health, Cuernavaca, in multiple enzyme systems that are involved in gene
Mexico; Kenneth H. Brown is also affiliated with Helen Keller expression, cell division and growth, and immunologic
International, Dakar, Senegal. and reproductive functions. As a consequence, zinc
Please direct queries to the corresponding author: Kenneth
H. Brown, Department of Nutrition, University of California, deficiency affects children’s physical growth and the risk
One Shields Ave., Davis, CA 95616, USA; e-mail: khbrown@ and severity of a variety of infections [1]. The results of
ucdavis.edu. multiple community-based intervention trials indicate
that zinc supplementation decreases the incidence of serum zinc concentration have been comprehensively
diarrhea and pneumonia among young children [3], reviewed [1, 12]. The prevalence of zinc deficiency
and clinical treatment studies have shown that zinc should be expressed as the percentage of the popula-
supplementation during diarrhea reduces the sever- tion with serum zinc concentration below the specific
ity and duration of such illnesses [4]. WHO and the lower cutoffs in relation to reference data for age,
United Nations Children’s Fund (UNICEF) now recom- sex, time of day, and fasting status of the individuals
mend that zinc supplementation should be included examined [13, 14]. When the prevalence of low serum
as a component in diarrhea treatment regimens [5], zinc concentration is greater than 20%, the risk of zinc
and efforts are under way in a number of countries deficiency is considered to be elevated and should be
to scale up zinc supplementation during diarrhea. addressed through public health nutrition interventions
In the above-mentioned report on the WHO work- to improve zinc status. This same indicator also can be
shop on zinc and mortality, the results of a meta-analysis used to assess the impact of an intervention program,
of available trials of preventive zinc supplementation by comparing the percentage of individuals with low
indicated that there was a statistically significant 9% serum zinc concentrations before and after initiation
reduction in overall mortality among young children of the intervention. Because serum zinc concentration
who received zinc supplementation [2]. The recent falls during the acute-phase response to infections, it
Lancet series on maternal and child undernutrition is advisable to include biochemical indicators of infec-
concluded that zinc deficiency is responsible for ~4% tion, such as C-reactive protein or α1-glycoprotein, to
of child mortality and disability-adjusted life-years [6]. avoid the possibility of overestimating the prevalence
Although the specific number of deaths that might of low serum zinc concentration due to concurrent
be averted by zinc-related interventions can still be infections [15, 16].
debated because of the numerous assumptions involved Inadequate dietary intake of absorbable zinc is one of
in such estimates, both of these foregoing analyses con- the major causes of zinc deficiency. Therefore, assess-
firm that zinc deficiency is an important risk factor for ment of the adequacy of zinc intakes through the use of
child morbidity and mortality [7, 8]. 24-hour recalls or weighed dietary records is an impor-
In addition to the effects of zinc on morbidity and tant component in evaluating the risk of zinc deficiency
mortality, a number of studies indicate that preven- in a population [11]. Dietary assessment can be used
tive zinc supplementation increases linear growth to identify subpopulations that have an elevated risk of
and weight gain in previously stunted or underweight zinc deficiency and to characterize dietary patterns that
children [9]. Thus, interventions to prevent zinc defi- contribute to inadequate zinc intakes, thus informing
ciency also can reduce the overall rates of childhood on the appropriate design of food-based interventions.
malnutrition, as defined by anthropometric criteria. The prevalence of the population with zinc intakes
For all of these reasons, global commitment is urgently less than the Estimated Average Requirement (EAR)
needed to implement policies and programs designed [1, 17] can be used as the specific indicator of the risk
to control zinc deficiency. of zinc deficiency in the population. Assessment of
the adequacy of zinc intakes should take into account
dietary zinc bioavailability, preferably through quanti-
Assessment of the risk of zinc deficiency fication of the phytate:zinc molar ratio of the diet [18]
or by using available equations to predict zinc absorp-
Because of the serious consequences of zinc deficiency, tion based on dietary zinc and phytate contents [19].
it is essential to quantify the risk of deficiency in those The risk of zinc deficiency is considered to be elevated
populations that are most likely to be affected by this and of public health concern when the prevalence of
problem. Regrettably, there are as yet very limited inadequate intakes is greater than 25%, in which case an
national-level data on the prevalence of zinc deficiency. intervention to increase dietary zinc intakes is recom-
To promote the inclusion of zinc status assessment in mended [11]. The change in prevalence of inadequate
the context of national health and nutrition surveys, zinc intakes can be used to assess the impact and effec-
guidelines on the assessment of population zinc status tive targeting of food-based interventions.
were recently published following a consensus confer- Although there are several adverse functional con-
ence convened by WHO, UNICEF, the International sequences of inadequate zinc intake, these outcomes
Atomic Energy Agency (IAEA), and IZiNCG [10]. are not specific to zinc deficiency. For example, the
The three main types of zinc status assessment that incidence of some types of infections can be reduced
were considered included biochemical, dietary, and by providing supplemental zinc [3, 20], but the disease
functional methods. rates are more closely linked to the level of expo-
Serum or plasma zinc concentration is considered sure to specific pathogens. Thus, a high incidence or
the best available biomarker of the risk of zinc defi- prevalence of particular infections, such as diarrhea,
ciency in populations [11]. Methods for collecting, may suggest that the population could benefit from
processing, and analyzing samples for determining interventions including zinc, but the illness rates would
Advances in zinc nutrition and health S7
not be very useful in quantifying the extent of zinc to 30%) are found in the Andean countries, some Cen-
deficiency in the population. Similarly, low height- tral American countries, Southern Africa, and some
for-age is not specific to zinc deficiency and could be countries in North Asia. As zinc deficiency is not the
attributable in part to maternal short stature, frequent only factor affecting children’s growth, assessment of
infections, and other nutritional deficiencies. Thus, dietary zinc intake and serum zinc levels can be used
providing zinc alone should not be expected to fully to confirm the risk of zinc deficiency in these countries
reverse childhood stunting. Nevertheless, a previous [11]. These assessments should be incorporated into
meta-analysis of randomized, controlled trials among existing public health and child nutrition monitoring
prepubertal children found that the severity of stunting programs whenever possible.
in the study populations predicted the response to zinc
supplementation [9]. Thus, the percentage of children
under 5 years of age with height-for-age z-score (HAZ) New advances in zinc metabolism
less than −2.0 SD with respect to the reference popula-
tion [21] has been recommended as the best functional Although this document focuses primarily on the
indicator to assess the likely risk of zinc deficiency in a recent progress that has been achieved with regard
population [11]. This risk is considered to be elevated to the role of zinc nutrition in public health, some of
and of public health concern when the prevalence of the advances that have occurred in our understanding
low height-for-age is greater than 20%, in which case of zinc metabolism and the factors that govern zinc
nutrition intervention strategies should include a homeostasis are also worth noting. A comprehensive
means to improve zinc status. review of new research on zinc metabolism is beyond
The validity of these indicators and proposed cutoffs the scope of this paper, but several new discoveries
is still provisional, so they should be evaluated further concerning zinc transport proteins will be described
when opportunities become available during national briefly, because they provide some insight into the
assessment surveys. As more experience is gained, complexities of zinc homeostasis, and this knowledge
these recommendations will need to be reviewed and eventually may yield useful information for estimat-
revised as necessary. ing dietary zinc requirements and for developing new
methods to assess zinc status.
The efficiency of zinc absorption from the diet usu-
Quantifying the risk of zinc deficiency ally ranges from about 15% to 35% in adults, depend-
ing on the amount consumed and the presence of
As with other micronutrient deficiencies, three main other dietary factors, such as phytate, that may inhibit
factors are responsible for the development of zinc absorption [25]. Active transport dominates at low or
deficiency in lower-income countries: inadequate normal intake, whereas passive diffusion contributes
dietary zinc intake or absorption from predominantly more significantly at high intake [26]. The extent of the
plant-based diets, as discussed above, or suboptimal homeostatic regulation of zinc metabolism in humans
breastfeeding practices; disease states that either induce is not well known, but both absorption and excretion
excessive losses or impair utilization of zinc; and appear to be involved. Studies in experimental animals
physiological states that increase zinc requirements, suggest that zinc homeostasis is closely regulated,
such as periods of rapid growth during childhood and although not to the same extent as for iron. The mecha-
pregnancy. These issues are reviewed in more detail nisms underlying the regulation of zinc absorption
elsewhere [1, 10, 22]. have long remained elusive.
Because so little information is available from nation- Understanding of the mechanisms regulating zinc
ally representative surveys on the prevalence of low absorption and homeostasis has progressed consid-
serum zinc concentration or inadequate dietary zinc erably because of the discovery of two families of
intake, current estimates of the extent of zinc defi- zinc transporters: the so-called ZIP proteins and the
ciency must rely on the prevalence of stunting among ZnT proteins. Members of the ZIP family of proteins
children under 5 years of age [11]. Fortunately, relevant (which are also referred to in the literature as Zrt-like
information is available at the national level for most proteins and Irt-like proteins, with systemic designa-
countries (fig. 1) [23]. Approximately 30% of children tion “SLC39”) transport zinc from the extracellular
under 5 years of age worldwide are stunted (HAZ space and intracellular organelles into the cytoplasm
< –2 SD with respect to the distribution of the reference [27]. Thus, the net effect of these transporters (or
population data). WHO recommends a prevalence of “zinc importer proteins”) is to increase cytoplasmic
stunting greater than 20% of the population to indicate zinc. There are 14 known members of the ZIP family
a public health concern [24]. The highest prevalence encoded by the human genome [28], but only a few of
rates of stunting (> 30%) are observed in countries in them have been characterized or evaluated with regard
sub-Saharan Africa, South Asia, Southeast Asia, and to physiological significance. ZIP-1 is expressed ubiq-
Central America. Intermediate prevalence rates (20% uitously in human tissues but is only localized to the
< 20%
20–30%
30–40%
≥ 40%
Not available
FIG. 1. Prevalence of nutritional stunting in children under 5 years of age. Source: IZiNCG [23]
plasma membrane in some cell types, possibly because has nine members in the human genome. These zinc
of zinc-responsive regulation of its subcellular localiza- transporters are primarily involved in cellular efflux of
tion [29, 30]. In zinc-deficient cells, ZIP-1 migrates to zinc and in uptake of zinc by intracellular organelles.
the plasma membrane, whereas in zinc-replete cells, Thus, the net effect of these transporters is to decrease
it is associated with intracellular compartments [31]. cytoplasmic zinc concentration. ZnT-1 expression
To date, ZIP-1 appears to be mostly involved in zinc in the intestine is regulated by dietary zinc [38] and
uptake by erythroleukemia (K562) cells and prostate has been implicated in the regulation of whole-body
cells [29, 32]. ZIP-2 and ZIP-3 have also been shown to zinc homeostasis by controlling zinc efflux from the
be involved in zinc uptake by some cell types, the latter enterocyte. ZnT-2 and ZnT-4 are involved in the flux
in mammary epithelial cells [33]. of zinc in the endosomes, possibly regulating intracel-
The ZIP-4 transporter has been shown to be a key lular trafficking of zinc. These membrane transporters
zinc importer in the intestinal cell. This transporter all have six transmembrane-spanning domains and a
was discovered when mutations in the ZIP-4 gene were conserved histidine-rich region predicted to have a
linked to the human genetic disorder acrodermatitis cytoplasmic loop that is likely to bind zinc [39, 40].
enteropathica [34]. Acrodermatitis enteropathica is Experiments showing zinc sequestration by endosomal
due to an autosomal recessive mutation of the gene vesicles during overexpression of ZnT-2 suggest that
that codes for ZIP-4, causing disrupted transport this transporter may be important for controlling
function and impaired zinc absorption. Patients with intracellular transport of zinc by the enterocyte [41].
acrodermatitis enteropathica need daily zinc supple- All three transporters are found primarily in intestinal
ments throughout life [35]. Because such supplements villus cells, and much less frequently in crypt cells.
alleviate the problems of patients with acrodermatitis As described for the ZIP proteins, individual mem-
enteropathica, it is obvious that there are other, but less bers of the ZnT family of transporters are located in
efficient, zinc transport mechanisms in the enterocyte. specific cell types. For example, in rats ZnT-1 is found
The ZIP-5 protein is expressed on the basolateral mem- mostly in the ileum, ZnT-2 is located primarily in the
brane of the enterocyte [36], where it may be responsi- duodenum and jejunum, and ZnT-4 is found through-
ble for zinc transport from the systemic circulation into out the small intestine [42]. ZnT-5, ZnT-6, and ZnT-7
the enterocyte, possibly as a means of drawing zinc into have been found to be involved in zinc homeostasis in
the intestinal cells when dietary zinc is low [27]. the pancreas, brain, and prostate, respectively [43–45];
Recently, ZIP-14 has been shown to be involved these transporters seem to be involved in zinc uptake in
in the uptake of zinc by the liver in response to acute the Golgi apparatus [45, 46]. ZnT-3 is localized to syn-
inflammation and infection [37]. Serum zinc concen- aptic vesicles in some types of neurons [47], and ZnT-8
tration falls during these conditions, whereas liver zinc is associated with the secretory granules of pancreatic
concentration increases, possibly in an attempt to with- beta-cells [48]. Thus, it is evident that the ZnT family is
hold zinc from pathogens. Liver ZIP-14 expression rises involved in multiple aspects of zinc homeostasis.
in response to the cytokine interleukin-6 (IL-6) during Several zinc transporters are involved in zinc secre-
the acute-phase response [37], suggesting that induction. ZnT-4, for example, has been shown to be involved
tion of ZIP-14 may be responsible for the hypozincemia in the secretion of zinc by the mammary gland, and
associated with infection. mutations of the gene cause the defect lethal milk in
The ZnT (“SLC30”) family of zinc transporters mice [49]. The milk of these animals has very low
zinc concentration, resulting in severe zinc deficiency future, however, to combine results from animal models
and high mortality among their pups. Several studies with data from compartmental modeling obtained in
have also shown that healthy, well-nourished lactating humans to unravel relevant information for under-
women can have abnormally low concentrations of standing human zinc requirements and developing new
zinc in their breastmilk [50]. Supplementation of these methods to assess zinc status.
women did not increase milk zinc concentrations, sug-
gesting a defect similar to that observed in mice with As discussed in the first IZiNCG technical docu-
the lethal milk defect, who are unable to secrete zinc ment [1], the recommended strategies to control zinc
into milk. A recent study of a family of women with low deficiency include supplementation, fortification, and
milk zinc contents, causing transient neonatal zinc defi- dietary diversification and modification. The present
ciency in their infants, showed that milk zinc secretion document reviews the current state of knowledge and
was impaired because of a mutation in ZnT-2 [51]. It is information gaps regarding each of these intervention
not yet known how common this type of mutation is in strategies. In particular, two papers discuss the available
lactating women or how often it causes zinc deficiency evidence regarding preventive zinc supplementation
in breastfed infants. among infants, preschoolers, and prepubertal chil-
The zinc transporters respond to conditions of low dren [53] and among pregnant and lactating women
or excessive zinc exposure and changes in zinc status, [54]. A third paper examines zinc supplementation as
presumably in an attempt to modulate their effects on adjunctive therapy in the treatment of diarrhea and
particular tissues and biological functions. For example, other diseases [55]. Another paper reviews the current
during zinc deficiency, the abundance of ZnT-1 protein state of knowledge concerning zinc fortification [56],
in the small intestine is reduced, decreasing endogenous and three separate papers describe the information
zinc losses, and the localization of ZIP-4 is changed to available on dietary diversification and modification
the entire villus, maximizing zinc uptake. In the liver, strategies. One of these latter papers focuses on general
ZnT-1 protein abundance is increased during zinc principles and approaches to dietary diversification
deficiency, most likely in an attempt to increase zinc in and modification [57], another describes the specific
the systemic circulation; however, liver zinc decreases, contribution of breastfeeding to maintaining adequate
as has been shown in animal studies [52]. Other tissues, zinc intakes among infants and young children [58],
such as the pancreas, muscle, and mammary gland, also and the third covers the potential of recently developed
respond to alterations in zinc status, but our knowledge biofortification approaches to improve zinc status [59].
regarding homeostasis in these tissues is more limited. The general conclusions of these reviews, their related
Furthermore, little is still known about the regulation of programmatic implications, and the most critical
zinc homeostasis and how different tissues contribute remaining research needs are summarized in the last
to this regulation in humans. It may be possible in the paper [60].
References
1. International Zinc Nutrition Consultative Group in developing countries: Pooled analysis of randomized
(IZiNCG), Brown KH, Rivera JA, Bhutta Z, Gibson controlled trials. Am J Clin Nutr 2000;72:1516–22.
RS, King JC, Lönnerdal B, Ruel MT, Sandtröm B, 5. World Health Organization/UNICEF. Clinical man-
Wasantwisut E, Hotz C. Assessment of the risk of zinc agement of acute diarrhoea. WHO/FCH/CAH/04.7.
deficiency in populations and options for its control. Geneva: WHO/UNICEF, 2004.
Food Nutr Bull 2004;25(1 suppl 2):S99–203. 6. Black RE, Allen LH, Bhutta ZA, Caulfield LE, de Onis
2. World Health Organization. Workshop to review the M, Ezzati M, Mathers C, Rivera J. Maternal and child
results of studies evaluating the impact of zinc supple- undernutrition: Global and regional exposures and
mentation on childhood mortality and severe morbidity. health consequences. Lancet 2008;371:243–60.
Conclusions and next steps. Geneva: WHO, 2006. 7. Ezzati M, Lopez AD, Rodgers A, Vander Hoorn S,
3. Zinc Investigators’ Collaborative Group, Bhutta ZA, Murray CJ. Selected major risk factors and global and
Black RE, Brown KH, Gardner JM, Gore S, Hidayat A, regional burden of disease. Lancet 2002;360:1347–60.
Khatun F, Martorell R, Ninh NX, Penny ME, Rosado 8. Jones G, Steketee RW, Black RE, Bhutta ZA, Morris SS.
JL, Roy SK, Ruel M, Sazawal S, Shankar A. Prevention How many child deaths can we prevent this year? Lancet
of diarrhea and pneumonia by zinc supplementation in 2003;362:65–71.
children in developing countries: Pooled analysis of ran- 9. Brown KH, Peerson JM, Rivera J, Allen LH. Effect
domized controlled trials. J Pediatr 1999;135:689–97. of supplemental zinc on the growth and serum zinc
4. Zinc Investigators’ Collaborative Group, Bhutta ZA, concentrations of prepubertal children: A meta-anal-
Bird SM, Black RE, Brown KH, Gardner JM, Hidayat A, ysis of randomized controlled trials. Am J Clin Nutr
Khatun F, Martorell R, Ninh NX, Penny ME, Rosado JL, 2002;75:1062–71.
Roy SK, Ruel M, Sazawal S, Shankar A. Therapeutic effects 10. de Benoist B, Darnton-Hill I, Davidsson L, Fontaine O.
of oral zinc in acute and persistent diarrhea in children Report of a WHO/UNICEF/IAEA/IZiNCG interagency
meeting on zinc status indicators, held in IAEA head- zinc. Biochim Biophys Acta 2006;1763:711–22.
quarters, Vienna, December 9, 2005. Food and Nutrition 28. Eide DJ. The SLC39 family of metal ion transporters.
Bulletin 2007;28:S399–S484. Pflugers Arch 2004;447:796–800.
11. de Benoist B, Darnton-Hill I, Davidsson L, Fontaine O, 29. Gaither LA, Eide DJ. The human ZIP1 transporter medi-
Hotz C. Conclusions of the joint WHO/UNICEF/IAEA/ ates zinc uptake in human K562 erythroleukemia cells.
IZiNCG interagency meeting on zinc status indicators. J Biol Chem 2001;276:22258–64.
Food Nutr Bull 2007;28:S480–S79. 30. Milon B, Dhermy D, Pountney D, Bourgeois M,
12. International Zinc Nutrition Consultative Group Beaumont C. Differential subcellular localization of
(IZiNCG). Assessing population zinc status with serum hZip1 in adherent and non-adherent cells. FEBS Lett
zinc concentration. IZiNCG Technical Brief No. 2. 2001;507:241–6.
Davis, CA: IZiNCG, 2007. 31. Wang F, Dufner-Beattie J, Kim BE, Petris MJ, Andrews
13. Hotz C, Peerson JM, Brown KH. Suggested lower cutoffs of G, Eide DJ. Zinc-stimulated endocytosis controls activity
serum zinc concentrations for assessing zinc status: Rean- of the mouse ZIP1 and ZIP3 zinc uptake transporters. J
alysis of the second National Health and Nutrition Exam- Biol Chem 2004;279:24631–9.
ination Survey data (1976–1980). Am J Clin Nutr 2003; 32. Franklin RB, Ma J, Zou J, Guan Z, Kukoyi BI, Feng P,
78:756–64. Costello LC. Human ZIP1 is a major zinc uptake trans-
14. Hess SY, Peerson JM, King JC, Brown KH. Use of serum porter for the accumulation of zinc in prostate cells. J
zinc concentration as an indicator of population zinc Inorg Biochem 2003;96:435–42.
status. Food Nutr Bull 2007;28:S403–29. 33. Kelleher SL, Lönnerdal B. Zip3 plays a major role
15. Brown KH. Effect of infections on plasma zinc con- in zinc uptake into mammary epithelial cells and
centration and implications for zinc status assess- is regulated by prolactin. Am J Physiol Cell Physiol
ment in low-income countries. Am J Clin Nutr 1998; 2005;288:C1042–7.
68:425S–9S. 34. Wang K, Zhou B, Kuo YM, Zemansky J, Gitschier J. A
16. Thurnham DI, Mburu AS, Mwaniki DL, De Wagt A. novel member of a zinc transporter family is defective
Micronutrients in childhood and the influence of sub- in acrodermatitis enteropathica. Am J Hum Genet
clinical inflammation. Proc Nutr Soc 2005;64:502–9. 2002;71:66–73.
17. Hotz C. Dietary indicators for assessing the ade- 35. Wang F, Kim BE, Dufner-Beattie J, Petris MJ, Andrews
quacy of population zinc intakes. Food Nutr Bull G, Eide DJ. Acrodermatitis enteropathica mutations
2007;28:S430–53. affect transport activity, localization and zinc-responsive
18. International Zinc Nutrition Consultative Group trafficking of the mouse ZIP4 zinc transporter. Hum
(IZiNCG). Determining the prevalence of zinc defi- Mol Genet 2004;13:563–71.
ciency: Assessment of dietary zinc intake. IZiNCG 36. Wang F, Kim BE, Petris MJ, Eide DJ. The mammalian
Technical Brief No. 3. Davis, CA: IZiNCG, 2007. Zip5 protein is a zinc transporter that localizes to the
19. Miller LV, Krebs NF, Hambidge KM. A mathematical basolateral surface of polarized cells. J Biol Chem
model of zinc absorption in humans as a function of 2004;279:51433–41.
dietary zinc and phytate. J Nutr 2007;137:135–41. 37. Liuzzi JP, Lichten LA, Rivera S, Blanchard RK, Aydemir
20. Fischer Walker CL, Black RE. Functional indicators TB, Knutson MD, Ganz T, Cousins RJ. Interleukin-6
for assessing zinc deficiency. Food Nutr Bull 2007; regulates the zinc transporter Zip14 in liver and contrib-
28:S454–79. utes to the hypozincemia of the acute-phase response.
21. World Health Organization. WHO growth standards: Proc Natl Acad Sci U S A 2005;102:6843–8.
Length/height-for-age, weight-for-age, weight-for- 38. McMahon RJ, Cousins RJ. Regulation of the zinc trans-
length, weight-for-height and body mass index-for-age: porter ZnT-1 by dietary zinc. Proc Natl Acad Sci U S A
Methods and development. Geneva: WHO, 2006. 1998;95:4841–6.
22. Gibson RS. Zinc: The missing link in combating micro- 39. Palmiter RD, Findley SD. Cloning and functional
nutrient malnutrition in developing countries. Proc characterization of a mammalian zinc transporter that
Nutr Soc 2006;65:51–60. confers resistance to zinc. EMBO J 1995;14:639–49.
23. International Zinc Nutrition Consultative Group 40. Murgia C, Vespignani I, Cerase J, Nobili F, Perozzi G.
(IZiNCG). Quantifying the risk of zinc deficiency: Cloning, expression, and vesicular localization of zinc
Recommended indicators. IZiNCG Technical Brief No. transporter Dri 27/ZnT4 in intestinal tissue and cells.
1. Davis, CA: IZiNCG, 2007. Am J Physiol 1999;277:G1231–9.
24. World Health Organization. Physical status: The use 41. Palmiter RD, Cole TB, Findley SD. ZnT-2, a mammalian
and interpretation of anthropometry. Report of a WHO protein that confers resistance to zinc by facilitating
expert committee. Technical Report Series No. 854. vesicular sequestration. EMBO J 1996;15:1784–91.
Geneva: WHO, 1995. 42. Liuzzi JP, Blanchard RK, Cousins RJ. Differential regula-
25. Sandström B, Arvidsson B, Cederblad Å, Björn-Ras- tion of zinc transporter 1, 2, and 4 mRNA expression by
mussen E. Zinc absorption from composite meals. I. dietary zinc in rats. J Nutr 2001;131:46–52.
The significance of wheat extraction rate, zinc, calcium, 43. Kambe T, Narita H, Yamaguchi-Iwai Y, Hirose J, Amano
and protein content in meals based on bread. Am J Clin T, Sugiura N, Sasaki R, Mori K, Iwanaga T, Nagao M.
Nutr 1980;33:739–45. Cloning and characterization of a novel mammalian zinc
26. Lee HH, Prasad AS, Brewer GJ, Owyang C. Zinc absorp- transporter, zinc transporter 5, abundantly expressed in
tion in human small intestine. Am J Physiol 1989; pancreatic beta cells. J Biol Chem 2002;277:19049–55.
256:G87–91. 44. Huang L, Kirschke CP, Gitschier J. Functional character
27. Eide DJ. Zinc transporters and the cellular trafficking of ization of a novel mammalian zinc transporter, ZnT6. J
Biol Chem 2002;277:26389–95. insulin-like growth factor-I, and the glucose/insulin

45. Kirschke CP, Huang L. ZnT7, a novel mammalian zinc axis in weanling rats. J Pediatr Gastroenterol Nutr
transporter, accumulates zinc in the Golgi apparatus. J 2005;41:72–80.
Biol Chem 2003;278:4096–102. 53. Brown KH, Peerson JM, Baker SK, Hess SY. Preven-
46. Huang L, Kirschke CP, Zhang Y, Yu YY. The ZIP7 gene tive zinc supplementation among infants, preschool-
(Slc39a7) encodes a zinc transporter involved in zinc ers, and older prepubertal children. Food Nutr Bull
homeostasis of the Golgi apparatus. J Biol Chem 2005; 2009;30:S12–40.
280:15456–63. 54. Hess SY, King JC. Effects of maternal zinc supplementa-
47. Wenzel HJ, Cole TB, Born DE, Schwartzkroin PA, tion on pregnancy and lactation outcomes. Food Nutr
Palmiter RD. Ultrastructural localization of zinc trans- Bull 2009;30:S60–78.
porter-3 (ZnT-3) to synaptic vesicle membranes within 55. Haider BA, Bhutta ZA. The effect of therapeutic zinc
mossy fiber boutons in the hippocampus of mouse and supplementation among young children with selected
monkey. Proc Natl Acad Sci U S A 1997;94:12676–81. infections: A review of the evidence. Food Nutr Bull
48. Chimienti F, Devergnas S, Favier A, Seve M. Identifica- 2009;30:41–59.
tion and cloning of a beta-cell-specific zinc transporter, 56. Hess SY, Brown KH. Impact of zinc fortification on zinc
ZnT-8, localized into insulin secretory granules. Diabe- nutrition. Food Nutr Bull 2009;30:S79–107.
tes 2004;53:2330–7. 57. Gibson RS, Anderson VP. A review of interventions
49. Huang L, Gitschier J. A novel gene involved in zinc based on dietary diversification or modification strate-
transport is deficient in the lethal milk mouse. Nat Genet gies with the potential to enhance intakes of total and
1997;17:292–7. absorbable zinc. Food Nutr Bull 2009;30:S108–43.
50. Atkinson SA, Whelan D, Whyte RK, Lönnerdal B. Abnor- 58. Brown KH, Engle-Stone R, Krebs NF, Peerson JM.
mal zinc content in human milk. Risk for development Dietary intervention strategies to enhance zinc nutri-
of nutritional zinc deficiency in infants. Am J Dis Child tion: Promotion and support of breastfeeding for infants
1989;143:608–11. and young children. Food Nutr Bull 2009;30:S144–71.
51. Chowanadisai W, Lönnerdal B, Kelleher SL. Identifi- 59. Hotz C. The potential to improve zinc status through
cation of a mutation in SLC30A2 (ZnT-2) in women biofortification of staple food crops with zinc. Food Nutr
with low milk zinc concentration that results in tran- Bull 2009;30:S172–78.
sient neonatal zinc deficiency. J Biol Chem 2006;281: 60. Brown KH, Baker SK, IZiNCG Steering Committee.
39699–707. Galvanizing action: Conclusions and next steps for
52. Hall AG, Kelleher SL, Lönnerdal B, Philipps AF. A graded mainstreaming zinc interventions in public health pro-
model of dietary zinc deficiency: Effects on growth, grams. Food Nutr Bull 2009;30:S179–84.
Preventive zinc supplementation among infants,
preschoolers, and older prepubertal children
Kenneth H. Brown, Janet M. Peerson, Shawn K. Baker, and Sonja Y. Hess
Abstract There were no significant effects on children’s behavioral

development, although the number of available studies
Zinc supplementation trials carried out among children is relatively small. The available evidence supports the
have produced variable results, depending on the specific need for intervention programs to enhance zinc status
outcomes considered and the initial characteristics of to reduce child morbidity and mortality and to enhance
the children who were enrolled. We completed a series child growth. Possible strategies for delivering preventive
of meta-analyses to examine the impact of preventive zinc supplements are discussed.
zinc supplementation on morbidity; mortality; physical
growth; biochemical indicators of zinc, iron, and copper
status; and indicators of behavioral development, along Key words: Children, growth, infants, iron status
with possible modifying effects of the intervention results. indicators, morbidity, mortality, prevention, zinc sup-
Zinc supplementation reduced the incidence of diarrhea plementation
by ~20%, but the impact was limited to studies that
enrolled children with a mean initial age greater than
12 months. Among the subset of studies that enrolled Background
children with mean initial age greater than 12 months,
the relative risk of diarrhea was reduced by 27%. Zinc A considerable number of intervention trials have been
supplementation reduced the incidence of acute lower conducted in a variety of settings to assess the impact
respiratory tract infections by ~15%. Zinc supplementa- of preventive zinc supplementation on children’s
tion yielded inconsistent impacts on malaria incidence, health and development. The results of these studies
and too few trials are currently available to allow defini- are inconsistent, possibly because of differences in
tive conclusions to be drawn. Zinc supplementation had a the underlying zinc status or other characteristics of
marginal 6% impact on overall child mortality, but there the study populations or discrepancies in the research
was an 18% reduction in deaths among zinc-supple- methods. In this paper, we examine the results of con-
mented children older than 12 months of age. Zinc sup- trolled supplementation trials to address the following
plementation increased linear growth and weight gain by questions:
a small, but highly significant, amount. The interventions
yielded a consistent, moderately large increase in mean Section 1: Does preventive zinc supplementation of
serum zinc concentrations, and they had no significant infants and young children affect their risk of selected
adverse effects on indicators of iron and copper status. illnesses, survival, physical growth, behavioral develop-
ment, and serum zinc concentration? Are these effects
modified by child- or dose-related factors?
Kenneth H. Brown, Janet M. Peerson, and Sonja Y. Hess are
affiliated with the Department of Nutrition and the Program Section 2: Are there adverse effects of preventive zinc
in International and Community Nutrition, University of supplementation?
California, Davis, California, USA; Kenneth H. Brown and
Shawn K. Baker are affiliated with Helen Keller International, Section 3: What are the opportunities to link preven-
Dakar, Senegal. tive zinc supplementation programs to existing health
H. Brown, Department of Nutrition, University of California, and nutrition programs, and what technical, social,
One Shields Ave., Davis, CA 95616, USA; e-mail: khbrown@ behavioral, and programmatic challenges must be
ucdavis.edu. confronted?
S12 Food and Nutrition Bulletin, vol. 30, no. 1 (supplement) © 2009, The United Nations University.
Preventive zinc supplementation in children S13
Section 1 in the meta-analyses, the analytic methods that were

used, and the specific outcomes of interest.
Does preventive zinc supplementation of infants and
young children affect their risk of selected illnesses, Identification of references. We sought information
survival, physical growth, behavioral development, and on controlled zinc supplementation trials conducted
serum zinc concentration? Are these effects modified by among prepubertal children by completing a compu-
child- or dose-related factors? terized bibliographic search in May 2007, using the
PubMed bibliographic database with the key word
Conclusions “zinc” and limiting for human studies, English lan-
guage, clinical trial, and randomized, controlled trials.
Preventive zinc supplementation reduces the incidence The results of the search were further expanded by
of diarrhea by ~20% among children in lower-income contacting experts in the field and examining subse-
countries, although current evidence indicates that this quent PubMed notifications and one conference report.
beneficial effect of zinc is limited to children greater The search strategy yielded a total of 1,625 individual
than ~12 months of age. Zinc supplementation also references for consideration (fig. 1).
lowers the incidence of acute lower respiratory tract
infections (ALRI), reducing pneumonia and ALRI by Selection of studies. The title or abstract of each article
~15%. Fewer studies have been completed to assess the was scanned by a research assistant and two of the
effects of zinc supplementation on the incidence and authors. Full articles were retrieved for further assess-
severity of malaria, but the limited available informa- ment if the available information suggested that zinc
tion suggests that zinc supplementation may reduce the was provided as a supplement (exclusive of infant
number of malaria episodes that result in clinic visits. formula), the presence or absence of zinc in the sup-
Overall, zinc supplementation produces a 6% reduc- plement was the only factor that differed between any
tion in child mortality, although this benefit may be two intervention groups, and zinc supplementation was
restricted to children 12 months of age or older, in whom provided for prevention of deficiency rather than for
the mortality reduction is approximately 18%. There is treatment of a current disease. Zinc supplementation
some information to suggest that zinc supplementation was considered to be therapeutic when it was provided
also may reduce mortality among small-for-gestational- as a component of the treatment regimen for diarrhea,
age (SGA) infants, but the number of available stud- pneumonia, malaria, or inpatient nutritional rehabilita-
ies and the numbers of children enrolled in each are tion of children with severe malnutrition (marasmus or
too small to allow definite conclusions to be drawn. kwashiorkor), and therefore studies of these conditions
Zinc supplementation produces a small, but signifi- were excluded from the present analysis. All other zinc
cant, increase in linear growth and weight gain. Zinc supplementation studies were considered preventive
supplementation consistently increases serum zinc zinc supplementation trials.
concentration, with a moderately large effect size. We A total of 95 references were identified from control-
did not find evidence of any overall impact of zinc led trials of preventive zinc supplementation in chil-
supplementation on mental or psychomotor develop- dren; 8 of these publications were excluded because the
ment. However, the number of available studies is still article described a prior meta-analysis [1, 2] or pooled
relatively small, and the duration of these studies may analysis [3], insufficient data were presented to address
be too short to permit detection of such outcomes. the questions of interest [4, 5], subjects were selected
Zinc supplementation programs should be consid- because of sickle-cell disease [6], or some subjects
ered for children in countries with an elevated risk of were no longer prepubertal [7, 8]. For two studies that
zinc deficiency to reduce their incidence of diarrhea, included both prepubertal and postpubertal individu-
pneumonia, and possibly other infections; reduce als, we were able to include the results just for the pre-
mortality among children 12 months of age or older pubertal children, either as presented in the paper [9]
and possibly among SGA infants; and increase growth or as provided subsequently by the authors [10].
velocity and thereby reduce their risk of nutritional The 87 acceptable articles were then screened to
stunting and underweight. combine results from those that presented data on the
same intervention trial by using key trial characteristics,
Detailed review of evidence such as the country site, supplementation scheme, and
study population. In some cases, several articles were
Overview published from the same study under the names of
To address the aforementioned set of questions, we different first authors, so we refer to individual studies
conducted a systematic review of relevant supplemen- by using the country site and year of first publication.
tation trials of infants and prepubertal children. The A total of 55 individual trials were identified, which
following sections describe the procedures used to enrolled a total of 202,692 children. If a trial included
identify individual studies and select those for inclusion more than two sets of treatment groups that differed
S14 K. H. Brown et al.
1,620 articles identified in PubMed; 1,538 articles rejected

5 articles from other sources
87 articles reporting on randomized,

controlled zinc-supplementation
trials in children
55 individual randomized trials
75 groupwise comparisons
Placebo vs. zinc (n = 35)

MMN vs. MMN + zinc (n = 9)
Iron vs. iron + zinc (n = 4)
Iron + folic acid vs. iron + folic acid + zinc (n = 1)
Iron + folic acid + vitamin A vs. iron + folic acid + vitamin A + zinc (n = 2)
Iron + vitamin A vs. iron + vitamin A + zinc (n = 1)
Iron + vitamin A + vitamin B2 vs. iron + vitamin A + vitamin B2 + zinc (n = 1)
Iron + vitamin A + vitamin C vs. iron + vitamin A + vitamin C + zinc (n = 1)
Iron + vitamin C vs. iron + vitamin C + zinc (n = 1)
Vitamin A vs. vitamin A + zinc (n = 10)
Vitamin A + vitamin B2 vs. vitamin A + vitamin B2 + zinc (n = 1)
Vitamin A + vitamin C vs. vitamin A + vitamin C + zinc (n = 1)
Vitamin B2 vs. vitamin B2 + zinc (n = 1)
Several B vitamins vs. several B vitamins + zinc (n = 3)
Vitamin C vs. vitamin C + zinc (n = 2)
Vitamin D vs. vitamin D + zinc (n = 1)
Calcium + vitamin A vs. calcium + vitamin A + zinc (n = 1)
FIG. 1. Number of articles and individual studies included in the meta-analysis

on preventive zinc supplementation in children. For groupwise comparisons,
the two treatment groups differed only by the presence or absence of zinc in
the supplement provided. MMN (multiple micronutrients) indicates at least
four micronutrients. For more information on characteristics of the studies
and participants, see table 1.
only by the presence of zinc (e.g., placebo vs. zinc the published report was obtained from the original
alone, iron vs. zinc plus iron, or multiple micronutri- author(s) of the article, if possible. One of the review-
ents (MMN) with or without zinc), each set of groups ers verified all extracted information by comparing the
that differed by zinc only was considered a separate data with the original publication. When there were
groupwise comparison (fig. 1). A total of 75 separate differences of opinion among reviewers, these were
controlled comparisons were identified. Of these com- discussed and resolved by consensus.
parisons, 73 were derived from studies considered to
be well designed because the treatments were randomly Data analyses. The outcomes examined were inci-
assigned to individuals and the research protocol used a dence of diarrhea; incidence of ALRI; incidence of
double-blind, controlled design. Only one study did not malaria; mortality; change in height (length or stature),
specifically indicate whether treatments were randomly expressed in centimeters or height-for-age z-score
assigned [11]. In the study Brazil 1998 [12, 13], one of (HAZ); change in body weight, expressed in kilograms
the three treatment groups (5 mg of zinc daily) was or weight-for-age z-score (WAZ); change in weight-for-
excluded from consideration because that group was height z-score; change in mid-upper-arm circumfer-
not enrolled concurrently with the placebo group. ence; change in serum or plasma zinc concentration;
final mental development index score; final physical
Data extraction and management. For studies that development index score; change in blood hemoglobin
fulfilled the inclusion criteria, three research assist- concentration; change in serum ferritin concentra-
ants summarized relevant information regarding the tion; and change in serum copper concentration. For
study population and intervention design, using a all outcomes, except morbidity variables and final
standard data-extraction template. Any relevant infor- developmental scores, studies were included in the
mation concerning the trial that was missing from analyses only if information was available for both the
children’s initial status and change during the course supplementation doses ranged from 1 to 70 mg per
of the intervention. Morbidity and mortality variables dose (median, 10 mg, with one dose unknown). These
were converted to rate ratios, and anthropometric, bio- doses were provided daily [9, 20–68], several times
chemical, and development variables were converted per week [10–13, 69–94], or once per week [94–99],
to effect size, which was calculated as the difference resulting in a daily dose equivalents ranging from 0.9
between the mean of the values for the zinc and the to 21.4 mg of zinc/day. Most studies provided zinc as
corresponding control group divided by their pooled zinc sulfate (n = 36), although a few distributed other
standard deviation. In general, effect sizes of ~0.2 are compounds, including zinc acetate (n = 5), zinc gluco-
considered of small magnitude, effect sizes of ~0.5 nate (n = 5), zinc amino acid chelates (n = 3), and zinc
are considered moderately large, and those of ~0.8 or oxide (n = 1). In four studies, the zinc compound was
greater are considered large [14]. not stated [22, 48, 71, 73], and in another study zinc
The overall mean effect size for each outcome vari- acetate was provided during the first phase of the study
able was estimated from a random-effects model [15]. and zinc gluconate was given later [11]. We attempted
This model assumes that the observed effect size or to evaluate the possible modifying effect of current
log(relative risk) from a particular study is the sum of breastfeeding on the response to zinc supplementation,
the true effect for that study plus a normally distributed but this was not possible because of the lack of relevant
random error term, which is related to the sample size information in the available reports.
and effect size or relative risk for that study and, in Selected initial characteristics of the subjects enrolled
turn, that the true effects are themselves normally dis- in the trials are presented in table 1. The mean initial
tributed. Because the total variance for the study effect age at enrollment varied greatly among studies. Some
size is different from one study to the next, the best studies enrolled infants within a few days after birth
estimate of the overall mean is a weighted mean effect [12, 27, 30, 78], whereas one study enrolled children
size, in which the weights are equal to the inverse of with a mean initial age of 11.1 years [10].
the total variance. The SAS for WINDOWS (release 9)
MIXED procedure was used to estimate the weighted Results
mean effect size and its standard error.
Additionally, the heterogeneity of responses was Diarrhea morbidity. Information on diarrhea incidence
assessed by using the chi-square test, as described by was available from 24 studies, which enrolled a total
Hedges [16]. We explored possible sources of heteroge- of 16,339 children. These studies provided 33 distinct
neity with random-effects meta-regression analyses, in comparisons of zinc supplements, with or without
which study characteristics were used to explain effect other nutrients, versus the same preparations with-
sizes [17, 18]. As with any regression, the number of out zinc. The treatment groups received just zinc or
possible explanatory variables was strictly limited by placebo in 16 comparisons. In five comparisons, both
the number of observations, which in this case is the groups also received iron, with or without vitamin A;
number of comparisons available. Explanatory vari- and in four comparisons, the treatment groups received
ables were examined separately in a series of bivariate vitamin A plus a micronutrient other than iron. Three
models; then a subset of explanatory variables was comparisons provided vitamin C or B vitamins, with or
entered into a regression model and nonsignificant without zinc. Five additional comparisons investigated
predictors were removed in a stepwise fashion. When MMN, with and without zinc. The mean age of the
appropriate, nonlinearity was initially assessed with study participants ranged from newborns to approxi-
polynomial models and, in one case (relation between mately 4 years.
diarrheal incidence and age), was followed up with There was a significant 20% lower incidence of
the use of a two-phase regression model. The SAS for diarrhea among children who received zinc supplemen-
WINDOWS MIXED procedure was used for all of these tation (relative risk, 0.80; 95% CI, 0.71 to 0.90; p = .0004,
procedures except the two-phase regression model, for random-effects model) (fig. 2). Because of significant
which the SAS NLMIXED procedure was used. heterogeneity among studies (p < .0001), a meta-regres-
sion analysis was completed. The mean initial age of the
Description of intervention trials study subjects was highly significantly associated with
the magnitude of the effect of zinc supplementation (p
The general characteristics of the studies included < .001), and the groupwise comparisons are displayed
in the meta-analyses are shown in table 1. Of the 55 by mean initial age in figure 3. Inspection of the figure
studies included in the analyses, 7 were from Africa, indicates that the beneficial effect of zinc supplements
23 were from Asia, 12 were from South America, 11 on diarrhea incidence was limited to studies of chil-
were from North America, 1 was from Australia, and 1 dren with a mean initial age greater than 12 months.
was from Europe. The supplementation periods ranged Among studies of children with mean age initial age
from 2 weeks [19] to 15 months [11], and the number greater than 12 months, the relative risk of diarrhea
of subjects ranged from 18 to 94,359. The periodic zinc incidence was 0.73 (95% CI, 0.61 to 0.87; p = .0014).
TABLE 1. Selected characteristics of double-blind, randomized, controlled trials in prepubertal children and study subjects for each group comparison S16
Supplementation scheme Mean initial characteristicsc
Serum zinc
Country, year Selection Dura- Zinc concen
[reference] criteria for study Sample tion dose Age tration
author populationa Group comparisonb sizec % male (mo) Frequency (mg) Other micronutrientsd (mo) (µg/dL) HAZ WAZ
Bangladesh, Unselected Placebo 325 50.5 0.46 Daily 20 None 23.7 73.5 –2.41 –2.35
2001 infants Zinc
[19–21] Rahman Vitamin A 328 55.5 0.46 Daily 20 200,000 IU vitamin Ae 23.7 70.9 –2.41 –2.42
Vitamin A + zinc
Bangladesh, Unselected Placebo 301 44.5 4.6 Daily 5 None 0.9 NA NA NA
2002 children Zinc
[22] Osendarp
[23] Hamadani
Bangladesh, Unselected Placebo 126 57.0 1.38 Daily 20 None 39.0 62.0 NA NA
2003a infants Zinc
[24] Albert Vitamin A 123 55.0 1.38 Daily 20 200,000 IU vitamin Ae 41.0 62.0 NA NA
Vitamin A + zinc
Bangladesh, Unselected Vitamin A + vitamin 318 44.4 6 Weekly 20 100,000 IU vitamin Ae; 6.3 67.9 –1.20 –1.00
2003b infantsf B2 1 mg vitamin B2
[95, 97] Baqui Vitamin A + vitamin
[96] Black B2 + zinc
Vitamin A + vitamin 327 44.9 6 Weekly 20 100,000 IU vitamin Ae; 6.3 66.7 –1.20 –1.05
B2 + iron 1 mg vitamin B2,
Vitamin A + vitamin 20 mg iron
B2 + iron + zinc
Bangladesh, Unselected Placebo 1,621 52.0 12 Weekly 70 None 5.3 64.5 –1.30 –1.62
2005 infants Zinc
[98] Brooks
Brazil, 1998 LBW infants Placebo 134 44.5 1.84 6/wk 1 None 0.0 NA NA NA
[12] Lira Zinc
[13] Ashworth
Brazil, 2000 Children with Placebo 18 50.0 6 Daily 42 None 118.2 100.5 –2.67 NA
[25] Sayeg Porto HAZ < –2 Zinc
Burkina Faso, Unselected pre- Placebo 685 49.5 6 6/wk 12.5 None 18.1 76.5 –1.55 –2.00
2001 schoolchildren Zinc
[69, 70] Müller
K. H. Brown et al.
Canada, 1989 Boys with HAZ Placebo 60 100.0 12 Daily 10 None 75.8 104.9 –1.39 –0.18
[26] Gibson < 15% Zinc
Chile, 1994 Children with Placebo 42 52.4 12 Daily 10 None 104.3 NA –2.42 NA
[9] Castillo- HAZ < 5% Zinc
Duran percentile
Chile, 1995 SGA infants MMN 68 47.1 6 Daily 3 1,500 IU vitamin A, 0.1 NA NA NA
[27] Castillo- MMN + zinc 50 mg vitamin C,
Duran 400 IU vitamin D;
1–2 mg/kg irong
Chile, 1997 Unselected pre- Placebo 98 50.0 14 Daily 10 None 39.8 114.1 –0.52 0.13
[28] Ruz schoolchildren Zinc
Chile, 2001 Infants with Placebo 112 50.9 12 Daily 5 1–2 mg/kg irong 0.3 NA NA NA
[29] Castillo- birthweight Zinc
Duran > 2,300 g
China, 1992 Infants of high- B vitamins 65 50.0 6 Daily 7.6 B vitaminsh 0.1 86.0 NA NA
[30] Hong risk pregnancy B vitamins + zinc

China, 1998 Unselected MMN 230 50.0 2.3 6/wk 20 2,500 IU vitamin A, 0.9 90.0 86.3 NA NA
[71] Sandstead children MMN + zinc mg vitamin B1, 1.1
[72] Penland mg vitamin B2, 12 mg
vitamin B3, 1.1 mg
vitamin B6, 35 µg folic
acid, 400 IU vitamin D,
7 mg vitamin E, 20 µg
vitamin K, 1 mg copper,
20 µg selenium, 90 µg
iodine, 1 mg fluoride,
1.5 mg manganese, 30
µg molybdenum, 30 µg
chromium
China, 2002 Preschoolchil- Placebo 61 49.2 12 5/wk 3.5 None 49.1 NA NA NA
[73] Yang dren with Zinc
HAZ < –1 Vitamin A + calcium 55 50.9 12 200 µg vitamin A, 150 mg 47.1 NA NA NA
5/wk 3.5
Vitamin A + calcium calcium
+ zinc
Ecuador, 1994 Unselected pre- Placebo 96 60.0 15 6/wk 10 None 31.5 74.3 –2.90 –1.76
[74] Dirren schoolchildren Zinc
Ecuador, 1996 Preschoolchil- Placebo 48 56.0 2 Daily 10 None 42.3 86.5 –2.00 –1.40
[31] Sempertegui dren with Zinc
WAZ < 10th
percentile or
HAZ < 10th
percentile
S17
continued
TABLE 1. Selected characteristics of double-blind, randomized, controlled trials in prepubertal children and study subjects for each group comparison (continued) S18
Serum zinc
Ecuador, 2008i Nonanemic Placebo 251 53.1 6 Daily 3 None 21.1 71.6 –2.3 –1.3
[32] Wuehler preschool- Zinc (3 mg)
children with
HAZ < –1.3 Placebo 253 53.1 6 Daily 7 None 21.0 71.7 –2.3 –1.3
for children Zinc (7 mg)
12–23 mo of
age, < –1.5 for Placebo 253 53.1 6 Daily 10 None 20.9 72.0 –2.3 –1.3
children 24–30 Zinc (10 mg)
Ethiopia, 2000 Stunted infants Placebo 90 53.3 6 6/wk 10 None 9.6 NA –2.81 –2.58
[75] Umeta with HAZ < –2 Zinc (stunted)
Nonstunted Placebo 94 46.8 6 6/wk 10 None 9.3 NA –0.64 –1.40
infants Zinc (nonstunted)
matched by
age and sex
France, 1992 Breastfed infants Vitamin D 57 52.7 3 Daily 5 Vitamin Dh 5.5 NA 0.12 0.76
[33] Walravens Vitamin D + zinc
Gambia, 1993 Unselected pre- Placebo 109 50.0 15 2/wk 70 None 17.7 NA NA NA
[11] Bates schoolchildren Zinc
Guatemala, 1993 Unselected MMN 162 50.0 5.75 6/wk 10 1.5 mg vitamin B1, 1.2 mg 81.8 93.5 –1.38 –0.85
[76] Cavan children MMN + zinc vitamin B2, 20 mg vita-
[77] Grazioso min B3, 10 mg vitamin
B5, 1 mg vitamin B6, 6
µg vitamin B12, 100 µg
folic acid, 100 mg vita-
min C, 10 µg vitamin
D, 3.3 mg vitamin E, 50
µg copper, 2 mg chro-
mium, 110 µg iodine,
50 µg selenium, 110 mg
magnesium
Guatemala, 1997 Unselected Placebo 89 57.1 7 Daily 10 None 7.6 NA –2.16 –1.18
[34] Ruel infants Zinc
K. H. Brown et al.
[35] Bentley
[36] Rivera
India, 1996 Preschoolchil- MMN 609 52.3 6 Daily 10 240 RE vitamin A, 0.6 mg 16.0 64.8 NA NA
[37–42] Sazawal dren with MMN + zinc vitamin B1, 0.5 mg vita-
diarrhea in min B2, 10 mg vitamin
past 24 h B3, 0.5 mg vitamin B6,
100 IU vitamin D, 3 mg
vitamin E
India, 2001 SGA infants Vitamin B2 584 50.0 9 Daily 5 0.5 mg vitamin B2 0.5 NA NA NA
[43] Sazawal Vitamin B2 + zinc
[44] Black
MMN 570 50.0 9 Daily 5 0.5 mg vitamin B2, 60 0.5 NA –1.80 NA
MMN + zinc µmol folic acid, 180 mg
calcium, 90 mg phos-
phorus, 10 mg iron
India, 2002 Unselected pre- Vitamin A 2,482 52.4 4 Daily 20j 200,000 IU vitamin Ae,j 15.3 62.0 NA NA
[45, 47] Bhandari schoolchildren Vitamin A + zinc

[46] Taneja
India, 2003ai Unselected pre- Placebo 189 46.1 3.68 5/wk 10 None 23.5 NA NA NA
[94] Gupta schoolchildren Daily zinc
Placebo 185 46.0 3.68 Weekly 50 None 23.5 NA NA NA
Weekly zinc
India, 2003b LBW infants B vitamins 100 50.0 12 5/wk 4.5 B vitaminsh 0.1 114.1 NA –2.14
[78] Sur B vitamins + zinc
India, 2007a Unselected Folic acid + iron 94,359 52.7 12 Daily 10k 12.5 mg iron,k 50 µg folic 11.7 64.1 NA NA
[48] Bhandari infants and Folic acid + iron + acidk
preschoolchil- zinc
dren
India, 2007b Unselected pre- B vitamins 1,712 49.0 6 Weekly 49.3 B vitaminsh NA NA NA NA
[99] Gupta schoolchildren B vitamins + zinc
Indonesia, 2001 Unselected Placebo 238 50.0 6 5/wk 10 None 4.2 NA –0.79 –0.05
[79] Dijkhuizen infants Zinc
[80, 81] Wieringa Iron 240 50.0 6 5/wk 10 10 mg iron 4.2 NA –0.90 –0.06
Iron + zinc
Vitamin A 129 50.0 6 5/wk 10 2.4 mg β-carotene 4.2 NA NA NA
Vitamin A + zinc
Indonesia, 2003 Unselected Vitamin C 336 53.0 6 Daily 10 30 mg vitamin C 6.2 59.3 –0.37 –0.39
[49, 50] Lind infantsf Vitamin C + zinc
Vitamin C + iron 330 50.5 6 Daily 10 30 mg vitamin C, 10 mg 6.2 58.6 –0.32 –0.39
Vitamin C + iron + iron
zinc
S19
continued
Serum zinc
Indonesia, 2007 Unselected Vitamin A 391 49.9 6 Daily 10 100,000 IU vitamin Ae 5.1 100.0 –0.99 –0.54
[51] Fahmida infantsf Vitamin A + zinc
Jamaica, 1998 Preschoolchil- MMN 61 42.6 2.76 Daily 5 1,500 IU vitamin A, 0.5 14.1 NA –2.85 NA
[52] Meeks dren with MMN + zinc mg vitamin B1, 0.8 mg
Gardner HAZ < –2 vitamin B2, 7 mg vita-
and WAZ < min B3, 1 mg vitamin
median B6, 30 mg vitamin C,
400 IU vitamin D
Jamaica, 2005 Preschoolchil- MMN 114 33.2 6 Daily 10 1,500 IU vitamin A, 0.5 18.8 NA –1.42 –2.16
[53] Meeks dren with MMN + zinc mg vitamin B1, 0.8 mg
Gardner WAZ < –1.5 vitamin B2, 7 mg vita-
min B3, 1 mg vitamin
B6, 2 µg vitamin B12,
1 mg folic acid, 30 mg
vitamin C, 400 IU vita-
min D, 8 mg iron
Mexico, 1997 Unselected pre- Placebo 109 48.2 12 6/wk 20 None 28.7 89.8 –1.71 –1.40
[82] Rosado schoolchildren Zinc
[83] Allen Iron 108 45.0 12 6/wk 20 20 mg iron 28.2 103.7 –1.55 –1.40
[84] Munoz Iron + zinc
Mexico, 2005 Unselected Placebo 252 56.8 6 5/wk 30 None 84.0 81.6 NA NA
[85] Kordas childrenl Zinc
[86] Rosado
Iron 265 56.2 6 5/wk 30 30 mg iron 84.0 77.6 NA NA
[87] Rico
Iron + zinc
Mexico, 2006 Unselected Placebo 364 49.5 12 Daily 20 None 9.9 NA 0.08 0.09
[54] Long infants Zinc
Vitamin A 372 54.0 12 Daily 20 20,000 IU vitamin A for 9.7 NA 0.13 0.10
Vitamin A + zinc ≤ 1 yr of age; 45,000 IU
for > 1 yr of age every
2 mo
K. H. Brown et al.
Nepal, 2006 Unselected Vitamin A 13,385 50.0 13.7 Daily 10j 200,000 IU vitamin Aj 12.4 NA NA NA
[55, 56] Tielsch infants and Vitamin A + zinc every 6 mo
preschoolchil- Vitamin A + folic 17,079 50.0 13.7 Daily 10j 50 µg folic acid j, 12.5 mg 12.4 NA NA NA
dren acid + iron iron j; 200,000 IU vita-
Vitamin A + folic min Aj every 6 mo
acid + iron + zinc
Papua New Unselected pre- Placebo 274 47.0 10.58 6/wk 10 None 31.4 70.5 –1.90 NA
Guinea, 2000 schoolchildren Zinc
[88] Shankar
Peru, 2004a Anemic pre- Iron 223 50.0 4.14 6/wk 7.3 3 mg/kg/day iron 17.4 NA –1.04 –0.25
[89] Alarcon schoolchildren Iron + zinc
(hemoglobin
70–99.9 g/L)
Peru, 2004b Children with Vitamin C 159 50.3 6 Daily 10 50 mg vitamin C 18.9 70.3 –1.56 –1.13
[57] Penny persistent Vitamin C + zinc
diarrhea (> 14
days)
Peru, 2007 Infants with MMN 200 48.5 6 Daily 3 225 µg RE vitamin A, 0.5 7.5 77.6 –1.19 –0.75
[58] Brown LAZ < −0.5 MMN + zinc mg vitamin B1, 0.38 mg
and WLZ > −3 vitamin B2, 3.8 mg vita-
min B3, 2.5 mg vitamin
B5, 0.5 mg vitamin B6,
50 µg biotin, 20 mg vita-
min C, 225 IU vitamin
D, 3.8 mg vitamin E, 0.7
mg ironm
South Africa, HIV-positive MMNn 96 48.9 6 Daily 10 1,000 IU vitamin A, 1.5 38.3 NA –1.60 NA
2005 preschoolchil- MMN + zincn mg vitamin B1, 1.2 mg
[59] Bobat dren vitamin B2, 10 mg vita-
min B3, 1 mg vitamin
B6, 50 mg vitamin C,
400 IU vitamin D
Tanzania, 2006 Unselected Vitamin A 42,546 50.3 12.7 Daily 10j 200,000 IU vitamin Aj 18.1 NA –1.48 –1.28
[60, 61] Sazawal infants and Vitamin A + zinc every 6 mo
[62] Olney preschoolchil-
Vitamin A + folic 16,070 50.5 12.7 Daily 10j 200,000 IU vitamin Aj 18.1 NA –1.45 –1.15
dren
acid + iron every 6 mo; 50 µg folic
Vitamin A + folic acid j, 12.5 mg iron j
acid + iron + zinc
S21
continued
Serum zinc
Thailand, 1992 Infants with Placebo 68 47.4 6 5/wk 25 None 110.5 86.3 NA NA
[90] Udom serum retinol Zinc
kesmalee concentration Vitamin A 65 69.5 6 25 1,500 RE vitamin A 113.0 85.3 NA NA
5/wk
[91] Kramer < 1.05 μmol/L Vitamin A + zinc
and serum zinc
concentration
< 12.2 μmol/L
Thailand, 2006 Unselected Vitamin A + vita- 304 51.5 6 Daily 10 1,500 RE vitamin A e, 4.5 73.8 –0.69 –0.18
[63] Wasantwisut infantsf min C 30 mg vitamin C
Vitamin A + vita-
min C + zinc
Vitamin A + vitamin 305 50.0 6 Daily 10 1,500 RE vitamin A e, 4.5 71.0 –0.66 –0.12
C + iron 30 mg vitamin C,
Vitamin A + vitamin 10 mg iron
C + iron + zinc
Uganda, 1998 Unselected Placebo 153 54.1 8 3.75/wk 10 None 55.8 NA –0.70 –0.41
[92] Kikafunda preschool and Zinc
school-aged
children
USA, 1983 Preschoolchil- Placebo 40 65.0 12 Daily 10 None 50.0 72.0 –2.07 –1.76
[64] Walravens dren with Zinc
HAZ < 10th
percentile
USA, 1989 Preschoolchil- Placebo 50 52.0 6 Daily 5.7 None 15.2 70.0 –1.35 –2.04
[65] Walravens dren with Zinc
documented
decline of ≥ 20
percentile in
WAZ
USA, 2006 Breastfed Placebo 82 50.0 6 Daily 5 None 4.0 73.3 0.37 0.61
[66] Heinig infants with Zinc
birthweight >
2,500 g
K. H. Brown et al.
Vietnam, 1996 Preschoolchil- Placebo 146 50.0 5 Daily 10 None 17.6 NA –2.91 –2.61
[67] Ninh dren with Zinc
WAZ < –2 and
HAZ < –2
Vietnam, 2006 Unselected Vitamin A 393 50.4 6 Daily 10 100,000 IU vitamin Ae 5.8 94.8 –1.01 –0.56
[68] Berger infantsf Vitamin A + zinc
Vitamin A + iron 391 47.7 6 Daily 10 100,000 IU vitamin Ae; 5.9 92.7 –1.06 –0.58
Vitamin A + iron 10 mg iron
+ zinc
Zimbabwe, Unselected Placebo 313 46.0 12.2 3.5/wk 40 None 133.8 77.8 –1.18 –1.27
1997 children Zinc
[10, 93] Friis

HAZ, height-for-age z-score; HIV, human immunodeficiency virus; LAZ, length-for-age z-score; LBW, low-birthweight; MMN, multiple micronutrients with at least four micronutrients; RE, retinol
equivalent; SGA, small-for-gestational age; WAZ, weight-for-age z-score; WLZ, weight-for-length z-score; NA, not available
a. Only major selection criteria are listed here. A study population is considered “unselected” if the inclusion and exclusion criteria were such that almost all screened participants were eligible to participate
in the study.
b. The study treatment groups included into a group comparison analysis differed only by the presence or absence of zinc in the supplement.
c. Total sample size and mean initial characteristics of both study groups combined, which were included into a group comparison.
d. All nutrients listed were provided to both study groups within a group comparison
e. Single-dose vitamin A was provided in all studies at baseline, except for Bangladesh 2001 [19, 20], in which it was provided at day 14.
f. Selected for breastfeeding, but more than 90% of the population was eligible for the study. The study population is therefore considered representative and defined as “unselected.”
g. Iron supplementation after 4 months [27] and 5 months [29] of age.
h. No information on amount of micronutrients provided.
i. If a study included several zinc groups (different dosage or frequency), each groupwise comparison includes the placebo group and the respective zinc group.
j. Half the dose for children less than 12 months of age.
k. Half the dose for children less than 6 months of age.
l. Children with blood lead levels greater than 45 µg/dL and hemoglobin less than 90 g/L were excluded. However, 99.8% of the children screened were eligible and are therefore considered representative
and defined as “unselected.”
m. Children were also provided iron in iron-fortified cereal porridge separate from an aqueous multivitamin dose (containing zinc in the zinc group).
n. Most children received multivitamin supplement and cotrimoxazole.
S23
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Besides mean age, the magnitude Bangladesh 2002 [22] (n=299)
of the effect of zinc supplementa- Bangladesh 2003b [95] (n=318)
tion on the incidence of diarrhea Bangladesh 2003b Fe [95] (n=327)

Bangladesh 2005 [98] (n=1,474)
was negatively associated with Burkina Faso 2001 [69] (n= 661)
the baseline anthropometric Chile 2001 [29] (n=112)
status of the study population China 2002 [73] (n=55)
(initial height, p = .033; initial China 2002 calcium + vitamin A [73] (n=50)
weight, p = .032) and positively Ecuador 2008 3mg [32] (n=233)
Ecuador 2008 7mg [32] (n=232)
associated with the mean initial
Ecuador 2008 10mg [32] (n=234)
serum ferritin concentration (p Ethiopia 2000 non-stunted [75] (n=94)
= .036). However, there were no Ethiopia 2000 stunted [75] (n=90)
significant correlations between Guatemala 1997 [34] (n=89)
the daily zinc dose, inclusion India 1996 [39] (n=579)
of other micronutrients in the India 2002 [47] (n=2,464)
India 2003a Daily [94] (n=189)
supplement preparation, or the
India 2003a Weekly [94] (n=185)
duration of supplementation and India 2003b [78] (n=100)
the impact of supplementation India 2007b [99] (n=1,712)
on diarrhea incidence. We also Jamaica 1998 [52] (n= 61)
examined whether methodologic Mexico 1997 [82] (n=110)
issues, such as the duration of Mexico 1997 iron [82] (n=109)
Mexico 2006 [54] (n=364)
recall for illness reporting and
Mexico 2006 vitamin A [54] (n=372)
whether the reports were based Nepal 2006 [56] (n= 4,070)
on specific signs of diarrhea or Nepal 2006 iron + folic acid [55] (n=1,200)
parental perceptions of illness, Peru 2004a [89] (n=213)
affected the conclusions. We Peru 2004b [57] (n=159)
found that there was no relation Peru 2007 [58] (n=198)

South Africa 2005 [59] (n=96)
between the specific study meth-
USA 2006 [66] (n=70)
ods and the effect size of the zinc Vietnam 1996 [67] (n=146)
response.
Data were available on the All (n=16,665)
duration of diarrhea from just –3 –2 –1 0 1 2
nine studies, which provided 13 Lower incidence Greater incidence
groupwise comparisons among in zinc group in zinc group
a total of 1,692 children. The ln (relative risk)

mean age of the study partici- FIG. 2. Effect of zinc supplementation on diarrhea incidence from 24 intervention
pants ranged from 6 to 29 months. trials with 33 groupwise comparisons in which the supplements differed only by
Unlike what has been reported the presence or absence of zinc.
previously from diarrhea treat-
ment studies [100], there was no rates based on the physicians’ examinations were the
significant effect of preventive zinc supplementation ones included in the analysis. We also considered a
on the duration of diarrhea in these community-based second tier of studies that reported ALRI based on
trials (effect size, 0.041; 95% CI, −0.216 to 0.299; p = .73, rapid breathing or difficulty breathing, as reported by
random-effects model). the caregiver. Information was available from seven
studies based on the former, objective criteria [32, 41,
Respiratory disease morbidity. Analyses related to respi- 45, 57–59, 98] and from five studies based on reported
ratory disease were restricted to studies that provided symptoms only [22, 54–56, 95].
information on the incidence of ALRI, using either The combined set of studies yielded a total of 16
the World Health Organization (WHO) definition of treatment comparisons from 12 studies with a total of
ALRI, based on age-specific elevated respiratory rates 12,144 subjects. The children’s mean initial age ranged
[101], or clinical (auscultory or radiologic) evidence from 0.9 to 49 months. Zinc supplements were com-
of pneumonia, as defined by the authors. When data pared with placebo in six treatment group comparisons.
were reported for elevated respiratory rates, both with Three comparisons were of MMN, with and without
and without associated severity signs, such as cough, zinc, one of vitamin C, with or without zinc, and six
difficulty breathing, fever, or lethargy, the illness rates of vitamin A with other micronutrients, such as iron,
based on the more severe degree of ALRI were used iron and folic acid, or vitamin B2, with and without
preferentially. Likewise, when information was avail- zinc. Overall, there was a significant 15% reduction in
able from both fieldworkers and physicians, illness ALRI (relative risk, 0.85; 95% CI, 0.75 to 0.97; p = .017,
-3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5
0
0 group (relative risk, 0.99; 95% CI, 0.91 to 1.08; p = .78,
random-effects model). When both factors (initial HAZ
and diagnostic rigor) were included in the explanatory
6 models, only HAZ remained statistically significant.
10 Malaria morbidity. The effects of zinc supplementa-

tion on the risk of malaria were examined in the first
12
technical document prepared by the International Zinc
Nutrition Consultative Group (IZiNCG) [102]. At that
time, the results of just three intervention trials were
18 available, two of which found 32% (Gambia 1993 [11])
20
and 38% (Papua New Guinea 2000 [88]) reductions
Mean age (mo)
in clinic visits for malaria, and one of which found

24 no impact on the incidence of cases detected by daily
home visits (Burkina Faso 2001 [69]). The former
IZiNCG review concluded that zinc supplementation
30 may ameliorate the severity of malaria infections, hence
30
reducing the number of clinic visits, possibly without
affecting the overall incidence of infections. However,
the number of available trials was too small to allow
36 definitive conclusions to be drawn.
Since then, only two new relevant studies have
40 become available, neither of which fulfilled the inclu-
42
sion criteria for the present review. A study in the
Peruvian Amazon enrolled children from 0.5 to 15
years of age, some of whom exceeded the age range
established for the present review [8]. Children who
48 received either zinc or zinc plus iron had ~15% fewer
50
–3 –2 –1 0 1 2
episodes of Plasmodium vivax infections, as assessed
Lower incidence Greater incidence by twice-weekly home visits, compared with the pla-
in zinc group in zinc group cebo group, although the results were not statistically
ln (relative risk) significant (p > .36). However, there was a significant
FIG. 3. Effect of zinc supplementation on diarrhea incidence, interaction between age group and treatment group,
according to mean initial age of study subjects in each trial. such that, among children less than 5 years of age,
The curve represents ln(risk ratio) = − 0.081 for age less than those who received zinc without iron had an incidence
15 months, − .081 − .032*(age − 15) for age greater than 15 rate ratio (IRR) of 0.43 (95% CI, 0.17 to 1.10; p = .079),
months and those who received zinc with iron had an IRR of
0.30 (95% CI, 0.12 to 0.80; p = .016), compared with
random-effects model) (fig. 4). There was significant the placebo group. However, among children aged 5
heterogeneity among studies (p = .008). The two factors years or older, there was no significant effect of zinc
that were significantly associated with the magnitude alone or zinc plus iron. In another study recently
of reduction of relative risk of ALRI following zinc completed in Burkina Faso, children were randomly
supplementation were the initial height-for-age z-score assigned to receive either daily zinc supplements plus
(HAZ) (p = .010) and the quality of ALRI diagnosis a single large dose of vitamin A or placebo supple-
(p = .024). Specifically, studies that enrolled children ments [103]. There was a 22% lower rate of fever in
who initially were more stunted found a greater impact the supplemented group, as diagnosed during daily
of zinc supplements on ALRI reduction, as did those home visits, and a 30% reduction in malaria incidence,
studies that relied on more rigorous diagnostic criteria. as determined during clinic visits. However, because
The relative risk for those studies that diagnosed ALRI of the intervention design, which included both
based on counting respiratory rate or a physician’s zinc and vitamin A, it was not possible to determine
examination was 21% less in the zinc group than in the whether the results were uniquely attributable to the
comparison group (relative risk, 0.79; 95% CI, 0.67 to zinc supplements.
0.94; p = .013, random-effects model). In contrast, the In summary, there is still insufficient evidence to
studies that based the diagnosis only on reported rapid allow definitive conclusions to be drawn regarding the
breathing or difficulty breathing (without a physician’s effect of zinc supplementation on the risk of malaria,
examination) found no significant difference between although the weight of currently available information
the group that received zinc and the comparison suggests that zinc may reduce the incidence of malaria,
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Bangladesh 2005 [98] (n=1474)
Ecuador 2008 3 mg [32] (n=233)
Ecuador 2008 7 mg [32] (n=232)
Ecuador 2008 10 mg [32] (n=234)
India 1996 [41] (n=578)
India 2002 [45] (n=2220)
Peru 2004b [57] (n=159)
Peru 2007 [58] (n=200)
Subgroup (n=5,426)
Bangladesh 2002 [22] (n=299)

Bangladesh 2003b [95] (n=318)
Bangladesh 2003b iron [95] (n=327)
Mexico 2006 [54] (n=364)
Mexico 2006 vitamin A [54] (n=372)
Nepal 2006 [56] (n= 4,070)
Nepal 2006 iron + folic acid [55] (n=1,200)
Subgroup (n= 6,950)
All (n=12,376)
–3 –2 –1 0 1 2
Lower incidence Greater incidence
in zinc group in zinc group
ln (relative risk)
FIG. 4. Effect of zinc supplementation on the incidence of acute lower respiratory tract
infection (ALRI)a from 12 intervention trials with 16 groupwise comparisons in which
the supplements differed only by the presence or absence of zinc.
a. Gray circles indicate studies in which ALRI was diagnosed by fieldworkers or physicians by using
objective clinical signs; black circles indicate studies in which the diagnosis was based on caregiver
reports of elevated respiratory rate or difficulty breathing
especially that of more severe cases that result in clinic statistical power to detect small differences in mor-
attendance. tality outcomes, so the results may be susceptible to
publication or reporting bias. Some of the studies also
Mortality. Thirteen pertinent groupwise comparisons provided a single high-dose vitamin A supplement at
of mortality outcomes were available from 10 studies. baseline [45] or every 6 months during the study period
Seven of these studies were carried out in unselected [55, 56, 60, 61]. Although the distribution of high-dose
study populations [45, 48, 55, 56, 60, 61, 69, 88, 98], vitamin A supplements was not reported in the other
one included only low-birthweight infants [12], one studies [48, 69, 88, 98], children may have received such
included only SGA infants [43], and one enrolled supplements as part of ongoing national programs.
only children with human immunodeficiency virus Overall, there were 1,407 deaths among the 100,081
(HIV) infection [59]. Three of the group comparisons children in the control groups (1.41%) and 1,328 deaths
completed among unselected children were from among the 101,535 children in the zinc-supplemented
large-scale studies carried out in Tanzania 2006 (n = groups (1.31%). The estimated relative risk of mortality
16,070 [60]), Nepal 2006 (n = 17,079 [55]), and India was 0.94 (95% CI, 0.86 to 1.02; p = .11, random-effects
2007a (n = 78,346 [48]), in which zinc plus iron and model) (fig. 5). There was significant heterogeneity in
folic acid was compared with iron and folic acid only, the results (p = .005), but the number of studies was
and two were from the same studies in Tanzania 2006 too small to explore systematically the specific sources
(n = 42,546 [61]) and Nepal 2006 (n = 25,018 [56]), in of heterogeneity.
which zinc was compared with placebo. Four smaller Because of the heterogeneity among studies and
studies also compared mortality outcomes following the fact that the results are dominated by the larger
supplementation with zinc or placebo (Bangladesh trials in Nepal 2006 [55, 56], Tanzania 2006 [60, 61],
2005, n = 1,474 [98]; India 2002, n = 2,482 [45]; Papua and India 2007a [48] (with five groupwise compari-
New Guinea 2000, n = 274 [88]; Burkina Faso 2001, sons), we reexamined the outcomes for specific age
n = 685 [69]) in unselected children. These latter subgroups presented within these three larger trials
studies were not originally designed with sufficient and whether or not iron and folic acid were provided
Bangladesh 2005 [98] (n=1,474)

Brazil 1998 [12] (n=134)
Burkina Faso 2001 [69] (n= 685)
India 2001 [43] (n=584)
India 2001 MMN [43] (n=570)
India 2007a [48] (n=78,346)
Nepal 2006 [56] (n= 41,276)
Nepal 2006 iron + folic acid [55] (n=17,079)
Papua New Guinea 2000 [88] (n=274)
Tanzania 2006 [61] (n= 42,546)
Tanzania 2006 iron + folic acid [60] (n=16,070)
All (n=20,1616)
–3 –2 –1 0 1 2
Lower mortality Greater mortality
ln (relative risk)
FIG. 5. Effect of zinc supplementation on childhood mortality from 10 intervention trialsa with 13
groupwise comparisons in which the supplements differed only by the presence or absence of zinc.
a. The figure does not include the study by Bhandari et al. [45] (India, 2002; n = 2,482) because this study had no
deaths in the zinc-supplemented group
along with zinc. The authors of these trials graciously of zinc on mortality in either age group. The combined
provided the results of their respective studies disag- results are summarized in table 2 by age group and
gregated by age group (< 12 months or ≥ 12 months) treatment group.
for the comparisons of zinc versus placebo and zinc In summary, when the results of these studies are
plus iron and folic acid versus iron and folic acid. We combined, zinc supplementation reduced mortality of
modeled the mortality data using mixed models with children 12 months of age or older by ~18% but had
log relative risk as the outcome variable, with possible no effect on younger children. However, when iron and
explanatory variables age group (as defined above), iron folic acid were provided in addition to zinc, the impact
and folic acid treatment, study (a random effect), and of zinc among older children was no longer evident.
their interactions. Notably, there was a significant (p = The remaining studies either enrolled only younger
.04) interaction between age and supplementation with children or did not present the results disaggregated
iron and folic acid, such that when zinc supplements by age, so it is not possible to explore this issue further
were compared with placebo, there was a significantly with the available information.
lower mortality rate among the older children who Among the studies of selected study populations, two
did not receive iron and folic acid as compared with enrolled low-birthweight [12] or SGA infants [43]. In
the other three groups. When iron and folic acid were the study Brazil 1998, low-birthweight infants received
provided along with zinc, there was no significant effect zinc or placebo [12], whereas the study India 2001
TABLE 2. Effect of supplementation with zinc only or zinc plus iron and folic acid on risk of death among children < 12 months
or ≥ 12 months of age: Combined analyses of results from three large-scale trials (with five groupwise comparisons)
[48, 55, 56, 60, 61]
Age Mortality
group Sample Child- rate per 1,000 Relative
(mo) Group comparisona size yr Deaths child-yr risk 95% CI p
< 12 Zinc 27,440 15,328 385 25.1 1.05 0.91–1.21 .52
Placebo 26,974 14,951 360 24.1
≥ 12 Zinc 14,802 43,595 332 7.6 0.82 0.70–0.96 .013
Placebo 14,606 43,343 406 9.4
< 12 Folic acid + iron + zinc 32,859 23,410 349 14.9 0.97 0.82–1.15 .72
Folic acid + iron 32,456 23,205 352 15.2
≥ 12 Folic acid + iron + zinc 38,441 34,681 242 7.0 1.05 0.90–1.24 .52
Folic acid + iron 37,721 33,977 230 6.8
a. Refers to comparisons of treatment groups that differed only by the presence or absence of zinc. (i.e., zinc versus placebo or zinc plus iron
and folic acid versus iron plus folic acid).
included two group comparisons in which SGA infants available from 37 studies, which contained 47 group-
received vitamin B2 with and without zinc or MMN wise comparisons. The mean initial HAZ ranged from
with or without zinc [43]. Although the sample sizes −2.9 [67] to 0.36 [66], and the mean initial age ranged
were relatively small, both studies found 52% to 68% from less than 1 month [12] to 134 months [10]. There
lower mortality rates among children who received zinc was a significantly greater change in height among chil-
(Brazil 1998, p = .33; India 2001, p = .04). These results dren who received zinc supplements, with an overall
are consistent with the analyses by birthweight in Nepal effect size of 0.170 (95% CI, 0.075 to 0.264; p = .001,
2006, where infants with birthweight less than 2,000 g random-effects model) (fig. 6). There was significant
who received zinc had a relative risk of mortality that heterogeneity among studies (p < .0001). The effect size
was nearly half that of their counterparts who did not for change in height was negatively correlated with con-
receive zinc (relative risk, 0.56; 95% CI, 0.30 to 1.04; current administration of iron (p = .04) and vitamin A
p = .06). These combined sets of results indicate that supplements (p = .04). Unlike the results of a previous
providing preventive zinc supplementation in settings meta-analysis of the effect of zinc supplementation on
where there is an elevated risk of zinc deficiency would children’s growth [1], which found a positive response
reduce mortality among children greater than 1 year to zinc only among those studies that enrolled children
of age and possibly among low-birthweight infants. whose initial mean HAZ was less than approximately
However, additional studies are needed to confirm −1.5 z, there was no correlation between mean initial
these two sets of results. HAZ and effect size in the present analysis, even when
Physical growth. Information on change in height was the analysis was restricted to the subset of studies that
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Bangladesh 2001 [20] (n=325)
Bangladesh 2001 vitamin A [20] (n=328)
Bangladesh 2002 [22] (n=271)
Bangladesh 2005 [98] (n=576)
Brazil 1998 [12] (n=112)
Brazil 2000 [25] (n=18)
Burkina Faso 2001 [70] (n= 661)
Canada 1989 [26] (n= 60)
Chile 1994 [9] (n= 42)
Chile 1995 [27] (n= 68)
Chile 1997 [28] (n=53)
China 1992 [30] (n= 65)
China 2002 [73] (n= 61)
China 2002 calcium + vitamin A [73] (n=55)
Ecuador 1994 [74] (n=96)
Ecuador 1996 [31] (n= 48)
Ecuador 2008 3 mg [32] (n=211)
Ecuador 2008 7 mg [32] (n=208)
Ecuador 2008 10 mg [32] (n=218)
Ethiopia 2000 non-stunted [75] (n=94)
Ethiopia 2000 stunted [75] (n=90)
France 1992 [33] (n=57)
Gambia 1993 [11] (n=97)
Guatemala 1993 [76] (n=156)
Guatemala 1997 [36] (n=89)
India 2001 MMN [44] (n=150)
India 2003b [78] (n=100)
Indonesia 2003 [49] (n=326)
Indonesia 2003 iron [49] (n=324)
Jamaica 1998 [52] (n=57)
Mexico 1997 [82] (n=95)
Mexico 1997 iron [82] (n=99)
Peru 2004a [89] (n=213)
Peru 2004b [57] (n=146)
Peru 2007 [58] (n=175)
Tanzania 2006 [62] (n=102)
Tanzania 2006 iron + folic acid [62] (n=110)
Thailand 2006 [63] (n=304)
Thailand 2006 iron [63] (n=305)
Uganda 1998 [92] (n=113)
USA 1983 [64] (n= 40)
USA 1989 [65] (n=50)
USA 2006 [66] (n=70)
Vietnam 1996 [67] (n=146)
Vietnam 2006 [68] (n=386)
Vietnam 2006 iron [68] (n=384)
Zimbabwe 1997 [10] (n=191)
All (n=7,945)
–1.0 –0.5 0 0.5 1.0 1.5 2.0

Less growth Greater growth
Effect size
FIG. 6. Effect of zinc supplementation on change in height in prepubertal children from 37 control-
led supplementation trials with 47 groupwise comparisons in which the supplements differed only
by the presence or absence of zinc.
lasted at least 6 months or to those studies that enrolled subjects.

children with a mean initial age less than 3 years. This Twenty-two studies, with 30 groupwise comparisons,
difference from the earlier meta-analysis may be due provided information on the effect of zinc supplemen-
to the exclusion of studies of hospitalized, severely tation on change in weight-for-height z-score (WHZ).
malnourished children from the present analysis. There was a small, marginally significant, positive
Thirty-five studies presented sufficient information effect of zinc on change in WHZ (fig. 8). The estimated
to permit assessment of the effect of zinc supplementa- effect size was 0.062 (95% CI, 0.000 to 0.123; p = .049,
tion on the change in weight from baseline to the end random-effects model), and there was no significant
of the intervention. These studies provided 45 group- heterogeneity among studies (p = .28).
wise comparisons, and the mean initial weight-for-age There were 11 studies and 14 groupwise comparisons
z-score (WAZ) ranged from −2.61 [67] to 0.76 [33]. of the effect of zinc supplementation on change in mid-
Zinc supplementation had a significant positive overall upper-arm circumference. Zinc supplementation did
impact on change in weight, with a mean effect size of not have a significant effect on change in mid-upper-
0.119 (95% CI, 0.048 to 0.190; p = .002, random-effects arm circumference (data not presented here).
model) (fig. 7). There was significant heterogeneity In summary, zinc supplementation produced a small,
among studies (p < .001). The effect size for change but highly statistically significant, positive impact on
in weight was negatively correlated with concurrent children’s linear growth and weight gain and a marginal
administration of iron supplements (p = .002), but not effect on weight-for-height. There was significant het-
with any other characteristics of the studies or study erogeneity in the results of studies of growth velocity,
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Bangladesh 2001 [20] (n=325)
Bangladesh 2002 [22] (n=271)
Bangladesh 2005 [98] (n=527)
Brazil 1998 [12] (n=112)
Burkina Faso 2001 [70] (n= 661)
Canada 1989 [26] (n= 60)
Chile 1994 [9] (n= 42)
Chile 1995 [27] (n= 68)
Chile 1997 [28] (n=53)
China 1992 [30] (n= 65)
China 2002 [73] (n= 61)
China MMN 2002 calcium + vitamin A [73] (n=55)
Ecuador 1994 [74] (n=96)
Ecuador 1996 [31] (n= 48)
Ecuador 2008 3 mg [32] (n=210)
Ecuador 2008 7 mg [32] (n=206)
Ecuador 2008 10 mg [32] (n=215)
France 1992 [33] (n=57)
Gambia 1993 [11] (n=103)
Guatemala 1993 [76] (n=156)
Guatemala 1997 [36] (n=89)
India 2001 MMN [44] (n=150)
Indonesia 2003 [49] (n=326)
Jamaica 1998 [52] (n=57)
Mexico 1997 [82] (n=95)
Mexico 1997 iron [82] (n=99)
Peru 2004a [89] (n=213)
Peru 2004b [57] (n=146)
Peru 2007 [58] (n=175)
Tanzania 2006 [62] (n=102)
Thailand 2006 [63] (n=304)
Uganda 1998 [92] (n=113)
USA 1983 [64] (n= 40)
USA 1989 [65] (n=50)
USA 2006 [66] (n=70)
Vietnam 1996 [67] (n=146)
Vietnam 2006 [68] (n=386)
Zimbabwe 1997 [10] (n=191)
All (n=7,778)
–1.0 –0.5 0 0.5 1.0 1.5 2.0

Less weight gain Greater weight gain
Effect size
FIG. 7. Effect of zinc supplementation on change in weight in prepubertal children from 35 sup-
plementation trials with 45 groupwise comparisons in which the supplements differed only by the
presence or absence of zinc.
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Bangladesh 2001 [20] (n=325)
Bangladesh 2005 [98] (n=570)
Burkina Faso 2001 [70] (n= 661)
Canada 1989 [26] (n= 60)
Chile 1997 [28] (n=53)
Ecuador 1994 [74] (n=96)
Ecuador 1996 [31] (n= 48)
Ecuador 2008 3 mg [32] (n=210)
Ecuador 2008 7 mg [32] (n=206)
Ecuador 2008 10 mg [32] (n=215)
Guatemala 1993 [76] (n=156)
Guatemala 1997 [36] (n=89)
Indonesia 2003 [49] (n=326)
Jamaica 1998 [52] (n=57)
Mexico 1997 [82] (n=95)
Mexico 1997 iron [82] (n=99)
Peru 2004a [89] (n=213)
Peru 2004b [57] (n=145)
Peru 2007 [58] (n=175)
Thailand 2006 [63] (n=304)
USA 1983 [64] (n= 40)
USA 2006 [66] (n=70)
Vietnam 2006 [68] (n=386)
Zimbabwe 1997 [10] (n=97)
All (n= 6,221)
–1.0 –0.5 0 0.5 1.0 1.5 2.0

Less change Greater change
Effect size
FIG. 8. Effect of zinc supplementation on change in weight-for-height z-score

(WHZ) in prepubertal children from 22 controlled supplementation trials with 30
groupwise comparisons in which the supplements differed only by the presence
or absence of zinc.
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Bangladesh 2001 [21] (n=168)

Bangladesh 2002 [22] (n=251)
Bangladesh 2005 [98] (n= 463)
Chile 1995 [27] (n= 68)
Chile 1997 [28] (n= 63)
China 1992 [30] (n= 64)
China 1998 [71] (n=218)
Ecuador 1994 [74] (n=96)
Ecuador 1996 [31] (n= 48)
Ecuador 2008 3 mg [32] (n=94)
Ecuador 2008 7 mg [32] (n=93)
Ecuador 2008 10 mg [32] (n=102)
Guatemala 1993 [76] (n=145)
India 1996 [42] (n= 447)
Indonesia 2003 [50] (n=277)
Indonesia 2007 [51] (n=59)
Mexico 1997 [82] (n=101)
Mexico 1997 iron [82] (n=93)
Mexico 2005 [86] (n=234)
Mexico 2005 iron [86] (n=241)
Peru 2004b [57] (n=134)
Peru 2007 [58] (n=153)
Thailand 1992 [90] (n= 68)
Thailand 2006 [63] (n=124)
Zimbabwe 1997 [10] (n=29)
All (n= 4,571)
–1.0 –0.5 0 0.5 1.0 1.5 2.0 2.5 3.0

Effect size
FIG. 9. Effect of zinc supplementation on change in serum or plasma zinc concentration in
children from 22 controlled supplementation trials with 30 groupwise comparisons in which the
supplements differed only by the presence or absence of zinc.
but the source of heterogeneity generally could not For the present analyses, we only considered studies
be identified, except for a negative association with that reported information on the mental development
concurrent supplementation of either iron or vitamin index (MDI) or psychomotor development index
A for change in height and a negative association with (PDI), using the Bayley Scales. Most studies did not
concurrent iron supplementation for change in weight. present intraindividual changes in developmental
There was no overall effect of zinc supplementation on scores during the course of the intervention, so only
mid-upper-arm circumference measurements. final values could be compared. Final MDI and PDI
Serum or plasma zinc concentration. Information on values were reported from seven studies that provided
the change in serum or plasma zinc concentration was nine groupwise comparisons. The study duration
available from 22 intervention trials consisting of 30 ranged from 1.9 to 12 months, and just two studies
groupwise comparisons (fig. 9). As in previous meta- lasted more than 6 months. Two comparisons evaluated
analyses [1, 104], there was a consistent, moderately the impact of zinc supplementation versus placebo [13,
large, statistically significant positive effect of zinc 23], and the others provided additional micronutrients,
supplementation on the change in serum zinc concen- such as iron [29, 49, 96] or vitamin A [46, 96], to both
tration, with an overall effect size of 0.602 (95% CI, groups. None of the studies found a significant positive
0.439 to 0.766; p < .0001, random-effects model). The effect of zinc on final MDI (fig. 10). The overall esti-
daily zinc dose equivalents ranged from 2.9 to 21.4 mg mated effect size was 0.021 (95% CI, −0.133 to 0.175;
of zinc/day, and the studies lasted from 2 weeks to 14 p = .76, random-effects model). There was marginally
months. There was significant heterogeneity of results significant heterogeneity among studies (p = .065), and
(p < .001), but the source of heterogeneity could not there was a significant association between effect size
be identified. for final MDI and the percentage of males enrolled in
Mental and motor development. The available studies the individual studies (p = .024).
that reported on children’s developmental outcomes in As with MDI, there was no significant overall impact
relation to zinc supplementation varied greatly with of zinc supplementation on final PDI (fig. 11). The
regard to their developmental assessment methods. estimated effect size was 0.025 (95% CI, –0.149 to 0.198,
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Bangladesh 2002 [23] (n=198)
Brazil 1998 [13] (n=92)
Chile 2001 [29] (n=109)
India 2001 MMN [44] (n=162)
India 2002 [46] (n=571)
Indonesia 2003 [49] (n=332)
All (n=1,973)
–1.0 –0.5 0 0.5 1.0 1.5 2.0

Lower final MDI Greater final MDI
Effect size
FIG. 10. Effect of zinc supplementation on final mental development index (MDI) among infants
and young children from seven intervention trials with nine groupwise comparisons in which the
supplements differed only by the presence or absence of zinc.
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Bangladesh 2002 [23] (n=198)

Brazil 1998 [13] (n=92)
Chile 2001 [29] (n=109)
India 2001 MMN [44] (n=156)
India 2002 [46] (n=571)
Indonesia 2003 [49] (n=332)
All (n=1,967)
–1.0 –0.5 0 0.5 1.0 1.5 2.0

Lower final PDI Greater final PDI
Effect size
FIG. 11. Effect of zinc supplementation on final psychomotor development index (PDI) in infants
and young children from seven zinc supplementation trials with nine groupwise comparisons in
which the supplements only by the presence or absence of zinc.
p = 0.75, random-effects model). There was significant provided simultaneously in aqueous solutions and in
heterogeneity among studies (p = .013), but there were disproportionate molar doses [106]. However, there is
no significant correlations between study or subject no evidence of interference when they are delivered
characteristics and effect size. in near isomolar amounts or with food [107]. Some
longer-term studies also suggest that when given
together each mineral may reduce the magnitude
Section 2 of the response observed with single-nutrient sup-
plementation [50, 68, 79], although nutritional status
Are there adverse effects of preventive zinc supplemen is still enhanced to a considerable extent despite the
tation? nutrient–nutrient interactions [108]. Less information
is available with regard to interactions between zinc
Conclusions and copper, but some studies have found a negative
effect of large-dose zinc supplementation on indicators
According to the previous studies that have been used of copper status in adults [109, 110].
to define the safe upper level of zinc intake [105], the Because some studies have noted negative effects of
first signs of excessive intake are perturbations of zinc supplementation on the absorption or status of
copper and iron metabolism, resulting in impaired other minerals, we completed a systematic analysis of
status of these nutrients. Thus, we have reviewed the overall impact of preventive zinc supplementation
available studies that examined the impact of zinc sup- trials on indicators of children’s iron status (namely,
plementation on indicators of iron and copper status. hemoglobin and serum ferritin concentrations) and
There are no overall adverse effects of zinc supplemen- copper status (serum copper concentration). Studies
tation on concentrations of hemoglobin, serum ferritin, were identified by using the same strategy described
and serum copper. above in Section 1.
Hemoglobin and iron status. A total of 11 studies,
Detailed review of evidence which included 19 groupwise comparisons, provided
information on the change in hemoglobin concentra-
A number of studies have examined the effects of zinc tion following zinc supplementation. The daily dose
supplementation on iron absorption and vice versa, equivalents for those 19 sets of observations ranged
either by using isotopic tracers during short-term from 2.9 to 21.4 mg of zinc/day. Iron supplements were
studies to assess mineral absorption or by assess- also provided in eight of these groupwise comparisons
ing biochemical and functional responses following [50, 58, 62, 63, 83, 86, 89, 97]. Considering all of the
longer-term supplementation. The tracer studies indi- available information, there is no overall effect of zinc
cate that each mineral may interfere to some extent supplementation on change in hemoglobin concen-
with absorption of the other, but only when they are tration (fig. 12). The estimated mean effect size was
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0

Ecuador 2008 3 mg [32] (n=101)
Ecuador 2008 7 mg [32] (n=94)
Ecuador 2008 10 mg [32] (n=106)
Indonesia 2003 [50] (n=301)
Indonesia 2007 [51] (n=303)
Mexico 1997 [83] (n=99)
Mexico 1997 iron [83] (n=103)
Mexico 2005 [86] (n=252)
Mexico 2005 iron [86] (n=265)
Peru 2004a [89] (n=213)
Peru 2004b [57] (n=134)
Peru 2007 [58] (n=172)
Tanzania 2006 [62] (n=102)
Thailand 2006 [63] (n=124)
All (n=3,160)
–1.0 –0.5 0 0.5 1.0 1.5 2.0

Effect size
FIG. 12. Effect of zinc supplementation on change in hemoglobin concentration

among children from 11 controlled zinc supplementation trials with 19 groupwise
comparisons in which the supplements differed only by the presence or absence
of zinc.
0.019 (95% CI, −0.132 to 0.170; p = .80, random-effects (p = .024), the mean initial hemoglobin concentration
model). There was significant heterogeneity among (p = .018), and the mean initial ferritin concentration
studies (p < .0001), but no particular characteristics of (p = .019).
the studies or the study subjects were associated with Copper status. Four studies involving eight groupwise
the magnitude of hemoglobin response. In particular, comparisons supplied results on the change in serum
neither the daily zinc dose nor the presence of iron in copper concentration following zinc supplementation
the supplement was correlated with effect size of the (fig. 14). There was no overall effect of zinc supplemen-
change in hemoglobin concentration due to zinc. tation on the change in serum copper concentration.
Similarly, there was no overall effect of zinc supple- The estimated effect size was −0.041 (95% CI, −0.213
mentation on the change in serum or plasma ferritin to 0.131; p = .59, random-effects model), and the daily
concentration among the 17 available groupwise com- zinc dose was not correlated with the change in serum
parisons derived from 10 studies (fig. 13), 7 of which copper concentration. However, it should be recog-
also provided iron [50, 58, 63, 83, 86, 89, 97]. The esti- nized that serum copper concentration is a relatively
mated effect size was 0.051 (95% CI, −0.150 to 0.252; insensitive biomarker of copper status [111]. It is pos-
p = 0.60, random-effects model). There was significant sible that more subtle changes in copper metabolism
heterogeneity among comparisons (p < .0001); the may have occurred, although such changes, if they
magnitude of the change in serum ferritin concentra- did occur, would be unlikely to have any functional
tion in relation to zinc supplementation was negatively significance.
correlated with the presence of iron in the supplement
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Ecuador 2008 3 mg [32] (n=71)
Ecuador 2008 7 mg [32] (n=78)
Ecuador 2008 10 mg [32] (n=110)
Indonesia 2003 [50] (n=277)
Indonesia 2007 [51] (n=58)
Mexico 1997 [83] (n=96)
Mexico 1997 iron [83] (n=98)
Mexico 2005 [86] (n=220)
Mexico 2005 iron [86] (n=232)
Peru 2004a [89] (n=213)
Peru 2004b [57] (n=134)
Peru 2007 [58] (n=144)
Thailand 2006 [63] (n=124)
All (n=2,415)
–1.0 –0.5 0 0.5 1.0 1.5 2.0

Effect size
FIG. 13. Effect of zinc supplementation on change in serum or plasma ferritin

concentration in children from 10 controlled zinc supplementation trials including
17 groupwise comparisons in which the supplements differed only by the presence
or absence of zinc.
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0

Ecuador 2008 3 mg [32] (n=94)
Ecuador 2008 7 mg [32] (n=94)
Ecuador 2008 10 mg [32] (n=103)
Mexico 2005 [86] (n=200)
Mexico 2005 iron [86] (n=210)
Peru 2007 [58] (n=154)
All (n=1,017)
–1.0 –0.5 0 0.5 1.0 1.5 2.0

Effect size
FIG. 14. Effect of zinc supplementation on change in serum copper concentra-

tion in children from four controlled zinc supplementation trials with eight
groupwise comparisons in which the supplements differed only by the presence
or absence of zinc.
Section 3 Twice-yearly vitamin A supplementation (VAS). Glo-

bally, the most successful micronutrient supplementa-
What are the opportunities to link preventive zinc sup- tion program for children less than 5 years of age is
plementation programs to existing health and nutrition VAS, which is increasingly integrated into twice-yearly
programs, and what technical, social, behavioral, and events for child survival (combining such interventions
programmatic challenges must be confronted? as deworming, vaccinations, distribution of insecticide-
treated bednets, etc.) [115]. It is estimated that 79%
Available evidence regarding the impact of preven- of children 6 to 59 months of age in sub-Saharan
tive zinc supplementation of infants and children Africa and 71% of children 6 to 59 months of age in
on morbidity (diarrhea, ALRI, and perhaps malaria), South Asia received at least one dose of vitamin A in
mortality in children greater than 12 months of age 2005 [116]. A recent publication that describes the
and possibly SGA infants, and physical growth argues progress and future directions of twice-yearly VAS in
for the need to develop programs to prevent zinc defi- West and Central Africa [117] documents the success
ciency in those countries where an elevated risk of zinc of such programs and calls for institutionalizing the
deficiency has been identified. There is no evidence child health day approach to deliver VAS and other
of adverse effects of preventive zinc supplementation low-cost, high-impact services for child survival and
on markers of iron and copper status, indicating that development. VAS programs have been very effective in
zinc supplements can be delivered safely, either alone reaching children 12 to 59 months of age, although they
or with other micronutrients. The challenges for scal- have been somewhat less successful in reaching infants
ing up zinc supplementation programs are similar to 6 to 11 months of age [118]. This platform probably
those faced by other programs that attempt to procure offers the most promising avenue for rapid scale-up of
and distribute nutritional supplements or medicines, delivery of preventive zinc products, but a number of
as discussed below. issues must be addressed:
It has been stated previously that zinc needs to be » What duration of dosing will caregivers be able to
provided on a daily basis for an extended period of administer correctly if the supplement supply is
time [102], although one study found equivalent ben- delivered only once every 6 months?
eficial effects when supplemental zinc was provided » What combination of zinc dose and duration of sup-
weekly [94]. In either case, the likely need for frequent plementation will result in optimal improvement in
administration of zinc supplements presents a number zinc status when delivered at 6-month intervals?
of programmatic challenges related to product delivery » What is the optimal presentation of the product (sup-
over an extended period of time and ensuring compli- plement, powder, paste) to maximize compliance and
ance. The most common, currently existing supple- minimize costs and logistical burden?
mentation program requiring daily dosing and high » Can existing twice-yearly VAS programs support
compliance is iron and folic acid supplementation for the additional input and logistical costs of adding
pregnant and lactating women. The main operational preventive zinc supplementation?
constraints to successful delivery of such supplements » What communication strategies are required during
have been described elsewhere [112] and include twice-yearly events and as follow-up to these events
procurement and distribution of supplements, limited to support optimal compliance by caregivers?
access to and poor utilization of health services by the » Twice-yearly VAS programs only need to address
target population, inadequate training and motivation coverage, since doses are consumed at delivery. Com-
of frontline health workers, inadequate counseling of pliance will be essential for effective preventive zinc
target recipients or their caregivers, and low compli- programs. How will programs be able to monitor and
ance of the intended beneficiaries. These are common evaluate compliance?
obstacles that will need to be addressed by any supple- » What is the effectiveness of these programs?
mentation program, including programs that distribute Growth monitoring and promotion (GMP). GMP pro-
potential products such as tablets, powders, and pastes, grams could be ideal platforms for delivering preven-
as discussed below. In addition, there are generic issues tive zinc supplements, because such programs provide
of introduction of any new product, which include the frequent contacts with young children, thereby allow-
regulatory environment, quality assurance and control, ing for delivery of zinc-containing products, counseling
costs, supply chain and storage, product acceptability on their use, and monitoring of compliance. A recent
and packaging [113, 114]. review of GMP programs concluded that these pro-
The following section examines existing delivery grams should “maximize their potential, strengthen the
platforms that can be tapped for distribution of zinc nutrition counseling elements, [and] combine growth
supplements and discusses issues that need to be monitoring with other health interventions” [119].
addressed to deliver preventive zinc supplements suc- Preventive zinc supplementation is certainly one such
cessfully, either alone or in multiple-micronutrient health intervention that could easily meld with GMP
products. activities. A particular advantage of such programs is
that they provide routine contacts that can be exploited level of zinc necessary in the context of specific diets
to ensure delivery of supplements over an extended to result in adequate improvement in zinc status [125].
period and to promote compliance. If such products are to include iron, there are several
Community-based or community-directed distribu- other issues that need to be addressed to ensure safety
tion programs. Various community-based distribution [126], although a full review of these issues is beyond
systems exist in which the supply system is an extension the scope of this paper
of the health services. In these systems, either health Reaching low-birthweight infants. Low-birthweight
workers visit communities to renew and supervise dis- infants have multiple special nutritional needs. They
tribution of supplies or community distributors report have a greater risk of breastfeeding difficulties and have
to the health center to renew stocks. For example, com- an elevated risk of iron deficiency [127, 128]. Thus,
munity-directed treatment with ivermectin (CDTI) is programs to address the special health and nutritional
active in 26 countries in sub-Saharan Africa to control needs of low-birthweight infants would have a scope
onchocerciasis. Community-directed distributors are far broader than just preventive zinc supplementation,
chosen by the community to provide once-yearly treat- although preventive zinc supplementation should be
ment. The scope of CDTI is being expanded to include included as a key component. Although birthweight is
elimination of lymphatic filariasis and delivery of other often measured in clinical settings, this information is
services [120, 121]. Another example is provided by seldom used to provide a special package of interven-
traditional birth attendants, who have been trained tions to meet the needs of low-birthweight infants.
to deliver a variety of services, including distribution Issues that would have to be addressed in designing a
of iron and folic acid tablets to pregnant and lactating strategy to target these infants would include:
women [122]. The issues of integrating preventive zinc » Systematic identification of low-birthweight infants
products are similar to those described for GMP pro- (ensuring accurate weighing at delivery for births
grams, particularly if the program has ongoing contact attended by trained personnel; integrating weighing
with the intended beneficiaries. For programs such as at the first contact for other infants, for example,
CDTI that only have intermittent contact operations, through the expanded program on immunization,
research is needed to see whether these systems can be etc.);
expanded to deliver products on an ongoing basis. » Definition of the minimum package of low-birth-
Social marketing. This strategy is increasingly used weight infant care;
to deliver products through commercial channels or » Identifying community and health service contacts
messages to intended beneficiaries. A 1992 review that can be mobilized to deliver the package;
defines social marketing as “a broader, systematic » Monitoring and evaluation of delivery of and compli-
approach to developing strategies to define acceptable ance with the package.
concepts, behaviors, or products, to promote them, and In summary, the available evidence on the impact of
in the case of products, to distribute and price them for preventive zinc supplementation supports the need for
the market. A complete social marketing strategy not intervention programs to enhance zinc status. There
only develops and promotes a good ‘product,’ but also are a number of available opportunities to deliver
achieves and maintains political support and trains and preventive zinc supplementation as one component
motivates program implementers” [123]. Prices are often of programs to prevent MMN deficiencies and to
subsidized or programs have cross-subsidies to enhance address other nutrition and health needs of infants
reach to low-socioeconomic groups. In addition to and children. Efforts are needed to test these delivery
issues cited for other delivery strategies, a specific issue mechanisms and evaluate their potential for providing
is the extent to which such approaches reach the poorest cost-effective preventive zinc supplementation to high-
and most remote beneficiaries and how well relevant risk target groups on a large scale.
messages on dosing issues can be communicated.
Point-of-use fortificants. There has been rapid devel-
opment of point-of-use fortificants, including powders Acknowledgments
(often called “Sprinkles”), dispersible or crushable
tablets, and lipid-based nutrient supplements, which We greatly appreciate the contributions of research
are designed to address deficiencies in MMN and assistants Reina Engle-Stone, Josh Jorgensen, and K.
sometimes essential fatty acids and proteins [124]. The Ryan Wessells, who participated in the bibliographic
programmatic issues of their delivery are not substan- search and data extraction for the meta-analyses. We
tially different from delivery of a supplement, and it is further acknowledge the following scientists and their
assumed that the above-mentioned platforms could collaborators for providing additional information on
be used. These products have the added advantage their respective studies N. Bhandari (India, 2007a),
of favoring a delivery strategy that copromotes the T. Lind (Indonesia, 2003), S. Sazawal (India, 2001), S.
product and optimal infant and young child feeding Sazawal (Tanzania, 2006), and E. Wasantwisut (Thai-
practices. An issue specific to zinc is determining the land 2006).
References
1. Brown KH, Peerson JM, Rivera J, Allen LH. Effect of 14. Cohen J. Statistical power analysis for the behavioral
supplemental zinc on the growth and serum zinc con- sciences, 2nd ed. Hillsdale, NJ, USA: Lawrence Erlbaum
centrations of prepubertal children: A meta-analysis Associates, 1988.
of randomized controlled trials. Am J Clin Nutr 2002; 15. Wolf F. Meta-analysis: Quantitative methods for research
75:1062–71. synthesis, Newbury Park, CA, USA: Sage Publications,
2. Brown KH, Peerson JM, Allen LH. Effect of zinc sup- 1986.
plementation on children‘s growth: A meta-analysis of 16. Hedges LV. Estimation of effect size from a series of
intervention trials. Bibl Nutr Dieta 1998;54:76–83. independent experiments. Psychol Bull 1982;92:490–9.
3. Zinc Investigators’ Collaborative Group, Bhutta ZA, 17. Greenland S. Quantitative methods in the review of
Black RE, Brown KH, Gardner JM, Gore S, Hidayat A, epidemiologic literature Epidemiol Rev 1987;9:1–30.
Khatun F, Martorell R, Ninh NX, Penny ME, Rosado 18. Petitti DB. Meta-analysis, decision analysis, and cost-
JL, Roy SK, Ruel M, Sazawal S, Shankar A. Prevention effective analysis: Methods for quantitative synthesis in
of diarrhea and pneumonia by zinc supplementation in medicine. New York: Oxford University Press, 1994.
children in developing countries: Pooled analysis of ran- 19. Rahman MM, Vermund SH, Wahed MA, Fuchs GJ,
domized controlled trials. J Pediatr 1999;135:689–97. Baqui AH, Alvarez JO. Simultaneous zinc and vitamin A
4. Mahloudji M, Reinhold JG, Haghshenass M, Ronaghy supplementation in Bangladeshi children: Randomised
HA, Fox MR, Halsted JA. Combined zinc and iron com- double blind controlled trial. BMJ 2001;323:314–8.
pared with iron supplementation of diets of 6- to 12-year 20. Rahman MM, Tofail F, Wahed MA, Fuchs GJ, Baqui
old village schoolchildren in southern Iran. Am J Clin AH, Alvarez JO. Short-term supplementation with zinc
Nutr 1975;28:721–5. and vitamin A has no significant effect on the growth of
5. Smith RM, King RA, Spargo RM, Cheek DB, Field JB, undernourished Bangladeshi children. Am J Clin Nutr
Veitch LG. Growth-retarded aboriginal children with 2002;75:87–91.
low plasma zinc levels do not show a growth response 21. Rahman MM, Wahed MA, Fuchs GJ, Baqui AH, Alva-
to supplementary zinc. Lancet 1985;1:923–4. rez JO. Synergistic effect of zinc and vitamin A on the
6. Zemel BS, Kawchak DA, Fung EB, Ohene-Frempong K, biochemical indexes of vitamin A nutrition in children.
Stallings VA. Effect of zinc supplementation on growth Am J Clin Nutr 2002;75:92–8.
and body composition in children with sickle cell dis- 22. Osendarp SJ, Santosham M, Black RE, Wahed MA,
ease. Am J Clin Nutr 2002;75:300–7. van Raaij JM, Fuchs GJ. Effect of zinc supplementation
7. Ronaghy HA, Reinhold JG, Mahloudji M, Ghavami P, between 1 and 6 mo of life on growth and morbidity
Fox MR, Halsted JA. Zinc supplementation of malnour- of Bangladeshi infants in urban slums. Am J Clin Nutr
ished schoolboys in Iran: Increased growth and other 2002;76:1401–8.
effects. Am J Clin Nutr 1974;27:112–21. 23. Hamadani JD, Fuchs GJ, Osendarp SJ, Khatun F, Huda
8. Richard SA, Zavaleta N, Caulfield LE, Black RE, Witzig SN, Grantham-McGregor SM. Randomized controlled
RS, Shankar AH. Zinc and iron supplementation and trial of the effect of zinc supplementation on the mental
malaria, diarrhea, and respiratory infections in chil- development of Bangladeshi infants. Am J Clin Nutr
dren in the Peruvian Amazon. Am J Trop Med Hyg 2001;74:381–6.
2006;75:126–32. 24. Albert MJ, Qadri F, Wahed MA, Ahmed T, Rahman AS,
9. Castillo-Duran C, Garcia H, Venegas P, Torrealba I, Ahmed F, Bhuiyan NA, Zaman K, Baqui AH, Clemens
Panteon E, Concha N, Perez P. Zinc supplementation JD, Black RE. Supplementation with zinc, but not vita-
increases growth velocity of male children and adoles- min A, improves seroconversion to vibriocidal antibody
cents with short stature. Acta Paediatr 1994;83:833–7. in children given an oral cholera vaccine. J Infect Dis
10. Friis H, Ndhlovu P, Mduluza T, Kaondera K, Sandstrom 2003;187:909–13.
B, Michaelsen KF, Vennervald BJ, Christensen NO. The 25. Sayeg Porto MA, Oliveira HP, Cunha AJ, Miranda G,
impact of zinc supplementation on growth and body Guimaraes MM, Oliveira WA, dos Santos DM. Linear
composition: A randomized, controlled trial among growth and zinc supplementation in children with short
rural Zimbabwean schoolchildren. Eur J Clin Nutr stature. J Pediatr Endocrinol Metab 2000;13:1121–8.
1997;51:38–45. 26. Gibson RS, Vanderkooy PD, MacDonald AC, Gold-
11. Bates CJ, Evans PH, Dardenne M, Prentice A, Lunn man A, Ryan BA, Berry M. A growth-limiting, mild
PG, Northrop-Clewes CA, Hoare S, Cole TJ, Horan SJ, zinc-deficiency syndrome in some southern Ontario
Longman SC, Stirling D, Aggett PJ. A trial of zinc sup- boys with low height percentiles. Am J Clin Nutr 1989;
plementation in young rural Gambian children. Br J 49:1266–73.
Nutr 1993;69:243–55. 27. Castillo-Duran C, Rodriguez A, Venegas G, Alvarez P,
12. Lira PI, Ashworth A, Morris SS. Effect of zinc sup- Icaza G. Zinc supplementation and growth of infants born
plementation on the morbidity, immune function, and small for gestational age. J Pediatr 1995;127:206–11.
growth of low-birth-weight, full-term infants in north- 28. Ruz M, Castillo-Duran C, Lara X, Codoceo J, Rebolledo
east Brazil. Am J Clin Nutr 1998;68:418S–24S. A, Atalah E. A 14-mo zinc supplementation trial in
13. Ashworth A, Morris SS, Lira PI, Grantham-McGregor apparently healthy Chilean preschool children. Am J
SM. Zinc supplementation, mental development and Clin Nutr 1997;66:1406–13.
behaviour in low birth weight term infants in northeast 29. Castillo-Duran C, Perales CG, Hertrampf ED, Marin
Brazil. Eur J Clin Nutr 1998;52:223–7. VB, Rivera FA, Icaza G. Effect of zinc supplementation
on development and growth of Chilean infants. J Pediatr 44. Black MM, Sazawal S, Black RE, Khosla S, Kumar J,
2001;138:229–35. Menon V. Cognitive and motor development among
30. Hong ZY, Zhang YW, Xu JD, Zhou JD, Gao XL, Liu XG, small-for-gestational-age infants: Impact of zinc sup-
Shi YY. Growth promoting effect of zinc supplementa- plementation, birth weight, and caregiving practices.
tion in infants of high-risk pregnancies. Chin Med J Pediatrics 2004;113:1297–305.
(Engl) 1992;105:844–8. 45. Bhandari N, Bahl R, Taneja S, Strand T, Molbak K,
31. Sempertegui F, Estrella B, Correa E, Aguirre L, Saa B, Ulvik RJ, Sommerfelt H, Bhan MK. Effect of routine
Torres M, Navarrete F, Alarcon C, Carrion J, Rodriguez, zinc supplementation on pneumonia in children aged
Griffiths JK. Effects of short-term zinc supplementation 6 months to 3 years: Randomised controlled trial in an
on cellular immunity, respiratory symptoms, and growth urban slum. BMJ 2002;324:1358.
of malnourished Equadorian children. Eur J Clin Nutr 46. Taneja S, Bhandari N, Bahl R, Bhan MK. Impact of zinc
1996;50:42–6. supplementation on mental and psychomotor scores of
32. Wuehler SE, Sempertegui F, Brown KH. Dose-response children aged 12 to 18 months: A randomized, double-
trial of prophylactic zinc supplements, with or without blind trial. J Pediatr 2005;146:506–11.
copper, in young Ecuadorian children at risk of zinc 47. Bhandari N, Bahl R, Taneja S, Strand T, Molbak K,
deficiency. Am J Clin Nutr 2008;87:723–33. Ulvik RJ, Sommerfelt H, Bhan MK. Substantial reduc-
33. Walravens PA, Chakar A, Mokni R, Denise J, Lemon- tion in severe diarrheal morbidity by daily zinc supple-
nier D. Zinc supplements in breastfed infants. Lancet mentation in young north Indian children. Pediatrics
1992;340:683–5. 2002;109:86–92.
34. Ruel MT, Rivera JA, Santizo MC, Lonnerdal B, Brown 48. Bhandari N, Taneja S, Mazumder S, Bahl R, Fontaine O,
KH. Impact of zinc supplementation on morbidity from Bhan MK. Adding zinc to supplemental iron and folic
diarrhea and respiratory infections among rural Guate- acid does not affect mortality and severe morbidity in
malan children. Pediatrics 1997;99:808–13. young children. J Nutr 2007;137:112–7.
35. Bentley ME, Caulfield LE, Ram M, Santizo MC, 49. Lind T, Lönnerdal B, Stenlund H, Gamayanti IL, Ismail
Hurtado E, Rivera JA, Ruel MT, Brown KH. Zinc sup- D, Seswandhana R, Persson LA. A community-based
plementation affects the activity patterns of rural Gua- randomized controlled trial of iron and zinc supple-
temalan infants. J Nutr 1997;127:1333–8. mentation in Indonesian infants: Effects on growth and
36. Rivera JA, Ruel MT, Santizo MC, Lönnerdal B, development. Am J Clin Nutr 2004;80:729–36.
Brown KH. Zinc supplementation improves the 50. Lind T, Lönnerdal B, Stenlund H, Ismail D, Seswand-
growth of stunted rural Guatemalan infants. J Nutr hana R, Ekstrom EC, Persson LA. A community-based
1998;128:556–62. randomized controlled trial of iron and zinc supplemen-
37. Sazawal S, Black RE, Bhan MK, Jalla S, Bhandari N, tation in Indonesian infants: Interactions between iron
Sinha A, Majumdar S. Zinc supplementation reduces and zinc. Am J Clin Nutr 2003;77:883–90.
the incidence of persistent diarrhea and dysentery 51. Fahmida U, Rumawas JS, Utomo B, Patmonodewo S,
among low socioeconomic children in India. J Nutr Schultink W. Zinc-iron, but not zinc-alone supplemen-
1996;126:443–50. tation, increased linear growth of stunted infants with
38. Sazawal S, Bentley M, Black RE, Dhingra P, George S, low haemoglobin. Asia Pac J Clin Nutr 2007;16:301–9.
Bhan MK. Effect of zinc supplementation on observed 52. Meeks Gardner J, Witter MM, Ramdath DD. Zinc sup-
activity in low socioeconomic Indian preschool chil- plementation: Effects on the growth and morbidity of
dren. Pediatrics 1996;98:1132–7. undernourished Jamaican children. Eur J Clin Nutr
39. Sazawal S, Black RE, Bhan MK, Jalla S, Sinha A, 1998;52:34–9.
Bhandari N. Efficacy of zinc supplementation in reduc- 53. Meeks Gardner J, Powell CA, Baker-Henningham H,
ing the incidence and prevalence of acute diarrhea—a Walker SP, Cole TJ, Grantham-McGregor SM. Zinc sup-
community-based, double-blind, controlled trial. Am J plementation and psychosocial stimulation: Effects on
Clin Nutr 1997;66:413–8. the development of undernourished Jamaican children.
40. Sazawal S, Jalla S, Mazumder S, Sinha A, Black RE, Am J Clin Nutr 2005;82:399–405.
Bhan MK. Effect of zinc supplementation on cell-medi- 54. Long KZ, Montoya Y, Hertzmark E, Santos JI, Rosado JL.
ated immunity and lymphocyte subsets in preschool A double-blind, randomized, clinical trial of the effect of
children. Indian Pediatr 1997;34:589–97. vitamin A and zinc supplementation on diarrheal disease
41. Sazawal S, Black RE, Jalla S, Mazumdar S, Sinha A, and respiratory tract infections in children in Mexico
Bhan MK. Zinc supplementation reduces the incidence City, Mexico. Am J Clin Nutr 2006;83:693–700.
of acute lower respiratory infections in infants and 55. Tielsch JM, Khatry SK, Stoltzfus RJ, Katz J, LeClerq SC,
preschool children: A double-blind, controlled trial. Adhikari R, Mullany LC, Shresta S, Black RE. Effect of
Pediatrics 1998;102:1–5. routine prophylactic supplementation with iron and
42. Sazawal S, Malik P, Jalla S, Krebs N, Bhan MK, Black RE. folic acid on preschool child mortality in southern
Zinc supplementation for four months does not affect Nepal: Community-based, cluster-randomised, placebo-
plasma copper concentration in infants. Acta Paediatr controlled trial. Lancet 2006;367:144–52.
2004;93:599–602. 56. Tielsch JM, Khatry SK, Stoltzfus RJ, Katz J, LeClerq SC,
43. Sazawal S, Black RE, Menon VP, Dinghra P, Caulfield LE, Adhikari R, Mullany LC, Black R, Shresta S. Effect of
Dhingra U, Bagati A. Zinc supplementation in infants daily zinc supplementation on child mortality in south-
born small for gestational age reduces mortality: A ern Nepal: A community-based, cluster randomised,
prospective, randomized, controlled trial. Pediatrics placebo-controlled trial. Lancet 2007;370:1230–9.
2001;108:1280–6. 57. Penny ME, Marin RM, Duran A, Peerson JM, Lanata
CF, Lönnerdal B, Black RE, Brown KH. Randomized of zinc supplementation on malaria and other causes of
controlled trial of the effect of daily supplementation morbidity in West African children: Randomised double
with zinc or multiple micronutrients on the morbidity, blind placebo controlled trial. BMJ 2001;322:1567–72.
growth, and micronutrient status of young Peruvian 70. Müller O, Garenne M, Reitmaier P, Van Zweeden AB,
children. Am J Clin Nutr 2004;79:457–65. Kouyate B, Becher H. Effect of zinc supplementation on
58. Brown KH, López de Romaña D, Arsenault JE, Peer- growth in West African children: A randomized double-
son JM, Penny ME. Comparison of the effects of zinc blind placebo-controlled trial in rural Burkina Faso. Int
delivered in a fortified food or a liquid supplement J Epidemiol 2003;32:1098–102.
on the growth, morbidity, and plasma zinc concen- 71. Sandstead HH, Penland JG, Alcock NW, Dayal HH,
trations of young Peruvian children. Am J Clin Nutr Chen XC, Li JS, Zhao F, Yang JJ. Effects of repletion
2007;85:538–47. with zinc and other micronutrients on neuropsychologic
59. Bobat R, Coovadia H, Stephen C, Naidoo KL, McKerrow performance and growth of Chinese children. Am J Clin
N, Black RE, Moss WJ. Safety and efficacy of zinc sup- Nutr 1998;68:470S–5S.
plementation for children with HIV-1 infection in South 72. Penland JG, Sandstead HH, Alcock NW, Dayal HH,
Africa: A randomised double-blind placebo-controlled Chen XC, Li JS, Zhao F, Yang JJ. A preliminary report:
trial. Lancet 2005;366:1862–7. Effects of zinc and micronutrient repletion on growth
60. Sazawal S, Black RE, Ramsan M, Chwaya HM, Stoltzfus and neuropsychological function of urban Chinese
RJ, Dutta A, Dhingra U, Kabole I, Deb S, Othman MK, children. J Am Coll Nutr 1997;16:268–72.
Kabole FM. Effects of routine prophylactic supplemen- 73. Yang YX, Han JH, Shao XP, He M, Bian LH, Wang Z,
tation with iron and folic acid on admission to hospital Wang GD, Men JH. Effect of micronutrient supplemen-
and mortality in preschool children in a high malaria tation on the growth of preschool children in China.
transmission setting: Community-based, randomised, Biomed Environ Sci 2002;15:196–202.
placebo-controlled trial. Lancet 2006;367:133–43. 74. Dirren H, Barclay D, Ramos JG, Lozano R, Montalvo MM,
61. Sazawal S, Black RE, Ramsan M, Chwaya HM, Dutta A, Davila N, Mora JO. Zinc supplementation and child growth
Dhingra U, Stoltzfus RJ, Othman MK, Kabole FM. Effect in Ecuador. Adv Exp Med Biol 1994;352:215–22.
of zinc supplementation on mortality in children aged 75. Umeta M, West CE, Haidar J, Deurenberg P, Hautvast JG.
1–48 months: A community-based randomised placebo- Zinc supplementation and stunted infants in Ethiopia: A
controlled trial. Lancet 2007;369:927–34. randomised controlled trial. Lancet 2000;355:2021–6.
62. Olney DK, Pollitt E, Kariger PK, Khalfan SS, Ali NS, 76. Cavan KR, Gibson RS, Grazioso CF, Isalgue AM, Ruz M,
Tielsch JM, Sazawal S, Black R, Allen LH, Stoltzfus Solomons NW. Growth and body composition of per-
RJ. Combined iron and folic acid supplementation iurban Guatemalan children in relation to zinc status:
with or without zinc reduces time to walking unas- A longitudinal zinc intervention trial. Am J Clin Nutr
sisted among Zanzibari infants 5- to 11-mo old. J Nutr 1993;57:344–52.
2006;136:2427–34. 77. Grazioso CF, Isalgue M, de Ramirez I, Ruz M, Solomons
63. Wasantwisut E, Winichagoon P, Chitchumroonchokchai NW. The effect of zinc supplementation on parasitic
C, Yamborisut U, Boonpraderm A, Pongcharoen T, reinfestation of Guatemalan schoolchildren. Am J Clin
Sranacharoenpong K, Russameesopaphorn W. Iron and Nutr 1993;57:673–8.
zinc supplementation improved iron and zinc status, but 78. Sur D, Gupta DN, Mondal SK, Ghosh S, Manna B,
not physical growth, of apparently healthy, breast-fed Rajendran K, Bhattacharya SK. Impact of zinc supple-
infants in rural communities of northeast Thailand. J mentation on diarrheal morbidity and growth pattern
Nutr 2006;136:2405–11. of low birth weight infants in Kolkata, India: A rand-
64. Walravens PA, Krebs NF, Hambidge KM. Linear growth omized, double-blind, placebo-controlled, community-
of low income preschool children receiving a zinc sup- based study. Pediatrics 2003;112:1327–32.
plement. Am J Clin Nutr 1983;38:195–201. 79. Dijkhuizen MA, Wieringa FT, West CE, Martuti S,
65. Walravens PA, Hambidge KM, Koepfer DM. Zinc Muhilal. Effects of iron and zinc supplementation in
supplementation in infants with a nutritional pattern Indonesian infants on micronutrient status and growth.
of failure to thrive: A double-blind, controlled study. J Nutr 2001;131:2860–5.
Pediatrics 1989;83:532–8. 80. Wieringa FT, Dijkhuizen MA, West CE, Northrop-
66. Heinig MJ, Brown KH, Lönnerdal B, Dewey KG. Zinc Clewes CA, Muhilal. Estimation of the effect of the acute
supplementation does not affect growth, morbidity, or phase response on indicators of micronutrient status in
motor development of US term breastfed infants at 4–10 Indonesian infants. J Nutr 2002;132:3061–6.
mo of age. Am J Clin Nutr 2006;84:594–601. 81. Wieringa FT, Dijkhuizen MA, West CE, Thurnham DI,
67. Ninh NX, Thissen JP, Collette L, Gerard G, Khoi HH, Muhilal, Van der Meer JW. Redistribution of vitamin A
Ketelslegers JM. Zinc supplementation increases growth after iron supplementation in Indonesian infants. Am J
and circulating insulin-like growth factor I (IGF-I) in Clin Nutr 2003;77:651–7.
growth-retarded Vietnamese children. Am J Clin Nutr 82. Rosado JL, Lopez P, Munoz E, Martinez H, Allen LH.
1996;63:514–9. Zinc supplementation reduced morbidity, but neither
68. Berger J, Ninh NX, Khan NC, Nhien NV, Lien DK, zinc nor iron supplementation affected growth or body
Trung NQ, Khoi HH. Efficacy of combined iron and zinc composition of Mexican preschoolers. Am J Clin Nutr
supplementation on micronutrient status and growth in 1997;65:13–9.
Vietnamese infants. Eur J Clin Nutr 2006;60:443–54. 83. Allen LH, Rosado JL, Casterline JE, Lopez P, Munoz E,
69. Müller O, Becher H, van Zweeden AB, Ye Y, Diallo DA, Garcia OP, Martinez H. Lack of hemoglobin response to
Konate AT, Gbangou A, Kouyate B, Garenne M. Effect iron supplementation in anemic mexican preschoolers
with multiple micronutrient deficiencies. Am J Clin Arifeen S, Le K, McNary SW, Parveen M, Hamadani
Nutr 2000;71:1485–94. JD, Black RE. Iron and zinc supplementation promote
84. Munoz EC, Rosado JL, Lopez P, Furr HC, Allen LH. motor development and exploratory behavior among
Iron and zinc supplementation improves indicators of Bangladeshi infants. Am J Clin Nutr 2004;80:903–10.
vitamin A status of Mexican preschoolers. Am J Clin 97. Baqui AH, Walker CL, Zaman K, El Arifeen S, Chowd-
Nutr 2000;71:789–94. hury HR, Wahed MA, Black RE, Caulfield LE. Weekly
85. Kordas K, Stoltzfus RJ, Lopez P, Rico JA, Rosado JL. Iron iron supplementation does not block increases in serum
and zinc supplementation does not improve parent or zinc due to weekly zinc supplementation in Bangladeshi
teacher ratings of behavior in first grade Mexican chil- infants. J Nutr 2005;135:2187–91.
dren exposed to lead. J Pediatr 2005;147:632–9. 98. Brooks WA, Santosham M, Naheed A, Goswami D,
86. Rosado JL, Lopez P, Kordas K, Garcia-Vargas G, Ron- Wahed MA, Diener-West M, Faruque AS, Black RE.
quillo D, Alatorre J, Stoltzfus RJ. Iron and/or zinc sup- Effect of weekly zinc supplements on incidence of pneu-
plementation did not reduce blood lead concentrations monia and diarrhoea in children younger than 2 years in
in children in a randomized, placebo-controlled trial. J an urban, low-income population in Bangladesh: Ran-
Nutr 2006;136:2378–83. domised controlled trial. Lancet 2005;366:999–1004.
87. Rico JA, Kordas K, Lopez P, Rosado JL, Vargas GG, 99. Gupta DN, Rajendran K, Mondal SK, Ghosh S, Bhat-
Ronquillo D, Stoltzfus RJ. Efficacy of iron and/or tacharya SK. Operational feasibility of implementing
zinc supplementation on cognitive performance of community-based zinc supplementation: Impact on
lead-exposed Mexican schoolchildren: A randomized, childhood diarrheal morbidity. Pediatr Infect Dis J
placebo-controlled trial. Pediatrics 2006;117:518–27. 2007;26:306–10.
88. Shankar AH, Genton B, Baisor M, Paino J, Tamja S, 100. Zinc Investigators’ Collaborative Group, Bhutta ZA,
Adiguma T, Wu L, Rare L, Bannon D, Tielsch JM, West Bird SM, Black RE, Brown KH, Gardner JM, Hidayat A,
KP Jr, Alpers MP. The influence of zinc supplementa- Khatun F, Martorell R, Ninh NX, Penny ME, Rosado
tion on morbidity due to Plasmodium falciparum: A JL, Roy SK, Ruel M, Sazawal S, Shankar A. Therapeutic
randomized trial in preschool children in Papua New effects of oral zinc in acute and persistent diarrhea in
Guinea. Am J Trop Med Hyg 2000;62:663–9. children in developing countries: Pooled analysis of
89. Alarcon K, Kolsteren PW, Prada AM, Chian AM, randomized controlled trials. Am J Clin Nutr 2000;
Velarde RE, Pecho IL, Hoeree TF. Effects of separate 72:1516–22.
delivery of zinc or zinc and vitamin A on hemoglobin 101. World Health Organization. Management of a child
response, growth, and diarrhea in young Peruvian chil- with serious infection or severe malnutrition: Guidelines
dren receiving iron therapy for anemia. Am J Clin Nutr for care at the first-referral level in developing countries.
2004;80:1276–82. Geneva: WHO, 2000.
90. Udomkesmalee E, Dhanamitta S, Sirisinha S, Charoen- 102. International Zinc Nutrition Consultative Group
kiatkul S, Tuntipopipat S, Banjong O, Rojroongwasinkul (IZiNCG), Brown KH, Rivera JA, Bhutta Z, Gibson
N, Kramer TR, Smith JC Jr. Effect of vitamin A and zinc RS, King JC, Lönnerdal B, Ruel MT, Sandtröm B,
supplementation on the nutriture of children in North- Wasantwisut E, Hotz C. Assessment of the risk of zinc
east Thailand. Am J Clin Nutr 1992;56:50–7. deficiency in populations and options for its control.
91. Kramer TR, Udomkesmalee E, Dhanamitta S, Siris- Food Nutr Bull 2004;25(1 suppl 2):S99–203.
inha S, Charoenkiatkul S, Tuntipopipat S, Banjong O, 103. Zeba AN, Sorgho H, Rouamba N, Zongo I, Rouamba
Rojroongwasinkul N, Smith JC Jr. Lymphocyte respon- J, Guiguemde RT, Hamer DH, Mokhtar N, Ouedraogo
siveness of children supplemented with vitamin A and JB. Major reduction of malaria morbidity with combined
zinc. Am J Clin Nutr 1993;58:566–70. vitamin A and zinc supplementation in young children
92. Kikafunda JK, Walker AF, Allan EF, Tumwine JK. in Burkina Faso: A randomized double blind trial. Nutr
Effect of zinc supplementation on growth and body J 2008;7:7–13.
composition of Ugandan preschool children: A rand- 104. Hess SY, Peerson JM, King JC, Brown KH. Use of
omized, controlled, intervention trial. Am J Clin Nutr serum zinc concentration as an indicator of population
1998;68:1261–6. zinc status. Food Nutr Bull 2007;28:S403–29.
93. Friis H, Ndhlovu P, Mduluza T, Kaondera K, Sandstrom 105. US Institute of Medicine. Dietary reference intakes
B, Michaelsen KF, Vennervald BJ, Christensen NO. The for vitamin A, vitamin K, arsenic, boron, chromium,
impact of zinc supplementation on Schistosoma mansoni iodine, iron, manganese, molybdenum, nickel, silicon,
reinfection rate and intensities: A randomized, control- vanadium, and zinc, Washington, DC: National Acad-
led trial among rural Zimbabwean schoolchildren. Eur emy Press, 2001.
J Clin Nutr 1997;51:33–7. 106. Sandström B, Davidsson L, Cederblad A, Lönner-
94. Gupta DN, Mondal SK, Ghosh S, Rajendran K, Sur D, dal B. Oral iron, dietary ligands and zinc absorption. J
Manna B. Impact of zinc supplementation on diarrhoeal Nutr 1985;115:411–4.
morbidity in rural children of West Bengal, India. Acta 107. Rossander-Hulten L, Brune M, Sandström B, Lön-
Paediatr 2003;92:531–6. nerdal B, Hallberg L. Competitive inhibition of iron
95. Baqui AH, Zaman K, Persson LA, El Arifeen S, Yunus M, absorption by manganese and zinc in humans. Am J
Begum N, Black RE. Simultaneous weekly supplementa- Clin Nutr 1991;54:152–6.
tion of iron and zinc is associated with lower morbidity 108. Fischer Walker C, Kordas K, Stoltzfus RJ, Black RE.
due to diarrhea and acute lower respiratory infection in Interactive effects of iron and zinc on biochemical and
Bangladeshi infants. J Nutr 2003;133:4150–7. functional outcomes in supplementation trials. Am J
96. Black MM, Baqui AH, Zaman K, Ake Persson L, El Clin Nutr 2005;82:5–12.
109. Yadrick MK, Kenney MA, Winterfeldt EA. Iron, Micronutrient Forum, Istanbul, Turkey, 2007. Abstract
copper, and zinc status: Response to supplementation T79. Available at: www.micronutrientforum.org/
with zinc or zinc and iron in adult females. Am J Clin Meeting2007
Nutr 1989;49:145–50. 119. Ashworth A, Shrimpton R, Jamil K. Growth moni-
110. Davis CD, Milne DB, Nielsen FH. Changes in toring and promotion: Review of evidence of impact.
dietary zinc and copper affect zinc-status indicators of Matern Child Nutr 2008;4(suppl 1):86–117.
postmenopausal women, notably, extracellular super- 120. Haddad D, Cross C, Thylefors B, Richards FO Jr.,
oxide dismutase and amyloid precursor proteins. Am J Bush S, Hopkins AD, Baker SK. Health care at the end
Clin Nutr 2000;71:781–8. of the road: Opportunities from 20 years of partnership
111. Milne DB. Copper intake and assessment of copper in onchocerciasis control. Global Public Health 2008;
status. Am J Clin Nutr 1998;67:1041S–5S. 3:187–96.
112. Allen LH, Gillespie SR. What works? A review of 121. Haselow N, Obadiah M, Akame J. The integration
the efficacy and effectiveness of nutrition interventions. of vitamin A supplementation into community-directed
Geneva: United Nations Administrative Committee treatment with ivermectin: A practical guide for Africa.
on Coordination/Sub-Committee on Nutrition (ACC/ New York: Helen Keller International, 2004.
SCN) and Manila: Asian Development Bank, 2001. 122. Mora JO. Iron supplementation: Overcoming tech-
113. Huffman SL, Baker J, Shumann J, Zehner ER. The nical and practical barriers. J Nutr 2002;132:853S–5S.
case for promoting multiple vitamin/mineral supple- 123. Favin M, Griffiths M, The Manoff Group. Social
ments for women of reproductive age in developing marketing of micronutrients in developing countries.
countries. Washington, DC: LINKAGES Project, Acad- Washington DC: World Bank, 1992.
emy for Educational Development, 1998. 124. Nestel P, Briend A, de Benoist B, Decker E, Fer-
114. World Health Organization/UNICEF/US Agency guson E, Fontaine O, Micardi A, Nalubola R. Comple-
for International Development/US Pharmacopeia/ mentary food supplements to achieve micronutrient
Drug Quality and Information Program/Johns Hopkins adequacy for infants and young children. J Pediatr
Bloomberg School of Public Health. Production of zinc Gastroenterol Nutr 2003;36:316–28.
tablets and oral rehydration solution. Guidelines for 125. Hess SY, Brown KH. Impact of zinc fortification on
programme managers and pharmaceutical manufactur- zinc nutrition. Food Nutr Bull 2009;30:S79–107.
ers. Geneva: WHO, 2007. 126. World Health Organization Secretariat on behalf
115. Dalmiya N, Palmer A, Darnton-Hill I. Sustain- of the participants to the Consultation. Conclusions
ing vitamin A supplementation requires a new vision. and recommendations of the WHO Consultation on
Lancet 2006;368:1052–4. prevention and control of iron deficiency in infants and
116. UNICEF. The state of the world’s children 2008: young children in malaria-endemic areas. Food Nutr
Child survival. New York: UNICEF, 2007. Bull 2007;28:S621–7.
117. Aguayo VM, Garnier D, Baker SK. Drops of life: 127. Lawn J, Kerber K, ed. Opportunities for Africa‘s
Vitamin A supplementation for child survival. Progress newborns: Practical data, policy and programmatic
and lessons learned in West and Central Africa. Dakar: support for newborn care in Africa. Cape Town, South
UNICEF Regional Office for West and Central Africa/ Africa: Partnership for Maternal, Newborn and Child
Helen Keller International Regional Office for Africa, 2007. Health, 2006.
118. Kavle J, Baker S, Bendech MA, Aguayo VM. Analy- 128. Stoltzfus RJ, Dreyfus ML. Guidelines for the use of
sis and actions in West and Central Africa: Using surveys iron supplements to prevent and treat iron deficiency
to improve vitamin A supplementation performance. anemia. Washington, DC: ILSI Press, 1998.
The effect of therapeutic zinc supplementation among
young children with selected infections: A review of
the evidence
Batool A. Haider and Zulfiqar A. Bhutta
Abstract Key words: Diarrhea treatment, infections, malaria,

therapeutic, tuberculosis, zinc supplement
Background. Zinc deficiency is now widely recognized
as a leading risk factor for morbidity and mortality and
is estimated to be responsible for approximately 800,000 Background
excess deaths annually among children under 5 years
of age. Zinc is vital for several body functions, including
Objective. To evaluate the impact of zinc supple- protein synthesis and cell growth and differentiation.
mentation as an adjunct in the treatment of diarrhea, Severe zinc deficiency is characterized by stunted
pneumonia, malaria, and tuberculosis in children under growth, hypogonadism, impaired immune function,
5 years of age. skin disorders, cognitive dysfunction, and anorexia
Methods. A comprehensive literature search of elec- [1]. Zinc deficiency is now widely recognized as a
tronic databases to identify randomized, controlled trials leading risk factor for morbidity and mortality [2, 3].
on the topic was undertaken in January 2008. Eligible It is estimated that zinc deficiency is responsible for
studies identified on search were reviewed by the authors approximately 800,000 excess deaths annually among
and data extraction was done. Statistical analyses were children under 5 years of age. These deaths are related
performed with the use of Review Manager software. to diarrhea (176,000), pneumonia (406,000), and
Results. Current analysis of the adjunctive therapeutic malaria (207,000) brought on by inadequate zinc
benefit of zinc in acute diarrhea corroborates existing intake [4]. Further, a loss of nearly 16 million global
reviews and provides evidence of reduction in the dura- disability-adjusted life years (DALYs) is attributed to
tion of acute diarrhea by 0.5 day (p = .002) in children zinc deficiency [2]. The aim of this review is to evaluate
under 5 years of age. However, zinc supplementation is the impact of zinc supplementation as an adjunct in the
found to have no beneficial impact in infants under 6 treatment of infectious diseases (diarrhea, pneumonia,
months of age. A beneficial effect of zinc as an adjunc- malaria, and tuberculosis) in children under 5 years
tive treatment is also found in persistent diarrhea, the of age.
duration of which is reduced by 0.68 day (p < .0001). This paper is divided into five sections, which
Evidence of the benefit of zinc supplementation in pneu- address the following questions in relation to thera-
monia and malaria is insufficient, whereas no studies are peutic zinc supplementation:
available in children with tuberculosis.
Section 1: What is the effect of therapeutic zinc sup-
Conclusions. The existing literature provides evidence
plementation as an adjunct in the treatment of children
of a beneficial effect of therapeutic zinc supplementation
with acute diarrhea on the duration of the disease?
in the reduction of the duration of acute and persistent
diarrhea. However, evidence for its impact on pneumo- Section 2: What is the effect of therapeutic zinc sup-
nia, malaria. and tuberculosis in children under 5 years plementation as an adjunct in the treatment of children
of age is insufficient and needs further evaluation. with persistent diarrhea on the duration and severity
of the disease?
The authors are affiliated with the Department of Pedi-
atrics and Child Health, the Aga Khan University, Karachi, Section 3: What is the effect of therapeutic zinc sup-
Pakistan. plementation as an adjunct in the treatment of children
Please direct queries to the corresponding author: Zulfiqar with pneumonia on the duration and severity of the
A. Bhutta, Department of Paediatrics and Child Health, the disease?
Aga Khan University, Karachi, Pakistan; email: zulfiqar.
bhutta@aku.edu. Section 4: What is the effect of therapeutic zinc
S42 B. A. Haider and Z. A. Bhutta
supplementation as an adjunct in the treatment of supplementation in the treatment of acute diarrhea in

children with malaria? children reduces the duration of diarrhea by 0.5 day
Section 5: What is the effect of therapeutic zinc sup- (WMD, −0.50; 95% confidence interval [CI], −0.82 to
plementation as an adjunct in the treatment of children −0.18). These results support existing recommenda-
with tuberculosis? tions that zinc supplementation should be included
as a component of treatment of children with acute
diarrhea.
Methods Detailed review of the evidence
To identify randomized, controlled trials evaluating Several clinical trials have studied the effect of admin-
the effect of zinc supplementation as an adjunctive istration of zinc to children with diarrheal illness to
treatment for diarrhea, pneumonia, malaria, and tuber- determine whether this intervention can reduce the
culosis in children under 5 years of age, we undertook severity and duration of diarrhea. In a systematic
a comprehensive search of the electronic published review and pooled analysis of randomized, control-
literature up to January 2008, including PUBMED, led trials of children under 5 years of age with acute
POPLINE, EMRO, and African Index Medicus. The diarrhea in the year 2000, the Zinc Investigators’ Col-
Cochrane Databases of Systematic Reviews and Clini- laborative Group [5] found four studies that evaluated
cal Trials were also searched. The search was conducted the effect of adjunctive therapy with zinc supplements
by the authors, who reviewed the titles and abstracts, containing at least 50% of the US recommended dietary
and, if required, full texts were retrieved to assess the allowance (RDA) per day. Pooled analysis of data
eligibility of the studies. All identified eligible stud- from three of these studies showed that children who
ies were then reviewed by the authors, and data were received zinc supplementation had a 15% lower prob-
extracted. ability of continuing diarrhea on a particular day than
Statistical analyses and meta-analyses were per- those not receiving zinc.
formed by using Review Manager software, version To update this previous analysis, a literature search
4.2.8 (Cochrane Collaboration, 2003). Weighted mean was undertaken in early January 2008. The studies
differences (WMD) for continuous outcomes and eligible for inclusion in the review were randomized,
relative risks (RR) for dichotomous outcomes were controlled trials that evaluated the adjunctive thera-
calculated to present impact estimates. Where neces- peutic effect of zinc supplementation in children
sary, data were pooled using a generic inverse variance under 5 years of age with acute diarrhea. The search
method. Logarithms of relative risk or hazards and identified 21 trials completed since 1988, of which 2
their standard errors were calculated by standard statis- were multicenter studies. The main characteristics of
tical methods. All impacts were presented as summary the included studies [5–26] are summarized in table 1.
estimates along with 95% confidence intervals. For Most of the studies were conducted in lower-income
each parameter, heterogeneity among the studies was countries where zinc deficiency is common, and most
tested by chi-square statistics, I2 statistics, and visual included children who presented with acute watery
inspection of forest plots. Heterogeneity was found to diarrhea of unspecified etiology. The data from six
be substantial if the chi-square statistic was significant studies could not be included in the analysis, because
(p < .01), I2 was more than 50%, and the visual inspec- they did not include our outcomes of interest [23], they
tion of forest plots was suggestive. In case of absence presented data in a form that precluded inclusion [15],
of significant heterogeneity, a fixed-effects model was they included cases of shigellosis [18] or cholera [26],
used; however, in the presence of significant heteroge- or the data were not disaggregated by age (< 5 years vs.
neity, a random-effects model was used for pooling of ≥ 5 years) or intervention group to allow for a specific
data. This was followed by subgroup analysis to explore focus on the independent effect of zinc in our age range
the possible causes of heterogeneity. of interest [12, 21]. Zinc was supplemented in doses
varying from 5 to 45 mg/day.
The results of the meta-analysis of the 14 included
Section 1 studies showed that the mean duration of acute diarrhea
was significantly less in the zinc-supplemented group
What is the effect of therapeutic zinc supplementation than in the placebo group (WMD, −0.50; 95% CI, −0.82
as an adjunct in the treatment of children with acute to −0.18, random-effects model) (fig. 1). Because of
diarrhea on the duration of the disease? the presence of significant heterogeneity among the
included trials, subgroup analysis by age group was
Conclusions undertaken using 6 months of age or greater as the
cutoff. Zinc supplementation was found to increase the
Existing evidence indicates that the inclusion of zinc mean duration of acute diarrhea in children under 6
TABLE 1. Randomized, controlled trials of the use of zinc in the treatment of acute diarrhea
Country, year
[reference] Age Study
author group participants Zinc Placebo Interventions Results
India, 1988 6–18 mo Children with n = 25 n = 25 Intervention: zinc sulfate 40 mg/ Zinc supplementation led to a nonsignificant 9% reduction
[8] Sachdev acute dehydrating day in the duration of diarrhea and an 18% reduction in stool
diarrhea < 4 days Control: placebo (glucose) frequency. Children with low rectal mucosal zinc concen-
trations at baseline had more marked reductions in dura-
tion of diarrhea (33%) and in stool frequency (33%) than
those with higher tissue zinc
India, 1995 6–35 mo Children with n = 456 n = 481 Intervention: zinc gluconate 20 Zinc-supplemented group had 23% (95% CI, 12% to 32%)
[9] Sazawal diarrhea < 7 days mg/day, vitamins A, B, D, and E, reduction in the risk of continued diarrhea on a given
and ORS day. There was a 39% (95% CI, 6% to 70%) reduction
Control: vitamins A, B, D, and E in the mean number of watery stools per day and a 21%
and ORS reduction (95% CI, 10% to 31%) in the number of days
with watery stools. Reductions in duration and severity
were greater in children with stunting than in those with-
out stunting
Bangladesh, 3–24 mo Children with acute n = 57 n = 54 Intervention: zinc acetate 20 mg/ The median total diarrheal stool output was 28% less in
1997 diarrhea < 3 days day and vitamins A, B, D, and C the zinc-supplemented group (p = .06). There was a 14%
[10] Roy and weight-for-age Control: vitamins A, B, D, and C reduction in the duration of diarrhea in the zinc-supple-
< 76th percentile of mented group. The subgroup with lower plasma zinc (<
NCHS medians 14 µmol/L) at baseline had a significant (22%) reduction
in duration of diarrhea
Indonesia, 1998 3–35 mo Children with acute n = 739 n = 659 Intervention: zinc acetate 4–5 mg/ Mean duration of diarrhea reduced by 10% (from 3.8 ± 2.6
[11] Hidayat diarrhea day days to 3.5 ± 2.4 days) (not significant)
Control: placebo
Bangladesh, 6–24 mo Children with acute (a) Vitamin n = 171 (a) Vitamin A, 4,500 µg RE daily; Zinc supplementation was associated with a reduced dura-
1999 diarrhea ≥ 3 days A (n = 172) (b) 14.2 mg elemental zinc as tion of diarrhea (13%, p = .03) and markedly reduced rate
[12] Faruque (b) Zinc acetate for zinc or zinc and vita- of prolonged diarrhea (> 7 days) (43%, p = .017)
(n = 170) min A group for the first 417
(c) Zinc + patients enrolled in the study
vitamin A and 40 mg for the remaining
(n = 171) 273 patients randomized to
these groups; (c) both vitamin
A and zinc at the above doses
daily; (d) placebo
continued
S43
TABLE 1. Randomized, controlled trials of the use of zinc in the treatment of acute diarrhea (continued) S44
Country, year
Meta-analysis of 3 trials, 2000 (5) Zinc Investigators group Zinc supplementation was associated with a 15% (95% CI,
(Sazawal 1995 [9], Roy 1997 [10], and Hidayat 1998 [11]) 5% to 24%) lower probability of continuing diarrhea on
a given day
India, 2000 3–24 mo Male children with n = 44 n = 36 Intervention: zinc sulfate 40 mg/ Significantly shorter duration of diarrhea (70.4 ± 10.0 vs.
[6] Dutta acute watery day and ORS 103.4 ± 17.1 h, p = .0001), less liquid stools (1.5 ± 0.7 vs.
diarrhea ≤ 3 Control: ORS 2.4 ± 0.7 kg, p = .0001), less need for ORS (2.5 ± 1.0 vs.
days with some 3.6 ± 0.8 L, p = .0001) and other liquids (867 ± 466 vs.
dehydration and 1,355 ± 676 mL, p = .0001)
< 80% Harvard
standard weight-
for-age
India, 2002 6–35 mo Children with (a) Zinc n = 401 Intervention: (a) 15 mg (6–11 mo) Children receiving zinc syrup had lower duration of
[13] Bahl acute diarrhea ≤ (n = 404) or 30 mg (12–35 mo) elemental diarrhea (relative hazard, 0.89; 95% CI, 0.80–0.99) and
3 days (b) Zinc zinc per day as a syrup; zinc glu- fewer total stools (rate ratio, 0.73; 95% CI, 0.70 to 0.77)
+ ORS conate, (b) ORS premixed with than controls. Children receiving zinc-ORS had fewer
(n = 402) zinc (40 mg/L), zinc gluconate total stools (rate ratio, 0.83; 95% CI, 0.71 to 0.96), and
Control: ORS only. had watery stools less often (odds ratio, 0.61; 95% CI,
0.39 to 0.95) than the control children. There was no sig-
nificant effect on the duration of diarrhea
Nepal, 2002 6–35 mo Acute watery Zinc n = 449 Intervention: 15 mg/day (infants) The relative hazards for termination of diarrhea were
[14] Strand diarrhea (n = 442) or 30 mg/day (older infants) 26% (95% CI, 8% to 46%), 21% (95% CI, 4% to 38%),
Zinc + elemental zinc and massive vita- and 19% (95% CI, 2% to 40%) higher in the zinc, zinc +
vitamin A min A dose at enrollment in the vitamin A, and zinc-caretaker groups, respectively, than
(n = 447) vitamin A group in the placebo group. The relative risks of prolonged
Zinc- Control: placebo diarrhea (duration > 7 days) in these groups were 0.57
caretaker (95% CI, 0.38% to 0.86), 0.53 (95% CI, 0.35% to 0.81),
(n = 448) and 0.55 (95% CI, 0.37% to 0.84), respectively. The effect
In the zinc- of zinc was not enhanced by concomitant vitamin A
caretaker administration
group, zinc
was admin-
istered by
the care-
taker, who
knew that
the child
B. A. Haider and Z. A. Bhutta
was receiv-
ing zinc
Bangladesh, 3–59 mo Children with n = 2,483 n = 2,502 Intervention: zinc 20 mg/day and About 40% (399/1,007) of diarrheal episodes were treated
2002 acute diarrhea ORS with zinc in the first 4 mo of the trial; the rate rose to
[15] Baqui Control: ORS 67% (350/526) in month 5 and to > 80% (364/434) in mo
7 and was sustained at that level. Children from the inter-
vention cluster had a shorter duration (hazard ratio, 0.76;
95% CI, 0.65 to 0.90) and lower incidence (rate ratio,
0.85; 95% CI, 0.76 to 0.96) of diarrhea than children in
the comparison group. Admission to hospital of children
with diarrhea was lower in the intervention group than
in the comparison group (rate ratio 0.76;95% CI, 0.59 to
0.98)
Brazil, 2003 3–60 mo Children with n = 37 n = 37 Intervention: elemental zinc 22.5 Reduction in duration of diarrhea (1.2 ± 0.8 vs. 2.5 ± 1.8
[7] Al-Sonboli acute diarrhea < mg/day (age 3–6 mo) and 45 mg/ days, p< .001) and duration of watery stools (0.4 ± 0.6 vs.
7 days and mild day (age 7–60 mo) 1.3 ± 1.5 days, p< .001) in the zinc-supplemented group.
dehydration Control: placebo group received Effect was more marked in children with low serum zinc
vitamin C levels
Turkey, 2003 2–29 mo Malnourished n = 52 n = 54 Intervention: 20 mg zinc per day The mean duration of diarrhea was shorter and the per-
[16] Polat (weight-for-age < Control: Placebo centage of children with consistent diarrhea for more
76th percentile of than 3–7 days was lower in the study group than in the
NCHS) children control group. Prolonged diarrhea was present in 12%
with normal zinc of children in the study group and in 37% of children in
levels suffering the control group. Stool frequency during the first 4 days
from acute watery after enrollment was lower in children in the study group
diarrhea
India, 2004 3–36 mo Male children with n = 143 n = 144 Intervention: zinc sulfate (15 mg/ Zinc treatment reduced total stool output (ratio of geomet-
[17] Bhatnagar acute diarrhea day for those aged up to 12 mo, ric means, 0.69; 95% CI, 0.48 to 0.99) and stool output
≤ 3 days with 30 mg/day for those older) and per day of diarrhea (ratio of geometric means, 0.76; 95%
mild or severe vitamin B complex (vitamins B, CI, 0.59 to 0.98). The risk of continued diarrhea was
dehydration C, D; and niacinamide) lower (relative hazard, 0.76; 95% CI, 0.59 to 0.97). Zinc
Control: vitamin B complex (vita- supplementation also reduced the proportion of diarrheal
mins B, C, D; and niacinamide) episodes lasting ≥ 5 days (odds ratio, 0.49; 95% CI, 0.25 to
0.97) or ≥ 7 days (odds ratio, 0.09; 95% CI, 0.01 to 0.73)
India, 2005 6–59 mo Children with n = 102 n = 98 Intervention: 40 mg/day zinc sul- The mean survival time with diarrhea was not significantly
[19] Patel acute diarrhea < fate and 5 mg/day copper sulfate different in the treatment group (4.34 ± 0.2 [SE] days)
7 days and ORS and the placebo group (4.48 ± 0.2 days), nor was there
Control: ORS any difference in the median time to cure. Cure was less
likely if the duration of diarrhea prior to enrollment was
greater (p < .001), if the time taken for rehydration was
greater (p = .001), and if intravenous fluids were used (p
= .03), regardless of the micronutrient supplementation.
The proportion of children with diarrhea > 4 days was
46% in the placebo group, with an adjusted odds ratio of
1.19 (95% CI, 1.58 to 0.9; p = 0.2), as compared with 39%
S45
in the supplemented group

continued
TABLE 1. Randomized, controlled trials of the use of zinc in the treatment of acute diarrhea (continued) S46
Country, year
Bangladesh, 1–6 mo Hospitalized male n = 86 n = 89 Intervention: (a) 20 mg zinc Neither duration of diarrhea nor mean stool volume dif-
2005 infants with acetate/day fered between groups. There were no significant differ-
[20] Brooks acute diarrhea ≤ (b) 5 mg zinc acetate/day ences between the groups in fluid intake, the need for
3 days and some Control: Placebo unscheduled intravenous fluid, weight gain, or vomiting
dehydration rates
Australia, 2005 < 11 yr Children hospital- (a) Zinc 108 Intervention: (a) zinc, (b) vitamin Supplementation with zinc, vitamin A, or combined zinc
[21] Valery ized with acute (n = 107) A, (c) zinc + vitamin A and vitamin A had no significant effect on duration of
diarrhea (b) Vitamin Control: placebo diarrhea or rate of readmission compared with placebo.
A (n = 109) For children < 12 mo, vitamin A The median duration of diarrhea after starting sup-
(c) Zinc + 50,000 IU (days 1 and 5) and zinc plementation was 3.0 days for the vitamin A group, the
vitamin A sulfate 20 mg daily for 5 days; zinc group, and the placebo group (p = 0.25 and 0.69,
(n = 112) for children 1–10 yr, vitamin A respectively, for the comparison of the vitamin A group
100,000 IU (days 1 and 5) and and the zinc group with the placebo group). The number
zinc sulfate 40 mg daily for 5 of readmissions did not differ significantly between those
days receiving vitamin A or zinc and the relevant placebo
groups (relative risk for vitamin A, 1.2; 95% CI, 0.7 to 2.1;
relative risk for zinc, 1.3; 95% CI, 0.8 to 2.1)
Turkey, 2006 6–60 mo Acute diarrhea n = 150 n = 130 Intervention: ORS as required plus Mean duration of diarrhea was 3.02 ± 2 days in the zinc
[22] Boran < 14 days (a 15 mg zinc (for children 6–12 group and 3.67 ± 3.2 days in the control group. There was
randomized, mo) and 30 mg (for children no significant difference between treatment groups in the
open-label non- 12–60 mo) as zinc sulfate duration of diarrhea (p> .05). The number of stools after
placebo-control- Control: ORS alone starting supplementation was 5.8 ± 3.7 and 5.1 ± 3.9 on
led trial) day 1, 2.9 ± 1.6 and 3.0 ± 2.2 on day 2, and 1.8 ± 1.1 and
1.6 ± 0.9 on day 3 in the zinc and control groups, respec-
tively. There was no significant difference between treat-
ment groups in the severity of diarrhea (p > .05)
India, Pakistan, 1–5 mo Children with n = 554 n = 556 Intervention: zinc (10 mg/day for The geometric mean duration of the diarrhea episode
Ethiopia, 2006 acute watery 14 days) with ORS was 0.21 days longer among infants receiving zinc than
[25] Fischer diarrhea < 3 days Control: placebo and ORS among those receiving placebo, but this difference
Walker was not statistically significant, and no difference was
observed after controlling for sex, exclusive breastfeed-
ing, and LAZ. There were no differences in reported stool
frequency or the proportion of episodes lasting > 7 days
Multiple coun- 2–59 mo Children with n = 1,010 n = 992 Intervention: zinc (20 mg/day for In 5 of 6 sites, ORS use in cases with continued diarrhea on
tries, 2006 acute watery 14 days) with ORS days 3–5 was the same in the 2 groups or higher in the
[23] Awasthi diarrhea < 7 days Control: ORS alone zinc group. Overall adherence to zinc supplementation
(Effective- (Outcomes was 83.8% (95% CI, 81% to 86%). Overall, less antibiotic
ness trials in reported included or antidiarrheal use occurred in the zinc group (absolute
Brazil, Ethio- ORS use on days difference, 3.8%; 95% CI, 1.7 to 5.9)
pia, Egypt, 3–5, adherence to
India, and zinc, and any use
Philippines) of an antibacterial
or antidiarrheal
up to day 14)
Philippines, 2–59 mo Children with n = 60 n = 57 Intervention: zinc (20 mg/day for There was significantly shorter duration of diarrhea from
2007 acute watery 14 days) with ORS the time of consultation in the zinc than in the ORS group
[24] Gregorio diarrhea < 7 days Control: ORS alone (mean ± SD, 2.98 ± 0.92 vs. 3.67 ± 1.63 days, p = .009) with
and no evidence a difference of 0.69 days (16.6 h). More patients in the zinc
of dehydration group than in the control group had diarrhea lasting < 4
days from admission (54/59 [92%] vs. 43/57 [75%]), but
the difference was not significant (p = .780)
Bangladesh, 12–59 mo Moderately n = 28 n = 28 Intervention: 20 mg/day elemental Children receiving zinc recovered significantly faster than
2008 malnourished zinc and a multivitamin contain- the control children (p< .05). The median numbers of days
[18] Roy children with cul- ing vitamins A and D, thiamine, to recovery and to disappearance of blood and mucus were
ture-confirmed riboflavin, nicotinamide, and 50% shorter in the zinc group than in the control group,
shigellosis calcium at twice the RDA and these differences were significant. The mean body
Control: Same multivitamins alone weight of zinc-supplemented children increased signifi-
Standard antibiotic therapy was cantly (p< .01) from 8.8 kg on admission to 9.2 kg at recov-
given to both groups ery; the body weight of the control children increased from
9.3 to 9.6 kg, a nonsignificant change (p = .12).
Bangladesh, 3–14 yr Children with n = 82 n = 82 Intervention: zinc (30 mg/day) More patients in the zinc group than in the control group
2008 cholera Control: placebo recovered by 2 days (49% vs. 32%, p = .032) and by 3 days
(26) Roy (81% vs. 68%, p = .03). Zinc-supplemented patients had
a 12% shorter duration of diarrhea than control patients
(64.1 vs 72.8 h, p = .028) and 11% less stool output (1.6
vs. 1.8 kg/day, p = .039)
LAZ, length-for-age z-score; NCHS, National Center for Health Statistics; ORS, oral rehydration solution; RDA, recommended dietary allowance; RE, retinol equivalent
S47
S48
Review: Therapeutic zinc supplementation

Comparison: 01 Zinc supplementation vs. placebo in the treatment of acute diarrhea
Outcome: 01 Mean duration of acute diarrhea in days (all age groups)
Study Treatment Control WMD (random) Weight WMD (random)

or subcategory N Mean ± SD N Mean ± SD 95% CI % 95% CI
01 All age groups

Al-Sonboli 2003 [7] 37 1.20 ± 0.80 37 2.50 ± 1.80 6.94 –1.30 (–1.93 to –0.67)
Bahl 2002 [13] 402 1.60 ± 2.20 401 1.70 ± 2.50 8.66 –0.10 (–0.43 to 0.23)
Bhatnagar 2004 [17] 132 2.32 ± 1.54 134 2.69 ± 1.90 8.21 –0.37 (–0.79 to 0.05)
Boran 2006 [22] 150 3.02 ± 2.00 130 3.67 ± 3.20 6.93 –0.65 (–1.29 to –0.01)
Brooks 2005 [20] 86 5.00 ± 4.73 89 5.00 ± 4.81 3.35 –0.00 (–1.41 to 1.41)
Dutta 2000 [6] 44 2.93 ± 0.42 36 4.31 ± 0.71 8.94 –1.38 (–1.64 to –1.12)
Gregorio 2007 [24] 60 2.98 ± 0.92 57 3.67 ± 1.63 7.83 –0.69 (–1.17 to –0.21)
Hidayat 1998 [11] 739 3.50 ± 2.40 659 3.80 ± 2.60 8.94 –0.30 (–0.56 to –0.04)
Patel 2005 [19] 102 4.34 ± 2.02 98 4.48 ± 1.98 7.42 –0.14 (–0.69 to 0.41)
Polat 2003 [16] 52 4.70 ± 1.40 54 5.20 ± 1.60 7.31 –0.50 (–1.07 to 0.07)
Roy 1997 [10] 37 5.10 ± 2.50 37 5.50 ± 2.70 4.15 –0.40 (–1.59 to 0.79)
Sachdev 1988 [8] 25 3.40 ± 1.80 25 3.80 ± 1.70 5.10 –0.40 (–1.37 to 0.57)
Sazawal 1995 [9] 456 4.50 ± 3.60 481 5.40 ± 3.40 8.02 –0.90 (–1.35 to –0.45)
Fischer Walker 2006 [25] 554 4.93 ± 3.90 556 4.49 ± 3.17 8.19 –0.44 (0.02 to 0.86)
Subtotal (95% CI) 2,876 2,794 100.00 –0.50 (–0.82 to –0.18)
Test for heterogeneity: χ2 = 82.27, df = 13 (p < .00001), I2 = 84.2%
Test for overall effect: Z = 3.04 (p = .002)
–4 –2 0 2 4
Favors treatment Favors control
FIG. 1. Impact of zinc supplementation on mean duration of acute diarrhea, all age groups. WMD, weighted mean difference
Therapeutic zinc supplementation in children S49
months of age in two studies (WMD, 0.40; 95% CI, 0.00 Workers. Given the observed benefits on duration of
to 0.81; p = .05) (fig. 2). However, given that several diarrhea in repeated episodes in these large-scale trials,
previous studies found a beneficial impact of zinc sup- a potential preventive benefit of repeated therapeutic
plementation in infancy also included children under courses of zinc for diarrhea cannot be excluded. Given
6 months of age but did not clearly specify outcomes the average of three or four episodes of diarrhea per
by age group, the currently reported lack of benefit of child per year, treatment of all episodes of diarrhea with
zinc among children under 6 months of age needs to be zinc may be an efficient way of providing a reasonable
further verified by studies evaluating alternative doses amount of zinc to deficient populations and may be of
and formulations in this age group. some preventive benefit.
The impact of zinc supplementation on the severity
of acute diarrhea episodes was also evaluated, using sev-
eral different indicators of disease severity. The study Section 2
by Dutta et al. [6] reported lower total stool output
(1.5 ± 0.7 vs. 2.4 ± 0.7 kg, p = .0001) and lower intakes What is the effect of therapeutic zinc supplementation as
of oral rehydration solution (2.5 ± 1.0 vs. 3.6 ± 0.8 L, an adjunct in the treatment of children with persistent
p = .0001) in the zinc-supplemented group than in diarrhea on the duration and severity of the disease?
the control group. The study by Al-Sonboli et al. [7]
demonstrated a shorter duration of watery stools in Conclusions
the zinc-supplemented group (0.4 ± 0.6 vs. 1.3 ± 1.5
days, p < .001). Pooled analysis for the outcome of Five studies have been conducted to evaluate the
episodes lasting for 7 days or more was conducted, adjunctive therapeutic effect of zinc supplementation in
including data from six studies. Zinc supplementation children with persistent diarrhea. Pooled analysis of the
was associated with a marginally significant reduction available data has identified the beneficial effect of zinc
in acute diarrhea episodes that lasted for 7 days or more supplementation in reducing the duration of disease by
(six studies; relative risk, 0.68; 95% CI, 0.46 to 1.01, 0.68 day (WMD, −0.68; 95% CI, −1.01 to −0.36).
random-effects model) (fig. 3).
In recent years, three large-scale effectiveness trials Detailed review of the evidence
using zinc for the treatment of diarrhea have been
completed in India, Pakistan, and Mali through both Persistent diarrhea, which is defined as diarrhea lasting
the public and the private sector health systems. Pre- for more than 14 days, is associated with significant
liminary evidence suggests that in all instances there morbidity and mortality [28]. Thus, rapid institution
was increased zinc usage by child-care providers, with of effective treatment is needed for these children to
substantial impact on diarrhea outcomes (Bhutta ZA, prevent complications and death.
Soofi S, Huusain A, Black RE, personal communica- A published systematic review and meta-analysis
tion, 2007). In rural Haryana, India, Bhandari et al. [27] [5] of randomized, controlled trials of zinc supple-
demonstrated utilization of zinc supplements in 36.5% mentation as an adjunct in the treatment of persistent
(n = 1,571) and 59.8% (n = 1,649) of diarrheal episodes diarrhea in children under 5 years of age included
occurring in the 4 weeks preceding interviews in the data from four studies. The meta-analysis showed
intervention areas. The prevalence rates of diarrhea and a significant mean reduction of 42% (range, 10% to
pneumonia during the preceding 14 days were lower 63%) in the rate of treatment failure or death. The
in the intervention communities during the third quar- benefit appeared to be larger in younger children (< 12
terly survey (odds ratio for diarrhea, 0.56; 95% CI, 0.41 months of age) than in older children, in boys than in
to 0.75; odds ratio for pneumonia, 0.55; 95% CI, 0.25 girls, and in children who were wasted or had lower
to 1.25). The numbers of hospitalizations for any cause, plasma zinc levels.
diarrhea, and pneumonia in the preceding 3 months We repeated a literature search using the PubMed
were reduced in the intervention compared with the bibliographic search tool in early January 2008 to iden-
control areas (survey 3) (odds ratio for diarrhea hos- tify all trials of therapeutic zinc supplementation. Stud-
pitalizations, 0.69; 95% CI, 0.50 to 0.95; odds ratio for ies eligible for inclusion were randomized, controlled
pneumonia hospitalizations, 0.29; 95% CI, 0.15 to 0.54) trials that assessed the adjunctive therapeutic benefits
[27]). In a similar large-scale project in Matiari district of zinc supplementation in children under 5 years of
in Pakistan, zinc supplements delivered by primary care age with persistent diarrhea. The search identified five
government health workers (Lady Health Workers) to studies [29–33], the main characteristics of which are
children with acute diarrhea produced a reduction in summarized in table 2. One trial was conducted in Peru
the subsequent incidence rates of diarrhea and mor- and four in south Asia. Outcomes that were assessed
tality*. These results have led to the incorporation of
oral zinc sulfate for the treatment of diarrhea by the *Bhutta ZA, Soofi S, Huusain A, Black RE, WHO Zinc in
national primary care health program’s Lady Health Health Systems Workshop Delhi, February 2008.
S50

Comparison: 01 Zinc supplementation vs placebo in the treatment of acute diarrhea
Outcome: 02 Mean duration of acute diarrhea in days (<6 and >6 months age group)
Study Treatment Control WMD (fixed) Weight WMD (fixed)

01 Children <6 months of age

Brooks 2005 [20] 86 5.00 ± 4.73 89 5.00 ± 4.81 8.05 0.00 (–1.41 to 1.41)
Fischer Walker 2006 [25] 554 4.93 ± 3.90 556 4.49 ± 3.10 91.95 0.44 (0.02 to 0.86)
Subtotal (95% CI) 640 645 100.00 0.40 (0.00 to 0.81)
Test for heterogeneity: c2 = 0.34, df = 1 (p = .56), I2 = 0%
02 Children >6 months of age

Bahl 2002 [13] 402 1.60 ± 2.20 401 1.70 ± 2.50 44.51 –0.10 (–0.43 to 0.23)
Boran 2006 [22] 150 3.02 ± 2.00 130 3.67 ± 3.20 11.66 –0.65 (–1.29 to –0.01)
Patel 2005 [19] 102 4.34 ± 2.02 98 4.48 ± 1.98 15.37 –0.14 (–0.69 to 0.41)
Sachdev 1988 [8] 25 3.40 ± 1.80 25 3.80 ± 1.70 5.01 –0.40 (–1.37 to 0.57)
Sazawal 1995 [9] 456 4.50 ± 3.60 481 5.40 ± 3.40 23.44 –0.90 (–1.35 to –0.45)
Subtotal (95% CI) 1,135 1,135 100.00 –0.37 (–0.59 to –0.16)
Test for heterogeneity: c2 = 9.40, df = 4 (p = .05), I2 = 57.5%
–4 –2 0 2 4
FIG. 2. Impact of zinc supplementation on mean duration of acute diarrhea, according to age group. WMD, weighted mean difference
Comparison: 01 Zinc supplementation vs. placebo in the treatment of acute diarrhea
Outcome: 03 Number of episodes ≥ 7 days duration
Study Treatment Control RR (random) Weight RR (random)

or subcategory n/N n/N 95% CI % 95% CI
Bahl 2002 [13] 17/402 28/401 16.42 0.61 (0.34 to 1.09)

Bhatnagar 2004 [17] 1/132 9/134 3.26 0.11 (0.01 to 0.88)
Polat 2003 [16] 8/52 20/54 13.79 0.42 (0.20 to 0.86)
Sazawal 1995 [9] 70/456 90/481 22.63 0.82 (0.62 to 1.09)
Strand 2002 [14] 33/442 58/443 20.19 0.57 (0.38 to 0.86)
Fischer Walker 2006 [25] 135/538 109/536 23.71
23.71 1.23 (0.99 to 1.54)
Total (95% CI) 2,022 2,049 100.00 0.68 (0.46 to 1.01)

Total events: 264 (treatment), 314 (control)
Test for heterogeneity: χ2 = 23.25, df = 5 (p = .0003), I2 = 78.5%
0.1 0.2 0.5 1 2 5 10

FIG. 3. Impact of zinc supplementation on proportion of acute diarrhea episodes lasting for 7 days or more. WMD, weighted mean difference; RR, relative risk
S51
TABLE 2. Randomized, controlled trials of zinc in the treatment of persistent diarrhea S52
Country, year Age

[reference] author group Study participants Zinc Placebo Intervention Results
India, 1990 6–18 mo Diarrhea > 2 wk n = 20 n = 20 Intervention: 40 mg/day zinc sulfate A nonsignificant (19%) reduction in the duration of
[29] Sachdev Control: placebo diarrhea was observed
Both groups received oral nalidixic
acid and similar milk-free feeding
schedule
Bangladesh, 1998 3–24 mo Persistent diarrhea n = 95 n = 95 Intervention: 20 mg/day in 3 daily Overall, there was a nonsignificant reduction in dura-
[30] Roy and weight-for-age divided doses; zinc acetate and vita- tion of diarrhea. Duration of illness was significantly
< 76th percentile of mins A, B, D, and C reduced (33%) with zinc supplementation among
NCHS median Control: vitamins A, B, D, and C underweight children (≤ 70% weight-for-age, p = .03).
Supplemented children maintained their body weight
(5.72 vs. 5.70 kg, p = 0.62) during hospitalization,
unlike control group, which lost weight (5.75 vs. 5.67
kg, p = .05). Deaths: 1/95 in zinc group and 5/95 in
control group (p = .06)
Pakistan, 1999 6–36 mo Persistent diarrhea n = 43 n = 44 Intervention: 3 mg/kg/day (approx. There was no significant difference in the duration of
[31] Bhutta > 14 days and 20 mg/day) zinc sulfate and vita- diarrheal episodes. A trend toward shorter episodes
weight-for-age ≤ mins A, B, D, and C in children with lower plasma zinc concentrations at
−2 SD Control: vitamins A, B, D, and C baseline was seen
Peru, 1999 6–36 mo Persistent diarrhea > (a) n = n = 136 Intervention: (a) 20 mg/day zinc glu- Duration of illness was significantly reduced by 28%
[32] Penny 14 days 137 conate, (b) 20 mg/day zinc gluco- in children in the zinc group (p = .01) and by 33% in
(b) n = nate and oral multivitamins girls in the zinc + multivitamins group (p = .04)
139 Control: placebo
Meta-analysis of 4 trials, 2000 (5) Zinc Investigators Group (Roy 1998 [30], Bhutta 1999 [31], There was a 42% (95% CI, 10% to 63%) reduction in
Penny 1999 [32] and another unpublished trial from Bangladesh) treatment failure or death. Effect appeared to be more
marked in children who were aged < 12 mo, were male,
or had wasting or lower baseline plasma zinc levels
[33] Bangladesh, 2001 6 mo –2 yr Diarrhea > 14 days n = 24 n = 24 Intervention: (a) zinc (20 mg elemen- The mean daily stool outputs from days 2 to 7 of therapy
Khatun and moderately in each tal), (b) vitamin A 100,000 IU for were significantly less in the zinc and zinc plus vitamin
malnourished chil- group children < 1 yr and 200,000 IU for A groups, but not in the vitamin A group, in compari-
dren (61%–75% children > 1 yr , (c) both zinc and son with the control group. The rate of clinical recovery
of NCHS median vitamin A of children within 7 days was significantly greater in
weight-for-age) Control: placebo in 2 doses daily the zinc group (88%) than in the control group (46%,
All groups received a multivitamin p = .002) and the vitamin A group (50%, p = .005) but
syrup was not significantly different from that in the zinc plus
vitamin A group (67%, p = .086)
NCHS, National Center for Health Statistics

included duration of diarrhea, stool frequency, stool

volume, and body weight. All zinc-supplemented chil-
dren received zinc dosages to provide at least two times
the US RDA daily during the treatment period. Zinc-
–0.60 (–1.19 to –0.01)

–0.80 (–1.31 to –0.29)
–0.68 (–1.01 to –0.36)

supplemented children had better clinical outcomes
–0.40 (–1.67 to 0.87)
–0.50 (–1.57 to 0.57)

–0.80 (–1.76 to 0.16)
WMD (fixed)
in four of the five trials. A subsequent meta-analysis of
95% CI
the included studies showed that the mean duration of
persistent diarrhea was significantly less in the zinc-
supplemented group than in the placebo group (WMD,
–0.68 days; 95% CI, –1.01 to –0.36) (fig. 4). Serum zinc
was measured in all studies at baseline, but there were
no reported differences in the effect of the treatment
associated with baseline serum zinc status.
Weight
6.61
9.35
31.08
41.48
11.49
100.00
The impact of zinc supplementation on recovery
%
from persistent diarrhea was also evaluated. Summary
10
estimates, which were calculated from data from four
studies, showed that children in the zinc-supplemented
Favors control
group had a 21% lower probability of continuation of
FIG. 4. Impact of zinc supplementation on the mean duration of persistent diarrhea (days). WMD, weighted mean difference
5
diarrhea on a given day than the control group (rela-
tive hazard, 0.79; 95% CI, 0.65 to 0.96) (fig. 5). Sum-
WMD (fixed)
mary impact estimates on the severity indicators of
95% CI
0
the disease, such as stool frequency and stool output,
Favors treatment
could not be calculated because of the inconsistent
approaches used to analyze these outcomes. The effect
–5
of supplementation on the body weight of children at
the end of 2 weeks of treatment was reported in only
two studies, the analysis of which demonstrated a non-
–10
significant impact on this outcome (WMD, –0.09; 95%
CI, –0.40 to 0.22) (fig. 6).
Mean ± SD
5.50 ± 2.70
3.50 ± 1.40
3.00 ± 2.50
7.00 ± 3.80
4.50 ± 1.90
Section 3
Control
02 Zinc supplementation vs. placebo in the treatment of persistent diarrhea
What is the effect of therapeutic zinc supplementation as

an adjunct in the treatment of children with pneumonia
on the duration and severity of the disease?
44
44
95
20
136
N
Conclusions
Few clinical trials have addressed the therapeutic role of

01 Mean duration of persistent diarrhea (days)
zinc supplementation as an adjunct in the treatment of

5.10 ± 3.30
2.90 ± 1.40
2.20 ± 1.70
6.50 ± 3.70
3.70 ± 1.10
Mean ± SD
Treatment
339
pneumonia; there is insufficient information available

Test for heterogeneity: χ2 = 0.64, df = 4 (p = .96), I2 = 0%
to assess the effect of zinc on pneumonia outcomes.

Therapeutic zinc supplementation
Detailed review of the evidence

Test for overall effect: Z = 4.11 (p < .0001)
The exact mechanisms by which zinc may affect respi-

N
43
44
95
20
139
341
ratory infections are not well understood. These effects

may be modulated by impacts on the immune system
as well as cell membranes. Zinc is known to induce the
production of interferon and modulate inflammatory
cytokines, which in turn may have beneficial effects on
Sachdev 1990 [29]
Khatun 2001 [33]
Bhutta 1999 [31]
symptoms of respiratory infection [34].

Penny 1999 [32]
Roy 1998 [30]
or subcategory
Total (95% CI)
We identified all trials of therapeutic zinc supple-

Comparison:
mentation in children with pneumonia. Studies eligible

Outcome:
Review:
for inclusion were randomized, controlled trials that

Study
assessed the adjunctive therapeutic benefits of zinc

S54
Comparison: 02 Zinc supplementation vs. placebo in the treatment of persistent diarrhea
Outcome: 02 Recovery from persistent diarrhea
Study Relative hazard (fixed) Weight Relative hazard (fixed)

or subcategory log[relative hazard] ± SE 95% CI % 95% CI
Bhutta 1999 [31] –0.0202 ± 0.2719 13.56 0.98 (0.58 to 1.67)

Khatun 2001 [33] –0.7985 ± 0.2806 12.73 0.45 (0.26 to 0.78)
Penny 1999 [32] –0.1984 ± 0.1590 39.66 0.82 (0.60 to 1.12)
Roy 1998 [30] –0.1625 ± 0.1716 34.05 0.85 (0.61 to 1.19)
Total (95% CI) 100.00 0.79 (0.65 to 0.96)

Test for heterogeneity: χ2 = 4.89, df = 3 (p = 0.18), I2 = 38.6%
Test for overall effect: Z = 2.38 (p = 0.02)
0.2 0.5 1 2 5
FIG. 5. Impact of zinc supplementation on rates of recovery from persistent diarrhea

Comparison: 02 Zinc supplementation vs. placebo in the treatment of persistent diarrhea
Outcome: 03 Body weight on day 14 (kg)
Study Treatment Control WMD (fixed) Weight WMD (fixed)

Bhutta 1999 [31] 43 6.67 ± 1.43 44 7.13 ± 1.42 25.99 –0.46 (–1.06 to 0.14)
Roy 1998 [30] 74 5.72 ± 1.20 75 5.68 ± 1.00 74.01 0.04 (–0.31 to 0.39)
Total (95% CI) 117 119 100.00 –0.09 (–0.40 to 0.22)

–4 –2 0 2 4
Favors control Favors treatment
FIG. 6. Impact of zinc supplementation in children with persistent diarrhea on body weight at the end of 2 weeks. WMD, weighted mean difference
supplementation in children under 5 years of age suf- 15 years of age and one study reported only sex-based
fering from severe acute lower respiratory infection estimates [37]. Analysis showed nonsignificant effects
(ALRI) or pneumonia. We identified five studies that of zinc supplementation on the duration of hospitaliza-
evaluated whether zinc administered for a few days, tion (relative risk, 0.85; 95% CI, 0.71 to 1.02) (fig. 7),
along with antibiotics, affected the outcome of the duration of respiratory rate > 50/minute (relative risk,
disease. All trials except one were conducted in South 0.87; 95% CI, 0.73 to 1.03) (fig. 8), and chest indrawing
Asia. The characteristics of these studies [35–39] are (relative risk, 0.86; 95% CI, 0.70 to 1.04) (fig. 9).
presented in table 3. The outcomes assessed were
time to recovery from pneumonia symptoms, time to
complete recovery, time to recovery from respiratory Section 4
rate > 50/minute, duration of hospital stay, hypoxia,
and inability to eat. All treatment doses were at least 2 What is the effect of therapeutic zinc supplementation as
US RDA daily given for 5 or 6 days or until the child an adjunct in the treatment of children with malaria?
recovered from the current episode of pneumonia.
Children receiving zinc showed significantly faster Conclusions
recovery from pneumonia than those receiving placebo
in two of the five trials. Serum zinc was measured at On the basis of the available data, there is no evidence
baseline in all studies, and no differences in the effect that including zinc supplementation in the treatment
of treatment associated with baseline zinc status were of malaria affects the course of the illness.
reported. Summary estimates were calculated by the
generic inverse variance method utilizing data from Detailed review of the evidence
two studies [36, 38]. The data from three studies could
not be used in calculating summary estimates because A literature search for randomized, controlled trials
two of these studies [35, 39] included children up to assessing the impact of zinc supplementation in

Comparison: 03 Zinc supplementation vs. placebo in the treatment of pneumonia
Outcome: 01 Hospitalization
Study Risk ratio (fixed) Weight Risk ratio (fixed)

or subcategory log [risk ratio] ± SE 95% CI % 95% CI
Bose 2006 [38] –0.0726 ± 0.1171 59.42 0.93 (0.74 to 1.17)

Brooks 2004 [36] –0.2877 ± 0.1417 40.58 0.75 (0.57 to 0.99)
Total (95% CI) 100.00 0.85 (0.71 to 1.02)

0.1 0.2 0.5 1 2 5 10

Favors intervention Favors control
FIG. 7. Impact of zinc supplementation on the duration of hospitalization for pneumonia

Outcome: 02 Tachypnea (respiratory rate > 50 per minute)

or subcategory log[risk ratio ] ± SE 95% CI % 95% CI
Bose 2006 [38] –0.0304 ± 0.1169 58.29 0.97 (0.77 to 1.22)

Brooks 2004 [36] –0.3011 ± 0.1382 41.71 0.74 (0.56 to 0.97)
Total (95% CI) 100.00 0.87 (0.73 to 1.03)

Test for heterogeneity: χ2 = 2.24, df = 1 ( p = .13), I2 = 55.3%
Test for overall effect: Z = 1.61 ( p = .11)
0.1 0.2 0.5 1 2 5 10

FIG. 8. Impact of zinc supplementation on duration of tachypnea (respiratory rate > 50 per minute) in pneumonia
TABLE 3. Randomized, controlled trials of zinc in the treatment of pneumonia
S56
Country, year [refer- Age Study
ence] author group participants Zinc Control Interventions Results
India, 2002 9 mo–15 yr Hospitalized n = 42 n = 43 Intervention: zinc acetate 20 mg twice daily Time-to-event analysis using the Cox proportional-
[35] Mahalanabis children with Control: Placebo hazards model showed that the times needed for
clinically severe All patients received standard treatment with the resolution of fever and tachypnea, the return
measles accom- antibiotics and an initial 100,000 IU dose of of appetite, and the achievement of a “much
panied by pneu- vitamin A orally improved” or “cured” status were not different
monia who had between the 2 groups
been ill for ≤ 7
days
Bangladesh, 2004 2–23 mo Hospitalized chil- n = 132 n = 131 Intervention: elemental zinc 20 mg/day The group receiving zinc had reduced duration of
[36] Brooks dren with severe Control: placebo severe pneumonia (relative hazard, 0.70; 95% CI,
pneumonia All patients received the hospital’s standard 0.51 to 0.98), including duration of chest indrawing
antimicrobial treatment (relative hazard, 0.80; 95% CI, 0.61 to 1.05), respira-
tory rate > 50/min (relative hazard, 0.74; 95% CI,
0.57 to 0.98), and hypoxia (relative hazard, 0.79;
95% CI, 0.61 to 1.04), and overall hospital duration
(relative hazard, 0.75; 95% CI, 0.57 to 0.99). The
mean reduction is equivalent to 1 hospital day for
both severe pneumonia and time in hospital
India, 2004 2–24 mo Hospitalized chil- n = 39 n = 38 Intervention: (a) 10 mg zinc as acetate (twice Recovery rates in zinc-treated boys from very ill
[37] Mahalanabis dren with severe daily for 5 days) plus vitamin A placebo; status and from fever were 2.6 times (p = .004) and
ALRI (b) vitamin A 10,000 µg RE (twice daily 3 times (p = .003) greater than those in non-zinc-
for 4 days) plus zinc placebo, (c) zinc plus treated children; feeding difficulty and tachypnea
vitamin A, or were not significantly different between groups
Control: zinc and vitamin A placebos after an adjusted analysis
India, 2006 2–23 mo Hospitalized chil- n = 150 n = 149 Intervention: 20 mg zinc sulfate at the time There were no clinically or statistically significant
[38] Bose dren with severe of enrollment. From day 2, 10-mg tablets of differences in the duration of tachypnea, hypoxia,
pneumonia zinc sulfate chest indrawing, inability to feed, lethargy, severe
Control: placebo illness, or hospitalization. Zinc supplementation
All received standard therapy for severe was associated with a significantly longer duration
pneumonia of pneumonia in the hot season (p = .015)
Australia, 2006 Children Hospitalized chil- n = 111 n = 104 Intervention: (a) zinc plus vitamin A, (b) zinc There was no clinical benefit of supplementation
[39] Chang aged < dren with ALRI plus vitamin A placebo, (c) zinc placebo with vitamin A, zinc, or the 2 combined, with no
11 yr episodes plus vitamin A supplement. Zinc sulfate significant difference between zinc and no zinc,
(20 mg at age < 12 mo, 40 mg at age ≥12 vitamin A and no vitamin A, or zinc + vitamin A
mo) was administered daily for 5 days and and placebo groups in time to resolution of fever or
vitamin A was administered on days 1 and tachypnea, or duration of hospitalization. Children
5 after admission (50,000 IU at age < 12 mo, given zinc had increased risk of readmission for
100,000 IU at age ≥12 mo) ALRI within 120 days (relative risk, 2.4; 95% CI,
Control: zinc placebo plus vitamin A placebo 1.003 to 6.1)

ALRI, acute lower respiratory infection; RE, retinol equivalent

Outcome: 03 Chest indrawing

or subcategory log[risk ratio] ± SE 95% CI % 95% CI
Bose 2006 [38] –0.0834 ± 0.1439 48.16 0.92 (0.69 to 1.22)

Brooks 2004 [36] –0.2231 ± 0.1387 51.84 0.80 (0.61 to 1.05)
Total (95% CI) 100.00 0.86 (0.70 to 1.04)

Test for heterogeneity: χ2 = 0.49, df = 1 (p = 0.48), I2 = 0%
Test for overall effect: Z = 1.56 (p = 0.12)
0.5 0.7 1 1.5 2

FIG. 9. Impact of zinc supplementation on duration of chest indrawing in pneumonia
children suffering from malaria identified only one zinc group and from 66% to 41% in the placebo group.
multicenter study, which was conducted by the Zinc The predictors of changes in plasma zinc from admis-
Against Plasmodium Study Group [40] in five country sion to 72 hours included baseline C-reactive protein
sites (Ecuador, Ghana, Tanzania, Uganda, and Zambia). concentration, change in C-reactive protein concen-
Children aged 6 months to 5 years with fever and 2,000 tration, treatment group, study site, and baseline zinc
or more asexual forms of Plasmodium falciparum per concentration.
microliter in a thick blood smear were eligible for inclu-
sion. All children were treated with chloroquine and
were randomly assigned to receive either a relatively Section 5
high dose of zinc (20 mg/day for infants, 40 mg/day
for older children) or placebo for 4 days. The results What is the effect of therapeutic zinc supplementa-
showed no effect of zinc on the median time to reduction as an adjunct in the treatment of children with
tion of fever (zinc group, 24.2 hours; placebo group, tuberculosis?
24.0 hours; p = .37), the percentage of patients with a
reduction of at least 75% in parasitemia from baseline Conclusions
in the first 72 hours (73.4% of the zinc group and 77.6%
of the placebo group, p = .11), or change in hemoglobin No studies have been completed to evaluate the role of
concentration during the 3-day period of hospitaliza- zinc in the treatment of tuberculosis in children.
tion and the 4 weeks of follow-up. The mean plasma
zinc concentrations were low in all children at baseline Detailed review of the evidence
(zinc group, 55.9 ± 25.7 µg/dL; placebo group, 54.5 ±
21.3 µg/dL), but children who received zinc supple- Our literature search did not identify any randomized,
mentation had higher plasma zinc concentrations at 72 controlled trials evaluating the impact of zinc supple-
hours than did those who received placebo (71.6 ± 23.7 mentation among children with tuberculosis. However,
vs. 66.4 ± 21.3 µg/dL, p < .001). trials have been conducted in adults with tuberculosis
Low serum zinc levels are common in patients with and have shown beneficial effects when zinc is sup-
acute malaria. However, these levels revert to normal plemented along with other micronutrients. In a study
within a few days after clinical recovery from malaria. from India, 15 patients with pulmonary tuberculosis
Duggan et al. assessed the relation between plasma who received a zinc supplement were compared with 24
zinc concentration and the acute-phase response in an controls. The patients who received zinc had an earlier
observational cohort study of 689 children with acute sputum clearance than the control patients; however, the
falciparum malaria [41]. Plasma zinc was measured by difference was not statistically significant [42]. In a dou-
atomic absorption spectrophotometry. On admission, ble-blind, placebo-controlled trial in adults aged 15 to 55
70% of subjects had low plasma zinc (< 60 µg/dL). years in Indonesia [43], 40 patients newly diagnosed with
On multivariate analysis, the predictors of admission tuberculosis received either 5,000 IU of vitamin A (as
plasma zinc included admission levels of C-reactive retinyl acetate) and 15 mg zinc (as zinc sulfate) daily for
protein (a marker of the acute-phase response), parasite 6 months (micronutrient group) or a placebo, in addition
density, and study site. The children were randomly to antituberculosis treatment. The micronutrient-supple-
assigned to receive either zinc supplements or a pla- mented group had significantly higher rates of sputum
cebo. The proportion of children with low plasma conversion and radiologic resolution of lesions, although
zinc at 72 hours decreased from 73% to 30% in the the results could not be attributed to zinc alone.
S58 B. A. Haider and Z. A. Bhutta
Summary Comparable evidence of benefits of zinc in the

treatment of pneumonia is not available. However, the
These studies and updated analyses of the effect of marked reductions in the prevalence of pneumonia
including zinc in the treatment of diarrhea corrobo- and the rate of hospitalization for pneumonia in one
rate existing reviews and indicate that the evidence effectiveness trial of zinc treatment for diarrhea sug-
is both consistent and robust. The two effectiveness gest that repeated courses of zinc for the treatment of
trials in Asia and the one in Africa indicate that scal- diarrheal episodes may also reduce the incidence and
ing up the use of zinc in health systems is feasible and severity of lower respiratory infections. It can there-
has demonstrable benefits. Thus, for the treatment of fore be argued that the distinction between preventive
diarrhea, the need is to implement the revised diarrhea and therapeutic uses of zinc for a range of disorders
treatment strategy, including low-osmolality oral rehy- (especially acute diarrhea and respiratory infections
dration solution and zinc in all cases. It is possible that in high-burden communities) may become blurred
these benefits may also accrue in developed countries, with repeated use.
but there are few studies of zinc supplementation in Given that our current knowledge of the mecha-
developed countries, and the benefit of zinc in such nisms of benefit of zinc in these diverse disorders is
circumstances needs further evaluation. Current limited, there is an urgent need for better studies of
evidence indicates that zinc treatment of infants with the mechanisms of action of zinc in such disorders to
diarrhea has no beneficial impact, although this needs enable better understanding of the range of disorders
further evaluation. and populations for which zinc may be beneficial.
References
1. International Zinc Nutrition Consultative Group Nutr 1988;7:877–81.
(IZiNCG), Brown KH, Rivera JA, Bhutta Z, Gibson 9. Sazawal S, Black RE, Bhan MK, Bhandari N, Sinha A, Jalla
RS, King JC, Lönnerdal B, Ruel MT, Sandtröm B, S. Zinc supplementation in young children with acute
Wasantwisut E, Hotz C. Assessment of the risk of zinc diarrhea in India. N Engl J Med 1995;333:839–44.
deficiency in populations and options for its control. 10. Roy SK, Tomkins AM, Akramuzzaman SM, Behrens
Food Nutr Bull 2004;25(1 suppl 2):S99–203. RH, Haider R, Mahalanabis D, Fuchs G. Randomised
2. Black RE, Allen LH, Bhutta ZA, Caulfield LE, de Onis controlled trial of zinc supplementation in malnour-
M, Ezzati M, Mathers C, Rivera J, for the Maternal and ished Bangladeshi children with acute diarrhoea. Arch
Child Undernutrition Study Group. Maternal and child Dis Child 1997;77:196–200.
undernutrition: Global and regional exposures and 11. Hidayat AA, Sunoto A, Soedarmo SP. The effect of
health consequences. Lancet 2008;371:243–60. zinc supplementation in children under three years of
3. Horton S, Alderman H, Rivera JA. Copenhagen con- age with acute diarrhea in Indonesia. Med J Indonesia
sensus challenge paper: Hunger and malnutrition. 1998;7:237–241.
Copenhagen: Copenhagen Consensus Center, 2008. 12. Faruque AS, Mahalanabis D, Haque SS, Fuchs GJ, Habte
4. Caulfield LE, Black RE. Zinc deficiency. In: Ezzati M, D. Double-blind, randomized, controlled trial of zinc or
Lopez AD, Rodgers A, Murray CJL, eds. Compara- vitamin A supplementation in young children with acute
tive quantification of health risks: Global and regional diarrhoea. Acta Paediatr 1999;88:154–60.
burden of disease attributable to select major risk fac- 13. Bahl R, Bhandari N, Saksena M, Strand T, Kumar GT,
tors. Geneva: World Health Organization, 2004. Bhan MK, Sommerfelt H. Efficacy of zinc-fortified oral
5. Zinc Investigators’ Collaborative Group (Bhutta ZA, rehydration solution in 6- to 35-month-old children
Bird SM, Black RE, Brown KH, Gardner JM, Hidayat with acute diarrhea. J Pediatr 2002;141:677–82.
A, Khatun F, Martorell R, Ninh NX, Penny ME, 14. Strand TA, Chandyo RK, Bahl R, Sharma PR, Adhikari
Rosado JL, Roy SK, Ruel M, Sazawal S, Shankar A). RK, Bhandari N, Ulvik RJ, Molbak K, Bhan MK, Som-
Therapeutic effects of oral zinc in acute and persistent merfelt H. Effectiveness and efficacy of zinc for the
diarrhea in children in developing countries: Pooled treatment of acute diarrhea in young children. Pediatrics
analysis of randomized controlled trials. Am J Clin Nutr 2002;109:898–903.
2000;72:1516–22. 15. Baqui AH, Black RE, El Arifeen S, Yunus M, Chakraborty
6. Dutta P, Mitra U, Datta A, Niyogi SK, Dutta S, Manna B, J, Ahmed S, Vaughan JP. Effect of zinc supplementation
Basak M, Mahapatra TS, Bhattacharya SK. Impact of zinc started during diarrhoea on morbidity and mortality
supplementation in malnourished children with acute in Bangladeshi children: Community randomised trial.
watery diarrhoea. J Trop Pediatr 2000;46:259–63. BMJ 2002;325:1059.
7. Al-Sonboli N, Gurgel RQ, Shenkin A, Hart CA, Cuevas 16. Polat TB, Uysalol M, Cetinkaya F. Efficacy of zinc
LE. Zinc supplementation in Brazilian children with supplementation on the severity and duration of
acute diarrhoea. Ann Trop Paediatr 2003;23:3–8. diarrhea in malnourished Turkish children. Pediatr Int
8. Sachdev HP, Mittal NK, Mittal SK, Yadav HS. A control- 2003;45:555–9.
led trial on utility of oral zinc supplementation in acute 17. Bhatnagar S, Bahl R, Sharma PK, Kumar GT, Saxena SK,
dehydrating diarrhea in infants. J Pediatr Gastroenterol Bhan MK. Zinc with oral rehydration therapy reduces
stool output and duration of diarrhea in hospitalized 31. Bhutta ZA, Nizami SQ, Isani Z. Zinc supplementation
children: A randomized controlled trial. J Pediatr Gas- in malnourished children with persistent diarrhea in
troenterol Nutr 2004;38:34–40. Pakistan. Pediatrics 1999;103:e42.
18. Roy SK, Raqib R, Khatun W, Azim T, Chowdhury 32. Penny ME, Peerson JM, Marin RM, Duran A, Lanata
R, Fuchs GJ, Sack DA. Zinc supplementation in the CF, Lönnerdal B, Black RE, Brown KH. Randomized,
management of shigellosis in malnourished children in community-based trial of the effect of zinc supplemen-
Bangladesh. Eur J Clin Nutr 2008;62:849–55. tation, with and without other micronutrients, on the
19. Patel AB, Dhande LA, Rawat MS. Therapeutic evaluation duration of persistent childhood diarrhea in Lima, Peru.
of zinc and copper supplementation in acute diarrhea in J Pediatr 1999;135:208–17.
children: Double blind randomized trial. Indian Pediatr 33. Khatun UH, Malek MA, Black RE, Sarkar NR, Wahed
2005;42:433–42. MA, Fuchs G, Roy SK. A randomized controlled clini-
20. Brooks WA, Santosham M, Roy SK, Faruque AS, Wahed cal trial of zinc, vitamin A or both in undernourished
MA, Nahar K, Khan AI, Khan AF, Fuchs GJ, Black RE. children with persistent diarrhea in Bangladesh. Acta
Efficacy of zinc in young infants with acute watery Paediatr 2001;90:376–80.
diarrhea. Am J Clin Nutr 2005;82:605–10. 34. Shankar AH, Prasad AS. Zinc and immune function:
21. Valery PC, Torzillo PJ, Boyce NC, White AV, Stewart PA, The biological basis of altered resistance to infection.
Wheaton GR, Purdie DM, Wakerman J, Chang AB. Zinc Am J Clin Nutr 1998;68:447S–463S.
and vitamin A supplementation in Australian Indig- 35. Mahalanabis D, Chowdhury A, Jana S, Bhattacharya MK,
enous children with acute diarrhoea: A randomised Chakrabarti MK, Wahed MA, Khaled MA. Zinc supple-
controlled trial. Med J Aust 2005;182:530–5. mentation as adjunct therapy in children with measles
22. Boran P, Tokuc G, Vagas E, Oktem S, Gokduman MK. accompanied by pneumonia: A double-blind, rand-
Impact of zinc supplementation in children with acute omized controlled trial. Am J Clin Nutr 2002;76:604–7.
diarrhoea in Turkey. Arch Dis Child 2006;91:296–9. 36. Brooks WA, Yunus M, Santosham M, Wahed MA, Nahar
23. Awasthi S. Zinc supplementation in acute diarrhea is K, Yeasmin S, Black RE. Zinc for severe pneumonia in
acceptable, does not interfere with oral rehydration, and very young children: Double-blind placebo-controlled
reduces the use of other medications: A randomized trial. Lancet 2004;363:1683–8.
trial in five countries. J Pediatr Gastroenterol Nutr 37. Mahalanabis D, Lahiri M, Paul D, Gupta S, Gupta A,
2006;42:300–5. Wahed MA, Khaled MA. Randomized, double-blind,
24. Gregorio GV, Dans LF, Cordero CP, Panelo CA. Zinc placebo-controlled clinical trial of the efficacy of treat-
supplementation reduced cost and duration of acute ment with zinc or vitamin A in infants and young chil-
diarrhea in children. J Clin Epidemiol 2007;60:560–6. dren with severe acute lower respiratory infection. Am
25. Fischer Walker CL, Bhutta ZA, Bhandari N, Teka T, J Clin Nutr 2004;79:430–6.
Shahid F, Taneja S, Black RE, the Zinc Study Group. Zinc 38. Bose A, Coles CL, Gunavathi, John H, Moses P,
supplementation for the treatment of diarrhea in infants Raghupathy P, Kirubakaran C, Black RE, Brooks WA,
in Pakistan, India and Ethiopia. J Pediatr Gastroenterol Santosham M. Efficacy of zinc in the treatment of severe
Nutr 2006;43:357–63. pneumonia in hospitalized children < 2 y old. Am J Clin
26. Roy SK, Hossain MJ, Khatun W, Chakraborty B, Chowd- Nutr 2006;83:1089–96; quiz 1207.
hury S, Begum A, Mah-e-Muneer S, Shafique S, Khanam 39. Chang AB, Torzillo PJ, Boyce NC, White AV, Stewart
M, Chowdhury R. Zinc supplementation in children PM, Wheaton GR, Purdie DM, Wakerman J, Valery
with cholera in Bangladesh: Randomised controlled PC. Zinc and vitamin A supplementation in Indigenous
trial. BMJ 2008;336:266–8. Australian children hospitalised with lower respiratory
27. Bhandari N, Mazumder S, Taneja S, Dube B, Agar- tract infection: A randomised controlled trial. Med J
wal RC, Mahalanabis D, Fontaine O, Black RE, Bhan Aust 2006;184:107–12.
MK. Effectiveness of zinc supplementation plus oral 40. Zinc Against Plasmodium Study Group. Effect of zinc
rehydration salts compared with oral rehydration salts on the treatment of Plasmodium falciparum malaria in
alone as a treatment for acute diarrhea in a primary children: A randomized controlled trial. Am J Clin Nutr
care setting: A cluster randomized trial. Pediatrics 2002;76:805–12.
2008;121:e1279–85. 41. Duggan C, MacLeod WB, Krebs NF, Westcott JL, Fawzi
28. Bhutta ZA, Ghishan F, Lindley K, Memon IA, Mittal S, WW, Premji ZG, Mwanakasale V, Simon JL, Yeboah-
Rhoads JM. Persistent and chronic diarrhea and malab- Antwi K, Hamer DH. Plasma zinc concentrations are
sorption: Working Group report of the Second World depressed during the acute phase response in children
Congress of Pediatric Gastroenterology, Hepatology, and with falciparum malaria. J Nutr 2005;135:802–7.
Nutrition. J Pediatr Gastroenterol Nutr 2004;39(suppl 42. Pant K, Biswas SK, Chawla R, Shah A, Singh MM. Zinc
2):S711–6. in active pulmonary tuberculosis. Indian J Chest Dis
29. Sachdev HP, Mittal NK, Yadav HS. Oral zinc supple- Allied Sci 1987;29:144–9.
mentation in persistent diarrhoea in infants. Ann Trop 43. Karyadi E, West CE, Schultink W, Nelwan RH, Gross R,
Paediatr 1990;10:63–9. Amin Z, Dolmans WM, Schlebusch H, van der Meer JW.
30. Roy SK, Tomkins AM, Mahalanabis D, Akramuzzaman A double-blind, placebo-controlled study of vitamin A
SM, Haider R, Behrens RH, Fuchs G. Impact of zinc sup- and zinc supplementation in persons with tuberculosis
plementation on persistent diarrhoea in malnourished in Indonesia: Effects on clinical response and nutritional
Bangladeshi children. Acta Paediatr 1998;87:1235–9. status. Am J Clin Nutr 2002;75:720–7.
Effects of maternal zinc supplementation on pregnancy
and lactation outcomes
Sonja Y. Hess and Janet C. King
Abstract for reducing the risk of premature delivery, the possible

positive impact of zinc supplementation on infant birth-
Observational studies in human populations suggest weight among undernourished women, and the lack
that maternal zinc deficiency during pregnancy may of reported adverse effects, zinc should be included in
cause adverse pregnancy outcomes for the mother and maternal supplements given during pregnancy in popula-
fetus. Therefore, we reviewed the current evidence from tions at risk for zinc deficiency.
studies of zinc supplementation, with or without other
micronutrients, during pregnancy and lactation to assess
its impact on maternal, fetal, and infant health. Key words: Lactation, maternal health, neonatal
A meta-analysis of supplementation trials indicates health, pregnancy, pregnancy outcome, prevention, zinc
a 14% reduction in premature delivery among zinc- deficiency, zinc supplementation
supplemented women. Most studies found no significant
impact of maternal zinc supplementation on infant birth-
weight, but a subset of studies conducted in underweight Background
or zinc-deficient women suggests that there may be a
positive effect of zinc supplementation in such women. The results of experimental studies conducted in ani-
However, the number of relevant studies is limited, and mals and observational studies in human populations
more information is needed to confirm these observa- show that maternal zinc deficiency can have adverse
tions. The results for other pregnancy outcomes are effects on reproduction, including infertility, congenital
inconsistent, and the number of available studies is small. anomalies, fetal growth retardation, prolonged labor,
Likewise, the impact of maternal zinc supplementation embryonic or fetal death, and early postnatal infant
during pregnancy on infant postnatal growth and risk immune dysfunction. The possible mechanisms and
of infection is variable, and few studies are available. pathways of maternal zinc deficiency and adverse
Thus, more research will be needed to allow definitive health effects on the mother and fetus were previously
conclusions to be drawn, especially for the second half of reviewed [1]. On the basis of the new recommendations
infancy and later childhood. by the World Health Organization (WHO), the United
Studies found no adverse effects of maternal zinc Nations Children’s Fund (UNICEF), the International
supplementation on iron status during pregnancy. More Atomic Energy Agency (IAEA), and the International
information is required on other potential adverse effects, Zinc Nutrition Consultative Group (IZiNCG) to use
particularly with regard to a possible modifying effect of stunting rates of children under five as an indica-
preexisting maternal zinc status. tor for the risk of zinc deficiency, it is estimated that
In view of the possible benefits of zinc supplementation approximately one-third of the world’s population live
in countries where the risk of zinc deficiency is high
[2]. The prevalence of zinc deficiency among pregnant
and lactating women worldwide is unknown.
Sonja Y. Hess is affiliated with the Department of Nutrition
and the Program in International and Community Nutrition, Zinc requirements during pregnancy and lactation
University of California, Davis, California, USA; Janet C. King have been estimated from the zinc content of tissues
is affiliated with the Children’s Hospital Oakland Research accrued during pregnancy and the zinc content of
Institute, Oakland, California, USA. milk secreted during lactation [3]. The estimated total
Please direct queries to the corresponding author: Janet C.
King, Children’s Hospital Oakland Research Institute, 5700 additional zinc needed for pregnancy is ~100 mg [4].
Martin Luther King Jr. Way, Oakland, CA 94609, USA; e-mail: In addition to the zinc accrued by the fetus, zinc is
jking@chori.org. deposited in the placenta, amniotic fluid, and uterine
S60 Food and Nutrition Bulletin, vol. 30, no. 1 © (supplement) 2009, The United Nations University.
Zinc supplementation during pregnancy and lactation S61
and mammary tissue. Approximately 60% of the total Section 1

zinc is accumulated in the conceptus and 40% in the
maternal tissue. The additional daily need increases What is the effect of preventive zinc supplementation
during gestation to meet the demands for fetal growth, during pregnancy on maternal and neonatal health (i.e.,
rising from ~0.1 mg/day additional zinc in the first maternal mortality, maternal morbidity, fetal growth,
quarter of pregnancy to ~0.7 mg/day additional zinc and postnatal growth and morbidity)?
in the fourth quarter. As reviewed by Brown et al. [5],
the mean amount of zinc transferred in the breastmilk Conclusions
to exclusively breastfed infants declines rapidly from
~4 mg/day during the first few days of life to ~1.75 The results of experimental studies in laboratory ani-
mg/day by 1 month. Zinc transfer declines more mals and observational studies in human populations
slowly thereafter to ~1 mg/day at 6 months. Studies both suggest that maternal zinc deficiency during
suggest that maternal zinc absorption is increased pregnancy can cause adverse pregnancy outcomes for
and/or exogenous zinc excretion is decreased during the mother, fetus, and/or newborn infant postnatally.
pregnancy and lactation, thereby enhancing maternal A number of controlled intervention trials have now
zinc availability for fetal growth and milk zinc excre- been completed in humans, but the results are difficult
tion [6, 7]. However, the ability of these homeostatic to interpret because of the relatively small number of
mechanisms to compensate for diets that are low in studies reporting on each of these specific outcomes,
total zinc or in bioavailable zinc appears to be limited, the failure to characterize or stratify according to the
so reproductive function may be compromised under women’s preexisting zinc status and general nutritional
these circumstances. In these cases, supplementation condition, and the variable times of initiation, duration,
of women with low zinc intakes may be necessary to and amount of zinc supplementation. With recognition
ensure optimal reproductive outcomes. of these limitations in the existing evidence base, the
WHO currently recommends that all pregnant following conclusions can be drawn from the available
women in areas of high prevalence of malnutrition studies.
should routinely receive iron and folic acid supple- A meta-analysis of supplementation trials indicates
ments, together with appropriate dietary advice, to a small but significant positive impact of maternal
prevent anemia [8]. In view of the above-mentioned zinc supplementation on the duration of pregnancy
importance of zinc for human health and reproduc- and a 14% reduction in premature delivery among
tion, we reviewed the impact of zinc supplementa- zinc-supplemented women. Most available studies
tion on various reproductive outcomes in women found no significant impact of maternal zinc sup-
from developed and developing countries to evaluate plementation on infant birthweight, but a subset of
whether the addition of zinc to the iron and folic acid studies conducted in underweight or zinc-deficient
supplement should be considered. We compared the women suggests that there may be a positive effect of
results of controlled intervention trials in which zinc zinc supplementation in such women. However, these
was provided, with or without other micronutrients, to results are quite limited, and more studies are needed
pregnant women. We further examined the impact of to confirm these observations. There are inconsistent
zinc supplementation during lactation on maternal and results with regard to other pregnancy outcomes, and
infant zinc status and zinc-related outcomes. the number of available studies is small, so no defini-
This paper is divided into four sections, which tive conclusions are possible. Likewise, the impacts of
address the following questions in relation to zinc sup- maternal zinc supplementation during pregnancy on
plementation during pregnancy and lactation: infant postnatal growth and the risk of infection are
inconsistent and the number of studies is quite limited,
Section 1: What is the effect of preventive zinc supple-
so more research will be needed to allow definitive
mentation during pregnancy on maternal and neonatal
conclusions to be drawn, especially for the second half
health (i.e., maternal mortality, maternal morbidity,
of infancy and later childhood.
fetal growth, and postnatal growth and morbidity)?
Section 2: What is the effect of preventive zinc supple- Detailed review of evidence
mentation during lactation on maternal and neonatal
health? Bibliographic search
Data sets were identified for this section by using a
Section 3: Are there any adverse effects of zinc sup-
computerized bibliographic search (PubMed) with
plementation during pregnancy or lactation?
the key words zinc; limiting for human, English, clini-
Section 4: What are the implications of these outcomes cal trial, meta-analysis, randomized, controlled trial.
for zinc supplementation programs during pregnancy A total of 1,618 articles were identified during the
and lactation and what are the remaining research PubMed search using these key words. Three addi-
needs? tional manuscripts identified in other reviews of zinc
S62 S. Y. Hess and J. C. King
supplementation during pregnancy [9, 10] were also
Reduced no. of deliveries with

No effect on growth-support-
included, resulting in a total of 1,621 individual refer-
ing property of amniotic

No effect on preeclampsia
Effect of supplement
ences. All titles and abstracts were reviewed. Of these,
44 articles evaluated the impact of zinc supplementa-
tion on pregnancy-related outcomes in the mother or
infant. We excluded 2 of these 44 studies, 1 because the
complications
pregnant women chose the supplement themselves and
1 because the anemia status differed between treatment
No effect
No effect
groups. The remaining 42 articles were then screened
fluid
to combine those that presented data from the same
intervention trial by using key characteristics, such
as country, study population, and supplementation
Comparison
scheme. A total of 22 different trials were identified.
Placebo
n = 258
n = 379
n = 248
n = 248
n = 24
group
Maternal morbidity and mortality
The effect of supplemental zinc, with and without
other micronutrients, on maternal morbidity and
mortality was evaluated in 10 studies done over the
past 30 years [11–22] (table 1). With the exception of
n = 249
n = 415
n = 246
one study in Nepal [22, 23], the supplement was taken
n = 32
n = 64
group
Zinc
Zinc
daily from midpregnancy to term. In the study in Nepal
[22, 23], the supplement was started early in the first
trimester. An additional study in Nepalese pregnant
women investigated the impact of short-term zinc
From before wk 20
From before wk 20
From before wk 20
supplementation
supplementation (3 weeks) in women who reported
to term; length
Duration of
nancy to term
30 or 90 mg/day From midpreg-
night-blindness [18]. The first zinc supplementation
From booking
TABLE 1. Effect of zinc supplementation during pregnancy on maternal morbidity and mortality
study by Jameson and Ursing [11] was not randomized,

to term
to term
to term
and the investigators were not blinded to treatment. In
varied
that study of Swedish women with anemia, 90 mg of
supplemental zinc per day reduced complications at
delivery, such as prolonged labor or excessive bleeding.
Amount of zinc
Mahomed et al. [13] and Jonsson et al. [14] followed up

supplement
on this observation by conducting two double-blind,

20 mg/day
44 mg/day
90 mg/day
randomized, controlled trials of zinc supplementa- 20 mg/day

tion during pregnancy in England and Denmark,
Supplementation with zinc alone
respectively. The women were given either 20 or 44

mg of zinc/day from booking to term. The outcomes
Randomized, double- Black South Afri-
Swedish women
evaluated included maternal bleeding, hypertension,

Randomized, double- Chilean adoles-
Randomized, double- Danish women

Population
with anemia
cent women
can women
and complications of labor and delivery. There was no

Randomized, double- UK women
evidence in either of these randomized, controlled trials

that supplemental zinc affected maternal morbidity
during pregnancy. Simmer et al. [16], however, found a
lower incidence of induced labor (13% vs. 50%) among
zinc-supplemented (22.5 mg/day) UK women in a
not controlled, not
[15] Castillo-Durán blind, controlled
blind, controlled
blind, controlled
blind, controlled
small, double-blind study with a total sample size of

Study design
Not randomized,
56 women. In contrast, Dijkhuizen and Wieringa [17]

found significantly more deliveries with complications
blinded
in the group receiving zinc (30 mg/day) plus iron and

folic acid than in the groups receiving iron and folic
acid alone, iron and folic acid plus β-carotene, or iron
and folic acid plus β-carotene and zinc. The differences
South Africa, 1979
[reference] author
among these studies may be related to the amount of

[12] Appelbaum
Denmark, 1996
[13] Mahomed
supplemental zinc (90 vs. 20 or 44 mg/day), the initial

Country, year
Sweden, 1976
[11] Jameson
[14] Jonsson
Chile, 2001
zinc status of the mothers, the presence of other micro-

UK, 1989
nutrient supplements, obstetric practices, or investiga-

tor bias in the case of the nonblinded trial.
Hunt and coworkers [19] evaluated the effects of
Folic acid +
Supplementation with folic acid, iron, and zinc, with or without β-carotene iron + zinc Folic acid + iron
Indonesia, 2001 Randomized, double- Indonesian rural 30 mg/day From before wk 20 Folic acid + Folic acid + iron Increased no. of deliveries with
[17] Dijkhuizen blind, controlled women to term iron + zinc (n = 42) complications
(n = 48)
β-Carotene + β-Carotene + folic No effect
folic acid + acid + iron
iron + zinc (n = 45)
(n = 44)
Nepal, 2003 Cluster-randomized, Nepalese rural 30 mg/day From wk 5–10 Folic acid + Placebo, folic acid, AGP concentration decreased
[22] Christian double-blind, women gestation to 3 mo iron + zinc folic acid + iron, most in groups receiving
controlled postpartum (n = 173) or MMN folic acid, folic acid + iron,
(n = 606) and folic acid + iron + zinc
(p < .05). Subclinical infec-
tion (CRP concentration)
reduced in group receiving

folic acid + iron + zinc (p <
.05).
UK, 1991aa Randomized, double- Lower-social-class 22.5 mg/day From 15–25 wk n = 30 n = 26 Reduced induction of labor
[16] Simmer blind, controlled Englishwomen prior to delivery
Vitamin A/
carotene + Vitamin A/
Supplementation with vitamin A/carotenoid and zinc zinc carotene
Nepal, 2001 Randomized, double- Nepalese women 25 mg/day From wk 27 gesta- n = 102 n = 100 No effect on night-blindness
[18] Christian blind, controlled with night- tion for 3 wk
blindness
Supplementation with zinc and MMN MMN + zinc MMN
Tanzania, 2005 Randomized, double- HIV-infected 25 mg/day From wk 12–27 ges- n = 198 n = 199 No effect on parasitemia in
[21] Villamor blind, controlled pregnant Tanza- tation to term maternal and cord blood.
nian women
USA, 1984 Randomized, double- Low-income His- 20 mg/day From about wk 19 n = 107 n = 106 Reduced incidence of preg-
[19] Hunt blind, controlled panic California gestation to term nancy-induced hypertension
women
USA, 1995 Randomized, double- Medically indi- 25 mg/day From wk 19 gesta- n = 231 n = 206 No effect on serum homo-
[20] Hogg blind, controlled gent African- tion to term cysteine or pregnancy-
American induced hypertension
women
AGP, α1-acid glycoprotein; CRP, C-reactive protein; MMN, multiple micronutrients
S63
a. Iron and folic acid supplement was prescribed if clinically indicated.

20 mg of supplemental zinc/day during the last half of gestation or serum albumin concentration, in some
of gestation on the incidence of pregnancy-induced cases, or to their relatively small sample sizes. Despite
hypertension. The incidence was significantly lower in these limitations, serum zinc concentration is still the
the zinc-supplemented women (2% vs. 12%). Hogg et al. recommended biochemical indicator of zinc status
[20] also evaluated the effect of supplemental zinc (25 during pregnancy at the population level [2], and this
mg/day) in a larger population of poor African-Amer- indicator can be used to assess the impact of zinc sup-
ican women in Alabama and found no effect on the plementation in populations. A recent meta-analysis
incidence of pregnancy-induced hypertension between of nine studies of zinc supplementation in pregnant
the two groups or on serum levels of homocysteine, an women found a significantly positive overall effect
amino acid associated with occlusive vascular disease. of supplementation on mean serum zinc concentra-
Similarly, Castillo-Durán et al. [15] found no effect of tion, with an effect size of 0.20 SD (95% CI, 0.051
zinc supplementation on the incidence of preeclampsia to 0.348) [26].
in pregnant Chilean adolescents who received 20 mg of There are no studies of zinc supplementation and
supplemental zinc/day. A recent meta-analysis combin- maternal mortality. The few studies of maternal mor-
ing the results of seven randomized, controlled trials bidity reviewed here do not provide evidence that zinc
on pregnancy-induced hypertension and preeclampsia supplements alone consistently reduce complications of
found no significant differences (relative risk, 0.83; labor and delivery, maternal hypertension, or infection.
95% CI, 0.64 to 1.08) between the women receiving Since these pregnancy complications may be associated
zinc and the women in the control group [10]. These with placentation problems in early gestation, zinc
limited data do not support a role of zinc in reducing supplementation prior to conception also should be
hypertension or preeclampsia during gestation. evaluated in relation to maternal morbidity outcomes.
Zinc plays an important role in maintaining normal However, no studies of zinc supplementation prior to
immune function [24]. Therefore, maternal infection and during gestation have been done to date.
may be related to zinc status during pregnancy, and
three studies have examined maternal infectious mor- Fetal mortality and growth
bidity in relation to zinc supplementation. The effect of supplemental zinc on stillbirth or neo-
Supplemental zinc did not alter the microbial growth- natal death has been reported in 7 studies, and 17
promoting properties of amniotic fluid [12] or the risk studies evaluated its effect on fetal growth [13, 15–17,
of malaria parasitemia among HIV-infected women 19, 23, 29–43] (table 2). A recent meta-analysis found
[21]. Christian et al. [22] investigated the impact of no overall impact on the rate of stillbirth or neonatal
micronutrient supplementation during pregnancy on death [10].
the concentrations of acute-phase response proteins Most studies of supplemental zinc and fetal growth
studied before supplementation and at 32 weeks of have used birthweight as the endpoint. Only three stud-
gestation in a large controlled trial in rural Nepal (779 ies found that supplemental zinc significantly increased
women in five groups). Serum α1-acid glycoprotein birthweight as compared with the control group [29, 30,
concentration decreased in all groups but decreased 41]. All three studies were done in populations where
more in the groups receiving folic acid, or folic acid maternal zinc depletion was likely. In the study com-
and iron with or without zinc. In contrast, C-reactive paring a zinc supplementation group with an untreated
protein increased in all groups from baseline to 32 control group in India [30], infants born to women in
weeks of gestation, but it was significantly lower in the the control group weighed only about 2.6 kg; those
group receiving folic acid and iron with zinc only [22]. born to zinc-supplemented mothers were about 0.3 to
This suggests that zinc given with folic acid and iron 0.8 kg heavier (p < .001), depending on the length of
may ameliorate the inflammatory process in pregnancy, time supplemental zinc was provided. Xie et al. [29]
which has positive implications for reproductive health compared three different levels of supplemental zinc
outcomes. (5, 10, and 30 mg/day) with placebo in rural Chinese
Serum zinc concentration declines progressively women and found that the infants of mothers receiv-
during the course of pregnancy in relation to blood- ing the highest amount of zinc supplement (30 mg/
volume expansion [25]. Thus, serum zinc concentration day) were on average 283 g heavier (p = .016) and had
values must be interpreted in relation to the stage of larger head circumferences (0.6 cm, p = .035) than the
pregnancy or serum albumin concentrations. Unlike infants of mothers in the placebo group. This difference
studies in young children, in whom serum zinc con- was not found in mothers receiving 5 or 10 mg of zinc/
centration nearly always increases in response to zinc day. Goldenberg et al. [41] studied the effect of zinc
supplementation [26], only 5 of the 12 zinc supplemen- supplementation on birthweight in a group of African-
tation studies that reported serum zinc concentrations American women who were medically indigent (with-
during pregnancy found a significant increase in the out health insurance or other health-care coverage).
supplemented group [18, 27–30]. The lack of response Only women with plasma zinc concentrations below
may have been due to failure to control for the stage the median for their population at 20 weeks of gestation
TABLE 2. Effect of zinc supplementation during pregnancy on fetal mortality and growth
Country, year Amount of zinc Duration of
[reference] author Study design Population supplement supplementation Zinc group Comparison group Effect of supplement
Supplementation with zinc alone Zinc Placebo
Bangladesh, 2000 Randomized, Poor, urban Bang- 30 mg/day From wk 12–16 n = 269 n = 290 No effect
[32] Osendarp double-blind, ladeshi women gestation to term
controlled
Chile, 2001 Randomized, Chilean adolescent 20 mg/day From before wk 20 n = 249 n = 258 Significantly lower propor-
[15] Castillo- double-blind, women gestation to term tion of LBW and premature
Duran controlled infants
China, 2001 Randomized, Rural Chinese 5 mg/day From before wk 12 5 mg zinc/day n = 40 Increased birthweight and
[29] Xie double-blind, women 10 mg/day gestation to term (n = 37) head circumference in
controlled 30 mg/day 10 mg zinc/day high-zinc group compared
(n = 40) with placebo group
30 mg zinc/day
(n = 39)
South Africa, 1979 Randomized, Black South Afri- 30 or 90 mg/day From midpreg- n = 32 n = 33 No effect
[31] Ross double-blind, can women nancy to term
controlled
UK, 1989 Randomized, UK women 20 mg/day From before wk 20 n = 246 n = 248 No effect
[13] Mahomed double-blind, gestation to term
controlled
Folic acid + iron
Supplementation with folic acid, iron, and zinc + zinc Folic acid + iron
India, 1993 Randomized, Urban Indian 45 mg/day From booking to n = 106 n = 62 Increased birthweight and
[30] Garg not blinded women term gestational age; fewer pre-
No placebo term infants
Indonesia, 2001 Randomized, Indonesian rural 30 mg/day From before wk 20 Folic acid + iron + Folic acid + iron No effect on birthweight
[17] Dijkhuizen double-blind, women gestation to term zinc (n = 48) (n = 42)
controlled
β-Carotene + folic β-Carotene + No effect on birthweight
acid + iron + zinc folic acid + iron
(n = 44) (n = 45)
Nepal, 2003 Cluster-rand- Nepalese rural 30 mg/day From wk 5–10 Folic acid + iron + Placebo, folic No beneficial effect of zinc
[42] Christian omized, double- women gestation to 3 mo zinc (n = 827) acid, folic acid + on fetal loss compared with
blind, controlled postpartum iron, or MMN other groups
(n = 3,295)
continued
S65
TABLE 2. Effect of zinc supplementation during pregnancy on fetal mortality and growth (continued) S66
Country, year Amount of zinc Duration of
[reference] author Study design Population supplement supplementation Zinc group Comparison group Effect of supplement
[43] Katz Folic acid + iron + Placebo, folic Zinc with folic acid + iron
zinc (n = 827) acid, folic acid + increased birthweight of
iron, or MMN infants in the 2,400–2,900
(n = 3,295) g range
[23] Christian Folic acid + iron + Placebo, folic Zinc with folic acid +
zinc (n = 827) acid, folic acid + iron had no effect on
iron, or MMN birthweight
(n = 3,295)
Pakistan, 2005 Randomized, Pakistani women 20 mg/day From wk 10–16 n = 121 n = 121 No effect on birthweight or
[38] Hafeez double-blind gestation to term on length or duration of
controlled pregnancy
Peru, 1999 Randomized, Poor Peruvian 15 mg/day From wk 10–24 n = 521 n = 495 No effect on birthweight or
[34] Caulfield double-blind women gestation to term on length or duration of
controlled pregnancy
[35] Merialdi n = 31 n = 24 Zinc-containing supple-
ment increased fetal heart
rate range and in utero
movement
Peru, 2004 Randomized, Poor Peruvian 25 mg/day From wk 10–16 n = 94 n = 101 Zinc-containing supplement
[36] Merialdi double-blind women gestation to term increased fetal femur dia-
controlled physis length
[37] Merialdi n = 94 n = 101 Zinc-containing supplement
increased fetal heart rate
variability
UK, 1991aa Randomized, Lower-social-class 22.5 mg/day From 15–25 wk n = 30 n = 22 Reduced incidence of IUGR
[16] Simmer double-blind English women prior to delivery
controlled
UK, 1991b Randomized, High-risk English 62 mg/day Before wk 18 ges- n = 72 n = 62 No effect on birthweight
[33] Robertson double-blind women tation to term
controlled
Supplementation with zinc and MMN MMN + zinc MMN
Tanzania, 2005 Randomized, HIV-infected preg- 25 mg/day From wk 12–27 n = 198 n = 199 No effect on birthweight,
[44] Fawzi double-blind, nant Tanzanian gestation to term duration of pregnancy, or
controlled women fetal mortality
S. Y. Hess and J. C. King
were included in the study. Thus, the intervention was
Zinc supplementation associ-
greater in women with BMI

head circumference; effect
ated with lower incidence
No effect on birthweight or
targeted toward women at risk for being zinc deficient.
term delivery in normal-

Reduced incidence of pre-
Increased birthweight and

other measures of preg-
of pregnancy-induced
In all women, supplemental zinc increased the infant’s
birthweight by 128 g (p = .03) and head circumference
by 0.4 cm (p = .02). The effect was greater in nonobese
nancy outcome
weight women
hypertension
women, among whom zinc supplementation increased
birthweight by 248 g (p = .005) and head circumference
by 0.7 cm (p = .007). Eleven other studies failed to find
< 26
a relationship between zinc supplementation and birth-
weight [13, 15, 17, 19, 31–34, 38, 39, 44]. However, none
of those 11 studies stratified the effects of supplemental
zinc on birthweight by maternal weight or zinc status.
n = 106
n = 288
n = 286
There was no evidence that whether zinc was given
n = 36
along with other micronutrients or alone influenced

the outcome. The recently published meta-analysis
of 14 studies found that zinc supplementation had no
significant impact on birthweight (WMD, −10.59; 95%
CI, −36.71 to 15.54) [10]. Future studies are needed
to determine if maternal pregravid or gravid weight
n = 107
n = 268
n = 294
n = 10
modifies the effect of supplemental zinc on fetal growth

and birthweight.
Simmer et al. [16] conducted a double-blind trial in
the United Kingdom of mothers at risk for delivering
infants with intrauterine growth retardation because
BMI, body mass index; IUGR, intrauterine growth retardation; LBW, low-birthweight; MMN, multiple micronutrients
From before wk 25
gestation to term
gestation to term
From wk 19 gesta-
From about wk 19
From mo 1–3 ges-
they had had a small-for-gestational age infant previ-

tation to term
ously, were underweight, or were smoking. Zinc sup-

tion to term
plementation significantly reduced the incidence of

intrauterine growth retardation (7% vs. 27%, p = .04).
There was no significant effect on overall birthweight.
Since this is a small study with insufficient power to
allow definite conclusions to be drawn, additional
research is needed to evaluate whether zinc supple-
mentation is more likely to improve fetal growth in
high-risk, underweight mothers.
15 mg/day
20 mg/day
30 mg/day
Medically indigent 25 mg/day
Merialdi et al. [36] used ultrasonography to exam-

ine the effect of administration of supplemental zinc
to pregnant women on fetal bone growth in utero.
a. Iron and folic acid supplement was prescribed if clinically indicated.
Although there were no effects of zinc supplementation

Colorado women
Low-income, His-
panic California
African-Ameri-
primarily black
(given along with iron and folic acid) on birthweight

US adolescents,
Middle-income
can women
or head circumference, the femur diaphysis length was

significantly longer in the fetuses of mothers receiving
women
zinc along with iron and folic acid than in fetuses of

mothers receiving iron and folic acid only at all time
points measured between 20 and 38 weeks of gestation.
No differences in birth length due to zinc supplemen-
No statement on
randomization
double-blind,
tation were observed, but the small effects on femur

double-blind
double-blind
and blinding
Randomized,
Randomized,
Randomized,
length probably would not be detected in measure-

controlled
controlled
controlled
Controlled
ments of birth length. There were no other effects of

supplemental zinc on growth at specific anatomical
sites. These findings are consistent with studies in
experimental animals suggesting that zinc has a very
specific effect on the growth of long bones [45, 46].
[41] Goldenberg
[39] Hambidge
Preterm birth can contribute to a reduction in birth-

[40] Cherry
weight. Zinc supplementation significantly reduced

USA, 1983
USA, 1984
USA, 1989
USA, 1995
[19] Hunt
the prematurity rate compared with placebo among

Chilean adolescents (6% vs. 12%, p = .016) [15]. Among
Indian women, there were significantly fewer preterm
infants in the zinc-supplemented group than in a with normal birthweight. The results from the Peru-
nonintervention control group who did not receive a vian maternal zinc supplementation trial, which were
placebo (2% vs. 11%, p < .05) [30]. Although most other published in a review article [9], also show that mater-
studies found no significant overall impact on preterm nal zinc supplementation reduced acute diarrhea and
delivery, some studies found a beneficial impact in dysentery, but the results were only significant for the
selected subgroups of women. In a meta-analysis of 13 period from 8 to 12 months of age. In a study in rural
studies, Mahomed et al. [10] found a small, but signifi- Nepal, the women received supplements from early
cant, reduction in preterm birth (relative risk, 0.86; 95% pregnancy to 3 months postpartum that contained
CI, 0.76 to 0.98). This 14% reduction in preterm birth vitamin A alone; vitamin A and folic acid; vitamin A,
among zinc-supplemented women was found primarily folic acid, and iron; vitamin A, folic acid, iron, and zinc;
in the subset of studies involving lower-income women, or a multiple micronutrients (MMN) supplement [48].
suggesting that zinc supplementation may be beneficial None of the combinations of antenatal micronutrient
in areas where poor nutrition and maternal infection supplements affected symptoms of neonatal morbidity
are more common. in the first 10 days of life or at 6 weeks of age [48], and
Animal studies show that zinc is essential for neu- there was no evidence that zinc had any effect on infant
rologic development in utero. In two studies in Peru, mortality throughout the first year [42]. These findings
Merialdi et al. [35, 37] evaluated the effect of supple- suggest that maternal zinc supplementation during
mental zinc on fetal heart rate and fetal movement, pregnancy may influence the infant’s risk of selected
using novel methods for measuring fetal cardiac and infections postnatally, but that these benefits may be
somatic activity, both of which are influenced by the restricted to older infants. More research is needed to
development of the autonomic nervous system. Fetuses address these issues.
of mothers who received supplemental zinc plus iron In two studies, no differences in infant growth were
and folic acid showed an increased range and variabil- observed in relation to maternal zinc supplementation
ity of fetal heart rate and an increased amount of time among Indonesian infants at 6 months postpartum
spent moving compared with those whose mothers [51] and among Bangladeshi children at 13 months
received iron and folic acid supplements without zinc. postpartum [47]. In contrast, infants born to Peruvian
The differences became significant at 36 weeks of gesta- mothers supplemented with zinc during pregnancy had
tion (p < .05). Although follow-up studies are needed to significantly greater anthropometric measures from
verify these observations and assess their longer-term months 4 to 12 [50]. On average, the infants from the
implications, the data suggest that adding zinc to pre- zinc group were 0.58 ± 0.12 kg/month heavier, with
natal iron and folic acid supplements may have a ben- weight accrued during the first year of life, than those
eficial effect on fetal neurobehavioral development. from the control group. The longitudinal effects of zinc
In summary, the effects of supplemental zinc on treatment remained significant for weight, calf and
maternal labor and delivery complications and on fetal chest circumferences, and calf muscle area after con-
growth and neurobehavioral development are limited trol for a range of covariates, including infant-feeding
and conflicting. There is some evidence that sup- practices and diarrhea morbidity. The reason for these
plemental zinc increases birthweight and duration of differences among studies is not known, and additional
gestation in underweight women living in areas where research will be needed to understand the effects.
zinc intake is low or zinc is poorly absorbed. However, Neurobehavioral development was evaluated in
a number of other studies failed to find a relationship infants born to zinc-supplemented Bangladeshi [49]
between supplemental zinc, with or without other and Peruvian women [9]. In Peru, some improve-
micronutrient supplements, and birthweight. The ments in infant neurobehavioral development (novelty
results of those studies were not stratified by maternal preference) were observed in the zinc-supplemented
body weight. group at 6 months of age, whereas no benefit was
observed in Bangladesh. In fact, Bangladeshi infants
Neonatal and early infant morbidity and mortality, growth, in the placebo group had higher scores on mental
and development development and psychomotor indexes than those
The effects of zinc supplementation on postnatal in the zinc-supplemented group. Tamura et al. [52]
growth, development, and rates of infections are sum- evaluated the effect of prenatal zinc supplementation
marized in table 3 [9, 42, 47–52]. Osendarp et al. [47] on the mental and psychomotor development of 355
evaluated the effects of zinc supplementation on infant children of African-American mothers who partici-
growth and morbidity at 6 months of age in Bangladesh. pated in a double-blind study of zinc supplementation
Maternal supplementation during pregnancy reduced in which the infants of zinc-supplemented mothers had
the risk of infant acute diarrhea (risk ratio, 0.84; 95% increased head circumference at birth. There was no
CI, 0.72 to 0.98), dysentery (risk ratio, 0.36; 95% CI, effect on mental or psychomotor development of the
0.25 to 084), and impetigo (risk ratio, 0.53; 95% CI, 0.34 children at 5 years of age, before or after stratification of
to 0.82) in low-birthweight infants but not in infants the sample by maternal body mass index. Identification
TABLE 3. Effect of zinc supplementation during pregnancy on infant and young child postnatal growth, morbidity, and neurobehavioral development
Amount
of zinc Duration of Comparison
Study Study design Population supplement supplementation Zinc group group Effect of supplement
Bangladesh, 2000 Randomized, Infants of poor, 30 mg/day From wk 12–16 gesta- n = 184 n = 199 No effect on growth at 6 mo postpar-
[47] Osendarp double-blind, urban Bangla- tion to term tum. Reduced incidence of acute
controlled deshi women diarrhea and dysentery
[49] Hamadani n = 83 n = 85 Placebo improved mental and psycho-
motor development. No effect of zinc
on behavior or growth
Folic acid + Folic acid +
Supplementation with folic acid, iron, and zinc, with or without β-carotene iron + zinc iron
Indonesia, 2001 Randomized, Infants of Indo- 30 mg/day From before wk 20 Folic acid + Folic acid + No effect on growth at 6 mo
[51] Dijkhuizen double-blind, nesian rural gestation to term iron + zinc iron (n = 42) postpartum
controlled women (n = 48)

β-Carotene + β-Carotene No effect on growth at 6 mo
folic acid + iron + folic acid + postpartum
+ zinc (n = 44) iron (n = 45)
Nepal, 2003 Cluster-rand- Nepalese rural 30 mg/day From wk 5–10 ges- Folic acid + Placebo, folic No beneficial effect of zinc on neona-
[42, 48] Christian omized, double- women tation to 3 mo iron + zinc acid, folic tal and infant morbidity and infant
blind, controlled postpartum (n = 827) acid + iron, mortality
or MMN
(n = 3,295)
Peru, 1999 Randomized, Infants of poor 15 mg/day From wk 10–24 gesta- n = 273 n = 273 Larger growth measures beginning at 4
[50] Iannotti double-blind, Peruvian tion to term mo through 12 mo
controlled women
[9] Osendarp n = 521 n = 495 Reduced risk of any diarrhea at 8–12
mo and of dysentery at 0–12 mo
[9] Osendarp NA NA Some improvements in neurobehavio-
ral development (novelty preference)
at 6 mo
Zinc and MMN supplementation MMN + zinc MMN
USA, 1995 Randomized, Infants of medi- 25 mg/day From wk 19 gestation n = 173 n = 182 No effect on mental and psychomotor
[52] Tamura double-blind, cally indigent to term development of children at 5 yr of age
controlled African-
American
women
S69
MMN, multiple micronutrients; NA, not available

of long-term effects of prenatal supplemental zinc on occurred without any increase in maternal plasma zinc
neurologic development may be undermined by poor levels. The other four studies gave supplemental zinc
environmental postnatal influences on this aspect of to lactating women along with other micronutrients.
development. Karra et al. [54] provided 25 mg of supplemental zinc
per day to lactating women for 6 months and found a
significantly lower decline in breastmilk zinc concen-
Section 2 tration in the supplemented group than in the placebo
group, accompanied by a significant increase in plasma
What is the effect of preventive zinc supplementation zinc concentration in the zinc-supplemented mothers.
during lactation on maternal and neonatal health? Salmenperä et al. [56] found no difference in breastmilk
zinc concentrations between the control group and the
Conclusions supplemented group receiving 20 mg of zinc per day.
However, breastmilk zinc concentration in the group
The limited data available on zinc supplementation receiving 40 mg per day declined significantly more
and lactation performance and infant growth and slowly—by 6 months—than in the other two groups.
zinc status fail to show any consistent benefit to the There was no correlation between maternal serum zinc
mother or child. However, studies are available only concentration and the total zinc transfer into milk in
from relatively healthy women in developed countries. any of the three groups (40, 20, and 0 mg/day) [56].
The effect of maternal zinc supplementation during Krebs et al. [57] and Moser-Veillon and Reynolds [55]
pregnancy and lactation or during lactation only on did not find an effect of supplemental zinc on milk zinc
maternal and neonatal health needs to be studied in concentration. These inconsistent findings suggest that
undernourished women. neither maternal zinc intake nor plasma zinc levels are
major determinants of milk zinc concentration.
Detailed review of evidence Infant growth and zinc status
Bibliographic search Only one study is available regarding the effect of
Data sets were identified as described above. Of the maternal zinc supplementation during lactation on
1,618 articles, there were 3 articles that evaluated the infant growth. Salmenperä et al. [56] found no effect
impact of zinc supplementation during lactation. Two of maternal zinc supplementation (20 or 40 mg/day vs.
additional articles not identified during the PubMed placebo) during lactation on serum zinc concentra-
search were added. The findings of the five articles are tions or growth of Finnish infants. Infant serum zinc
summarized in the following section (table 4). concentrations throughout the first year of life (mean
All five studies of the effect of maternal zinc sup- ± SD, 67 ± 4 μg/dL) tended to decline during periods
plementation on milk zinc concentration and infant of rapid growth, especially in boys. There were no
growth [53–57] were done in women living in devel- associations between serum zinc concentrations and
oped countries (Finland and the United States). Zinc growth rates.
supplementation was initiated at birth and continued
daily throughout the duration of the studies, which
varied from 6 to 12 months. It is not clear whether the Section 3
mothers were exclusively breastfeeding for the entire
study period or whether breastfeeding was supple- Are there any adverse effects of zinc supplementation
mented with formula and complementary foods during during pregnancy or lactation?
the later stages of lactation.
Milk volume and zinc concentration Conclusions
Krebs et al. [53] found that the rate of decline in milk There is very little information available on adverse
zinc concentrations was significantly less in mothers effects of zinc supplementation during pregnancy. We
receiving supplemental zinc without any other micro- reviewed the effects of providing zinc in a prenatal
nutrients than in non-zinc-supplemented mothers supplement on iron status in nine randomized, double-
(p = .02). Log-transformed monthly milk zinc con- blind, controlled trials and on copper status in four
centrations were compared between the two groups trials. None of these studies found a difference in final
to determine the mean differences in rate of decline hemoglobin, serum ferritin, transferrin receptor, or
between birth and 9 months of age. Milk zinc concen- serum copper concentration between the zinc-supple-
tration decreased by 0.69 ± 0.27 µg/mL in the non- mented group and the control group. The four studies
zinc-supplemented group, whereas it declined by 0.54 ± reporting on copper status provided information on
0.14 µg/mL in the supplemented group. These changes serum copper concentration only, which is a relatively
in milk zinc concentration with zinc supplementation insensitive biomarker of copper status. Further studies
TABLE 4. Effect of maternal zinc supplementation during lactation on breastmilk zinc concentration
Amount
Country, year Age of of zinc Duration of Zinc Placebo
[reference] author Study design Population infants supplement supplementation group group Effect of supplement
USA, 1985 Untreated control Mothers in Colo- 1–12 mo 15 mg/day From 1 to 12 n = 14a n = 25a Rate of decline in milk zinc significantly
[53] Krebs group. No statement rado, USA mo less than in nonsupplemented group
on randomization
and blinding
MMN +
Supplementation with zinc and MMN zinc MMN
Finland, 1994 Controlled. No state- Finnish moth- 0–12 mo 20 mg/day From days 4–5 n= 73b n = 94b No effect on infant serum zinc concen-
[56] Salmenperä ment on randomiza- ers exclusively after delivery tration or growth
tion and blinding breastfeeding to 12 mo No difference in milk zinc between con-
trols and group receiving 20 mg zinc/

day
40 mg/day n = 27b Smaller decline in milk zinc in group
receiving 40 mg zinc/day by 6 mo
USA, 1988 Controlled. No state- Mothers in Indi- 0–6 mo 25 mg/day From day 1 after n = 24d n = 25d Rate of decline in milk zinc significantly
[54] Karra ment on randomiza- ana, USAc delivery to 6 less than in nonsupplemented group
tion and blinding mo
USA, 1990 Randomized, double- Mothers in Mar- 0–9 mo 25 mg/day From day 1 after n = 20e n = 20e No effect on infant plasma zinc,
[55] Moser- blind, controlled yland, USA delivery to 9 infant erythrocyte zinc, or milk zinc
Veillon mo concentration
USA 1995, Randomized, double- Mothers in Colo- 1–7 mo 15 mg/day From wk 2 n = 40f n = 31f No effect on milk zinc concentration
[57] Krebs blind, controlled rado, USA postpartum to
7 mo
MMN, multiple micronutrients
a. Sample size from enrollment to 12 months decreased from 14 to 2 in the zinc group and from 25 to 2 in the placebo group.
b. Sample size from enrollment to 12 months decreased from 73 to 2 in the group receiving 20 mg of zinc and from 94 to 5 in the placebo group. By 9 months, the group receiving 40 mg of zinc was reduced
to 4.
c. Study included a comparison group of lactating women in Egypt. Data not presented here.
d. Sample size from enrollment to 12 months decreased from 24 to 19 in the zinc group and from 25 to 22 in the placebo group.
e. Sample size from enrollment to 9 months decreased from 20 to 9 in the zinc group and from 20 to 11 in the placebo group.
f. The number of patients who completed the study is given.
S71
using more sensitive indicators of copper status are the addition of zinc to the prenatal supplement had no
required. On the basis of available results, it can be impact on hemoglobin or plasma ferritin concentration
concluded that the addition of zinc to a prenatal iron in mothers and infants. On the basis of these results, it
and folic acid supplement does not adversely affect can be concluded that the addition of zinc to a prenatal
iron status. iron and folic acid or a MMN supplement in the range
of 15 to 30 mg of zinc/day does not adversely affect
Detailed review of evidence iron status.
The studies reporting on copper status provide infor-
Any adverse effects related to the addition of zinc to mation on serum copper concentration only, which is
the prenatal iron and folic acid supplement would be the least sensitive biomarker of copper status [68]. Of
of concern. It is important, therefore, to consider the the three studies reporting on copper status (table 5),
risks, as well as the benefits, of zinc supplementation none of the studies found a significant difference in
during pregnancy and lactation. However, manifesta- final serum copper concentration between the two
tions of acute toxicity symptoms, such as nausea and treatment groups receiving MMN with or without zinc.
vomiting, occur only at very high zinc intake levels of Similarly, Hambidge et al. [39] stated that there was not
~225 to 450 mg/day or more [58]. An adverse effect a difference between groups at 10 months of gestation,
of zinc on copper metabolism has been shown only at but they did not provide the data disaggregated by
zinc intakes above 50 mg/day in adults, as measured treatment group. This limited information, as judged
by a decrease in erythrocyte superoxide dismutase by serum copper levels, indicates that the provision of
activity [59]. There is some evidence from interven- supplemental zinc at a level of 20 to 30 mg/day during
tion trials in young children that iron indicators do pregnancy does not have a negative effect on copper
not improve as much when supplemental iron is given status. Further studies using more sensitive indicators
with zinc as when iron is given alone [60], although no of copper status are needed.
overall effect was found in a recent meta-analysis [61]. None of the zinc supplementation trials in lactating
A recent animal study showed that supplementation women reported results on iron or copper status in
of zinc-adequate dams with additional zinc induced a the mothers. However, on the basis of the lack of any
immunosuppressive response in the offspring, a find- adverse effects on iron and copper status in pregnant
ing that needs further investigation but implies a need women and in children [61], it can be assumed that
for caution [62]. providing a zinc supplement during lactation at the
The impact of adding zinc to prenatal supplements recommended dosage does not have an adverse effect
on iron and copper nutriture was reviewed. A total on maternal iron or copper status.
of nine randomized, double-blind controlled trials in
pregnant women [13, 17, 19, 32, 44, 63–67] reporting
results on either hemoglobin concentration or other Section 4
iron status indicators are summarized in table 5. One
study was not considered in this analysis because the What are the implications of these outcomes for zinc sup-
supplementation period was only 3 weeks [18] and, plementation programs during pregnancy and lactation
therefore, was not likely to have an impact on hemo- and what are the remaining research needs?
globin concentration. For all other studies, we present
the data of the two treatment groups differing by zinc Adding zinc to prenatal supplements
only, if the studies included more than two groups.
Two studies [13, 32] compared zinc versus placebo, As mentioned above, WHO recommends that all
four studies compared the addition of zinc to an iron pregnant women living in areas of high prevalence of
[63] or an iron and folic acid supplement [17, 65] with malnutrition should routinely receive iron and folic
vitamin A or β-carotene [17, 66], and four studies pro- acid supplements to prevent anemia as part of the inte-
vided a MMN supplement containing iron and folic grated management of pregnancy and childbirth [8].
acid, with or without zinc [19, 44, 64, 67]. One study The purpose of this review was to evaluate the potential
found that the addition of zinc to an iron supplement benefits and adverse effects of adding zinc to the pre-
had a beneficial impact on hemoglobin concentration natal supplement containing iron and folic acid and to
in anemic pregnant Iranian women after 12 weeks of draw conclusions for programmatic implications.
supplementation [63]. None of the other studies found There is currently no evidence of adverse effects
a difference in final hemoglobin, serum ferritin, or of supplemental zinc on iron status or the response
transferrin receptor concentration between the treat- to iron supplementation during pregnancy. A meta-
ment group receiving zinc and those receiving no zinc analysis showed that zinc supplementation during
(with or without other micronutrients). Dijkhuizen pregnancy reduced the risk of preterm birth by 14%
et al. [51] also evaluated maternal and infant micro- [10]. However, the effects of supplemental zinc on labor
nutrient status 6 months postpartum and found that and delivery complications, birthweight, and postnatal
TABLE 5. Possible adverse effects of zinc supplementation during pregnancy on hemoglobin and iron and copper status indicators
Time of Initial hemo- Final hemo- Initial serum Final serum Initial serum Final serum
Country, year Time of gestation at Micronutri- globin con- globin con- ferritin con- ferritin con- copper con- copper con-
[reference] gestation at final blood ent content of Sample centration centration centration centration centration centration
author enrollment collection supplement Study group size a (g/L)b (g/L)b (μg/L)c (μg/L)c (μg/dL)b (μg/dL)b,d
Supplementation with zinc alone
UK, 1989 Before wk 20 Wk 28–32 — Control 232 — 115 — — — —
[13] Mahomed (NS)e
20 mg zinc Zinc 238 — 117 — — — —
Bangladesh, Wk 12–16 Wk 28–32 — Control 232 115 ± 11 108 ± 11 — — — —
2000 (NS)e
[32] Osendarp
30 mg zinc Zinc 214 114 ± 14 108 ± 13 — — — —
Supplementation with folic acid, iron, and zinc

Iran, 2005 Wk 16–20 Wk 28–32 100 mg iron Iron 58 104 ± 2 119 ± 9 — — — —
[63] (p < .05)
Mahmoudian
100 mg iron, Iron + zinc 60 105 ± 2 128 ± 10 — — — —
15 mg zinc
Peru, 1999 Wk 10–24 Wk 37–38 250 μg folic Folic acid + 320 115 ± 14 115 ± 13 19.7 17.8 — —
[65] Zavaleta acid, 60 mg iron (NS) (7.7, 50) (8.2, 38.9)
iron (NS)
250 μg folic Folic acid + 325 116 ± 12 114 ± 13 21.8 17.6 — —
acid, 60 mg iron + zinc (8.9, 53.3) (8.1, 38.5)
iron, 15 mg
zinc
Indonesia, 2001 Wk 10–20 Wk 32 400 μg folic Folic acid + 34 118 ± 14 109 ± 12 24 (8.7, 39.6) 14 (7.8, 22.8) — —
[17] Dijkhuizen acid, 30 mg iron (NS)f (NS)f
iron
400 μg folic Folic acid + 45 114 ± 11 106 ± 14 28 14 (7.0, 22.3) — —
acid, 30 mg iron + zinc (14.2, 46.7)
iron, 30 mg
zinc
continued
S73
TABLE 5. Possible adverse effects of zinc supplementation during pregnancy on hemoglobin and iron and copper status indicators (continued) S74
Time of Initial hemo- Final hemo- Initial serum Final serum Initial serum Final serum
Country, year Time of gestation at Micronutri- globin con- globin con- ferritin con- ferritin con- copper con- copper con-
[reference] gestation at final blood ent content of Sample centration centration centration centration centration centration
author enrollment collection supplement Study group size a (g/L)b (g/L)b (μg/L)c (μg/L)c (μg/dL)b (μg/dL)b,d
4.5 mg β-Carotene 42 118 ± 12 107 ± 12 25 16.2 — —
β-carotene, + folic acid (16.4, 39.6) (8.5, 24.4)
400 μg folic + iron
acid, 30 mg
iron
4.5 mg β-Carotene 43 116 ± 12 109 ± 11 21 16 — —
β-carotene, + folic acid (10.1, 43.2) (8.8, 25.3)
400 μg folic + iron +
acid, 30 mg zinc
iron, 30 mg
zinc
Nepal, 2003g Wk 10 Wk 32 Vitamin A Vitamin A + 202 115 ± 18 116 ± 15 11.5 (18.2) 14.4 (12.9) 148 ± 44 217 ± 48
[22, 66] (1,000 μg RE), folic acid + (NS) (NS) (NS)h
Christian 400 μg folic iron
acid, 60 mg
iron
Vitamin A Vitamin A + 253 116 ± 22 116 ± 15 13.7 (16.7) 13.7 (15.3) 151 ± 44 217 ± 43
(1,000 μg RE), folic acid +
400 μg folic iron + zinc
acid, 60 mg
iron, 30 mg
zinc
Supplementation with zinc and MMN
USA, 1984i Average wk Average wk 1 mg folic acid, MMN 37 119 ± 9 123 ± 9 (NS) — — 225 ± 40 236 ± 41
[19] Hunt 19 36 20 mg iron, (NS)
other MMN
1 mg folic acid, MMN + zinc 27 122 ± 10 123 ± 13 — — 225 ± 32 238 ± 34
20 mg iron,
other MMN,
20 mg zinc
USA, 1985 About wk 17 About wk 36 1 mg folic acid, MMN 26 123 ± 13 124 ± 11 29 ± 24.4 10.6 ± 5.7 — —
[67] Hunt 20 mg iron, (NS) (NS)
other MMN
S. Y. Hess and J. C. King
1 mg folic acid, MMN + zinc 36 119 ± 11 120 ± 11 32.9 ± 27.7 10.5 ± 7.5 — —
20 mg iron,
other MMN,
20 mg zinc
USA, 1995 Average wk Average Folic acid and MMN 32 — — — — 222 ± 38 230 ± 46
[64] Tamura 20 wk 40 other MMNj (NS)
(delivery)
Folic acid and MMN + zinc 31 — — — — 215 ± 34 225 ± 42
other MMN,j
25 mg zinc
Tanzania, 2005 Average wk Wk 6 5 mg folic acid, MMN 193 98 ± 14 114 ± 17 — — — —
[44] Fawzi 23 postpartum 120 mg iron, (NS)
other MMN
5 mg folic acid, MMN + zinc 192 100 ± 12 112 ± 16 — — — —
120 mg iron,
other MMN,
25 mg zinc
MMN, multiple micronutrients; NS, not significant; RE, retinol equivalents
a. Sample size for measurement of initial hemoglobin concentration. Sample size for other indicators may vary.
b. Results for hemoglobin and serum copper concentrations are reported as mean ± SD, except for UK 1989 [13], where SD was not reported.
c. Results for serum ferritin are reported as geometric mean (–1 SD, +1 SD) for Peru 1999 [65], as median (interquartile range) for Indonesia 2001 [17] and Nepal 2003 [22, 66], and as mean ± SD for USA
1985 [67].
d. The study by Hambidge et al. [39] is not shown in this table because the data on copper concentration are not disaggregated between the two treatment groups. However, the authors state that there was
no significant difference between final concentrations at 10 months of gestation.
e. No significant difference in final concentration between two treatment groups
f. No significant difference in final concentration among all four treatment groups.
g. Additional information provided by author (Christian P, personal communication, 2007).
h. P-value calculated on the basis of changes in copper concentration.
i. Results reported only for women who were studied long enough to take supplements for more than 60 days.
j. MMN content of supplement not defined.
S75
growth are inconsistent. birthweight. The association between maternal stat-

In view of the lack of a clear benefit of supplemental ure, zinc status, and birthweight also needs further
zinc during gestation, a recommendation to add zinc investigation.
to the usual prenatal supplement of iron and folic acid » Since pregnancy complications may be associated
cannot be made. Nevertheless, since toxic effects of with placentation problems in early gestation, zinc
supplemental zinc have not been identified, it may be supplementation prior to conception also should
prudent to include zinc in the prenatal supplement in be evaluated in relation to maternal morbidity
areas at high risk for zinc deficiency, as indicated by a outcomes.
stunting rate of more than 20% among children under » Future studies on zinc supplementation during
five [2] or a maternal body mass index under 18.5. pregnancy should include a follow-up during early
childhood to further evaluate the impact on growth,
Zinc supplementation during lactation development, and morbidity in infants of mothers
who have received zinc supplementation during
There is currently insufficient evidence for any benefit pregnancy.
in providing zinc supplements to lactating women. » It is further important to assess the risks as well as the
Further research is needed to assess the impact of zinc benefits of zinc supplementation during pregnancy
supplementation on the lactating mothers and their and lactation.
infants, in particular among undernourished women » The effect of maternal zinc supplementation during
living in lower-income countries. lactation on maternal and neonatal health needs to
be studied in undernourished women.
Research needs
» There is a need for further research on the addition Acknowledgments

of zinc to prenatal supplements of iron and folic acid
for undernourished or low-weight women in lower- We appreciate the contributions of Reina Engle-Stone,
income countries. Josh Jorgensen, and K. Ryan Wessells, who assisted with
» Future studies are needed to determine if maternal the bibliographic search and data extraction, and Janet
pregravid or gravid weight is a determinant of the M. Peerson, who provided statistical advice.
effect of supplemental zinc on fetal growth and
References
1. Caulfield LE, Zavaleta N, Shankar AH, Merialdi M. supplementation. Integrated management of pregnancy
Potential contribution of maternal zinc supplementation and childbirth (IMPAC). Geneva: WHO, 2006.
during pregnancy to maternal and child survival. Am J 9. Osendarp SJ, West CE, Black RE. The need for maternal
Clin Nutr 1998;68:499S–508S. zinc supplementation in developing countries: An unre-
2. de Benoist B, Darnton-Hill I, Davidsson L, Fontaine O, solved issue. J Nutr 2003;133:817S–27S.
Hotz C. Conclusions of the joint WHO/UNICEF/IAEA/ 10. Mahomed K, Bhutta Z, Middleton P. Zinc supplemen-
IZiNCG interagency meeting on zinc status indicators. tation for improving pregnancy and infant outcome.
Food Nutr Bull 2007;28:S480–S79. Cochrane Database Syst Rev 2007;CD000230.
3. King JC, Turnlund JR. Human zinc requirements. In: 11. Jameson S, Ursing I. Low serum zinc concentrations in
Mills CF, ed. Zinc in human biology. London: Springer- pregnancy, results of investigations and treatment. Acta
Verlag, 1989:335–50. Med Scand Suppl 1976;593:50–64.
4. Swanson CA, King JC. Zinc and pregnancy outcome. 12. Appelbaum PC, Ross SM, Dhupelia I, Naeye RL. The
Am J Clin Nutr 1987;46:763–71. effect of diet supplementation and addition of zinc
5. Brown KH, Engle-Stone R, Krebs NF, Peerson JM. in vitro on the growth-supporting property of amni-
Dietary intervention strategies to enhance zinc nutri- otic fluid in African women. Am J Obstet Gynecol
tion: Promotion and support of breastfeeding for infants 1979;135:82–4.
and young children. Food Nutr Bull 2009;S142–69. 13. Mahomed K, James DK, Golding J, McCabe R. Zinc
6. Fung EB, Ritchie LD, Woodhouse LR, Roehl R, supplementation during pregnancy: A double blind
King JC. Zinc absorption in women during pregnancy randomised controlled trial. BMJ 1989;299:826–30.
and lactation: A longitudinal study. Am J Clin Nutr 14. Jonsson B, Hauge B, Larsen MF, Hald F. Zinc sup-
1997;66:80–8. plementation during pregnancy: A double blind ran-
7. Donangelo CM, Zapata CL, Woodhouse LR, Shames DM, domised controlled trial. Acta Obstet Gynecol Scand
Mukherjea R, King JC. Zinc absorption and kinetics 1996;75:725–9.
during pregnancy and lactation in Brazilian women. 15. Castillo-Duran C, Marin VB, Alcaraz LS, Iturralde H, Ruz
Am J Clin Nutr 2005;82:118–24. MO. Controlled trial of zinc supplementation in Chilean
8. World Health Organization. Iron and folate pregnant adolescents. Nutr Res 2001;21:715–24.
16. Simmer K, Lort-Phillips L, James C, Thompson RP. A 30. Garg HK, Singhal KC, Arshad Z. A study of the effect
double-blind trial of zinc supplementation in pregnancy. of oral zinc supplementation during pregnancy on
Eur J Clin Nutr 1991;45:139–44. pregnancy outcome. Indian J Physiol Pharmacol
17. Dijkhuizen MA, Wieringa FT. Vitamin A, iron and zinc 1993;37:276–84.
in Indonesia: Micronutrient interactions and effects of 31. Ross SM, Nel E, Naeye RL. Differing effects of low and
supplementation. Doctoral dissertation, Wageningen high bulk maternal dietary supplements during preg-
Agriculture University, Wageningen, Netherlands, 2001. nancy. Early Hum Dev 1985;10:295–302.
18. Christian P, Khatry SK, Yamini S, Stallings R, LeClerq SC, 32. Osendarp SJ, van Raaij JM, Arifeen SE, Wahed M,
Shrestha SR, Pradhan EK, West KP Jr. Zinc supplemen- Baqui AH, Fuchs GJ. A randomized, placebo-controlled
tation might potentiate the effect of vitamin A in restor- trial of the effect of zinc supplementation during preg-
ing night vision in pregnant Nepalese women. Am J Clin nancy on pregnancy outcome in Bangladeshi urban
Nutr 2001;73:1045–51. poor. Am J Clin Nutr 2000;71:114–9.
19. Hunt IF, Murphy NJ, Cleaver AE, Faraji B, Swendseid ME, 33. Robertson JS, Heywood B, Atkinson SM. Zinc sup-
Coulson AH, Clark VA, Browdy BL, Cabalum T, Smith plementation during pregnancy. J Public Health Med
JC Jr. Zinc supplementation during pregnancy: Effects 1991;13:227–9.
on selected blood constituents and on progress and out- 34. Caulfield LE, Zavaleta N, Figueroa A, Leon Z. Maternal
come of pregnancy in low-income women of Mexican zinc supplementation does not affect size at birth or
descent. Am J Clin Nutr 1984;40:508–21. pregnancy duration in Peru. J Nutr 1999;129:1563–8.
20. Hogg BB, Tamura T, Johnston KE, Dubard MB, Golden- 35. Merialdi M, Caulfield LE, Zavaleta N, Figueroa A, DiPi-
berg RL. Second-trimester plasma homocysteine levels etro JA. Adding zinc to prenatal iron and folate tablets
and pregnancy-induced hypertension, preeclampsia, improves fetal neurobehavioral development. Am J
and intrauterine growth restriction. Am J Obstet Gyne- Obstet Gynecol 1999;180:483–90.
col 2000;183:805–9. 36. Merialdi M, Caulfield LE, Zavaleta N, Figueroa A,
21. Villamor E, Msamanga G, Aboud S, Urassa W, Hunter DJ, Costigan KA, Dominici F, Dipietro JA. Randomized
Fawzi WW. Adverse perinatal outcomes of HIV-1- controlled trial of prenatal zinc supplementation and
infected women in relation to malaria parasitemia in fetal bone growth. Am J Clin Nutr 2004;79:826–30.
maternal and umbilical cord blood. Am J Trop Med Hyg 37. Merialdi M, Caulfield LE, Zavaleta N, Figueroa A,
2005;73:694–7. Dominici F, Dipietro JA. Randomized controlled
22. Christian P, Jiang T, Khatry SK, LeClerq SC, Shrestha SR, trial of prenatal zinc supplementation and the devel-
West KP Jr. Antenatal supplementation with micro- opment of fetal heart rate. Am J Obstet Gynecol
nutrients and biochemical indicators of status and 2004;190:1106–12.
subclinical infection in rural Nepal. Am J Clin Nutr 38. Hafeez A, Mehmood G, Mazhar F. Oral zinc supple-
2006;83:788–94. mentation in pregnant women and its effect on birth
23. Christian P, Khatry SK, Katz J, Pradhan EK, LeClerq SC, weight: A randomised controlled trial. Arch Dis Child
Shrestha SR, Adhikari RK, Sommer A, West KP Jr. Fetal Neonatal Ed 2005;90:F170–1.
Effects of alternative maternal micronutrient supple- 39. Hambidge KM, Krebs NF, Jacobs MA, Favier A,
ments on low birth weight in rural Nepal: Double blind Guyette L, Ikle DN. Zinc nutritional status during
randomised community trial. BMJ 2003;326:571–6. pregnancy: A longitudinal study. Am J Clin Nutr
24. Shankar AH, Prasad AS. Zinc and immune function: 1983;37:429–42.
The biological basis of altered resistance to infection. 40. Cherry FF, Sandstead HH, Rojas P, Johnson LK, Batson
Am J Clin Nutr 1998;68:447S–63S. HK, Wang XB. Adolescent pregnancy: Associations
25. Hotz C, Peerson JM, Brown KH. Suggested lower cutoffs among body weight, zinc nutriture, and pregnancy
of serum zinc concentrations for assessing zinc status: outcome. Am J Clin Nutr 1989;50:945–54.
Reanalysis of the second National Health and Nutrition 41. Goldenberg RL, Tamura T, Neggers Y, Copper RL,
Examination Survey data (1976–1980). Am J Clin Nutr Johnston KE, DuBard MB, Hauth JC. The effect of
2003;78:756–64. zinc supplementation on pregnancy outcome. JAMA
26. Hess SY, Peerson JM, King JC, Brown KH. Use of serum 1995;274:463–8.
zinc concentration as an indicator of population zinc 42. Christian P, West KP, Khatry SK, Leclerq SC, Pradhan EK,
status. Food Nutr Bull 2007;28:S403–29. Katz J, Shrestha SR, Sommer A. Effects of maternal
27. Caulfield LE, Zavaleta N, Figueroa A. Adding zinc to micronutrient supplementation on fetal loss and infant
prenatal iron and folate supplements improves maternal mortality: A cluster-randomized trial in Nepal. Am J
and neonatal zinc status in a Peruvian population. Am J Clin Nutr 2003;78:1194–202.
Clin Nutr 1999;69:1257–63. 43. Katz J, Christian P, Dominici F, Zeger SL. Treatment
28. Neggers YH, Goldenberg RL, Tamura T, Johnston KE, effects of maternal micronutrient supplementation vary
Copper RL, DuBard M. Plasma and erythrocyte zinc by percentiles of the birth weight distribution in rural
concentrations and their relationship to dietary zinc Nepal. J Nutr 2006;136:1389–94.
intake and zinc supplementation during pregnancy in 44. Fawzi WW, Villamor E, Msamanga GI, Antelman G,
low-income African-American women. J Am Diet Assoc Aboud S, Urassa W, Hunter D. Trial of zinc supplements
1997;97:1269–74. in relation to pregnancy outcomes, hematologic indica-
29. Xie LM, Chen X, Pan J. The effects of zinc supplementa- tors, and T cell counts among HIV-1-infected women in
tion to Chinese rural pregnant women and their preg- Tanzania. Am J Clin Nutr 2005;81:161–7.
nancy outcome. Journal of Shanghai Second Medical 45. Yamaguchi M, Yamaguchi R. Action of zinc on bone
University 2001;13:199–24. metabolism in rats. Increases in alkaline phosphatase
activity and DNA content. Biochem Pharmacol 57. Krebs NF, Reidinger CJ, Hartley S, Robertson AD,
1986;35:773–7. Hambidge KM. Zinc supplementation during lactation:
46. Golub MS, Gershwin ME, Hurley LS, Saito WY, Hen- Effects on maternal status and milk zinc concentrations.
drickx AG. Studies of marginal zinc deprivation in Am J Clin Nutr 1995;61:1030–6.
rhesus monkeys. IV. Growth of infants in the first year. 58. Fosmire GJ. Zinc toxicity. Am J Clin Nutr 1990;51:
Am J Clin Nutr 1984;40:1192–202. 225–7.
47. Osendarp SJ, van Raaij JM, Darmstadt GL, Baqui AH, 59. International Zinc Nutrition Consultative Group
Hautvast JG, Fuchs GJ. Zinc supplementation during (IZiNCG), Brown KH, Rivera JA, Bhutta Z, Gibson
pregnancy and effects on growth and morbidity in low RS, King JC, Lönnerdal B, Ruel MT, Sandtröm B,
birthweight infants: A randomised placebo controlled Wasantwisut E, Hotz C. Assessment of the risk of zinc
trial. Lancet 2001;357:1080–5. deficiency in populations and options for its control.
48. Christian P, Darmstadt GL, Wu L, Khatry SK, Leclerq SC, Food Nutr Bull 2004;25(1 suppl 2):S99–203.
Katz J, West KP Jr, Adhikari RK. The effect of maternal 60. Fischer Walker C, Kordas K, Stoltzfus RJ, Black RE.
micronutrient supplementation on early neonatal Interactive effects of iron and zinc on biochemical and
morbidity in rural Nepal: A randomised, controlled, functional outcomes in supplementation trials. Am J
community trial. Arch Dis Child 2008;93:660–4. Clin Nutr 2005;82:5–12.
49. Hamadani JD, Fuchs GJ, Osendarp SJ, Huda SN, 61. Brown KH, Peerson JM, Baker S, Hess SY. Preventive zinc
Grantham-McGregor SM. Zinc supplementation during supplementation among infants, preschoolers, and older
pregnancy and effects on mental development and prepubertal children. Food Nutr Bull 2009;30:S12–40.
behaviour of infants: A follow-up study. Lancet 2002; 62. Raqib R, Hossain MB, Kelleher SL, Stephensen CB, Lön-
360:290–4. nerdal B. Zinc supplementation of pregnant rats with
50. Iannotti LL, Zavaleta N, Leon Z, Shankar AH, adequate zinc nutriture suppresses immune functions
Caulfield LE. Maternal zinc supplementation and growth in their offspring. J Nutr 2007;137:1037–42.
in Peruvian infants. Am J Clin Nutr 2008;88:154–60. 63. Mahmoudian A, Khademloo M. The effect of simulta-
51. Dijkhuizen MA, Wieringa FT, West CE, Muhilal. Zinc neous administration of zinc sulfate and ferrous sulfate
plus beta-carotene supplementation of pregnant women in the treatment of anemic pregnant women. Journal of
is superior to beta-carotene supplementation alone in Research in Medical Sciences 2005;10:205–9.
improving vitamin A status in both mothers and infants. 64. Tamura T, Olin KL, Goldenberg RL, Johnston KE,
Am J Clin Nutr 2004;80:1299–307. Dubard MB, Keen CL. Plasma extracellular superoxide
52. Tamura T, Goldenberg RL, Ramey SL, Nelson KG, dismutase activity in healthy pregnant women is not
Chapman VR. Effect of zinc supplementation of preg- influenced by zinc supplementation. Biol Trace Elem
nant women on the mental and psychomotor develop- Res 2001;80:107–13.
ment of their children at 5 y of age. Am J Clin Nutr 65. Zavaleta N, Caulfield LE, Garcia T. Changes in iron
2003;77:1512–6. status during pregnancy in Peruvian women receiving
53. Krebs NF, Hambidge KM, Jacobs MA, Rasbach JO. The prenatal iron and folic acid supplements with or without
effects of a dietary zinc supplement during lactation on zinc. Am J Clin Nutr 2000;71:956–61.
longitudinal changes in maternal zinc status and milk 66. Christian P, Shrestha J, LeClerq SC, Khatry SK, Jiang T,
zinc concentrations. Am J Clin Nutr 1985;41:560–70. Wagner T, Katz J, West KP Jr. Supplementation with
54. Karra MV, Kirksey A, Galal O, Bassily NS, Harrison GG, micronutrients in addition to iron and folic acid does
Jerome NW. Zinc, calcium, and magnesium concentra- not further improve the hematologic status of pregnant
tions in milk from American and Egyptian women women in rural Nepal. J Nutr 2003;133:3492–8.
throughout the first 6 months of lactation. Am J Clin 67. Hunt IF, Murphy NJ, Cleaver AE, Faraji B, Swendseid ME,
Nutr 1988;47:642–8. Browdy BL, Coulson AH, Clark VA, Settlage RH, Smith
55. Moser-Veillon PB, Reynolds RD. A longitudinal study JC Jr. Zinc supplementation during pregnancy in
of pyridoxine and zinc supplementation of lactating low-income teenagers of Mexican descent: Effects on
women. Am J Clin Nutr 1990;52:135–41. selected blood constituents and on progress and out-
56. Salmenperä L, Perheentupa J, Nanto V, Siimes MA. come of pregnancy. Am J Clin Nutr 1985;42:815–28.
Low zinc intake during exclusive breast-feeding does 68. Milne DB. Copper intake and assessment of copper
not impair growth. J Pediatr Gastroenterol Nutr status. Am J Clin Nutr 1998;67:1041S–5S.
1994;18:361–70.
Impact of zinc fortification on zinc nutrition
Sonja Y. Hess and Kenneth H. Brown
Abstract it will be important to evaluate the outcomes of such

programs.
Food fortification is increasingly recognized as an effec-
tive approach to improve a population’s micronutrient
status. The present report provides a critical review of Key words: Bioavailability, household fortification,
the scientific evidence currently available on the impact zinc absorption, zinc fortification, zinc status
of zinc fortification on zinc nutrition.
The available studies clearly show that zinc fortifica-
tion can increase dietary zinc intake and total daily zinc Background
absorption. Most absorption studies also indicate that
adding zinc to food does not adversely affect the absorp- Food fortification can be defined as the deliberate addi-
tion of other minerals, such as iron. Despite the positive tion of one or more nutrients to particular foods so as
effect of zinc fortification on total zinc absorption, only a to increase the intake of these nutrients and correct or
few studies have found positive impacts of zinc fortifica- prevent a demonstrated deficiency, thereby providing
tion on serum zinc concentrations or functional indica- some health benefit [1]. Food fortification is often
tors of zinc status. The reasons for these inconsistent considered the most economical approach to reduce
results are uncertain but may relate to the choice of food nutritional deficiencies in settings where suitable
vehicles, the age group and zinc status of the study popu- food vehicles are available, the food industry is suffi-
lations, or particular aspects of the study design. Thus, ciently developed to be able to produce and distribute
additional research is needed to determine the impact of these foods, and higher-risk subgroups of vulnerable
zinc fortification, with or without other micronutrients, populations have access to adequate amounts of these
in populations at risk for zinc deficiency. foods. The recent World Health Organization (WHO)
Because of the benefits of increasing intake in popula- publication on food fortification distinguishes among
tions at high risk for zinc deficiency, the documented three possible approaches: mass, targeted, and market-
increase in total zinc absorption that occurs following driven fortification [1]. Mass fortification refers to
zinc fortification, the absence of any adverse effects, and the addition of micronutrients to edible products that
the relatively low cost of adding zinc, public health plan- are consumed regularly by a large proportion of the
ners should consider including zinc in mass and targeted general public, such as cereal flours, vegetable oils
fortification programs in such populations. Because of and fats, milk, and condiments. Targeted fortification
the limited available information on program impact, is defined as the fortification of foods designed for
specific population subgroups, such as complemen-
tary foods for young children, foods for institutional
programs aimed at schoolchildren or preschoolers, and
The authors are affiliated with the Department of Nutrition
and the Program in International and Community Nutrition, foods used for emergency situations. Fortification is
University of California, Davis, California, USA; Kenneth considered market-driven when a food manufacturer
H. Brown is also affiliated with Helen Keller International, takes the initiative to add one or more micronutrients
Dakar, Senegal. to processed and branded foods.
H. Brown, Department of Nutrition, University of California, Food fortification has become an increasingly attrac-
One Shields Ave., Davis, CA 95616, USA; e-mail: khbrown@ tive strategy in lower-income countries, and a growing
ucdavis.edu. number of programs are being implemented. However,
S80 S. Y. Hess and K. H. Brown
surprisingly few of these fortification programs have Section 5: What are the steps in implementing zinc
been formally evaluated for their impact on the target fortification programs and what additional research
population’s nutrient intake and subsequent nutritional is needed?
status [1]. The present report provides a critical review
of the scientific evidence currently available on the
impact of zinc fortification of foods on zinc nutrition. Section 1
Although food fortification is most commonly
understood to refer to the addition of nutrients to food Does zinc fortification affect total zinc intake and frac-
at the industrial site of food processing or production, tional and total absorption of zinc? Are these effects mod-
fortification can also occur at the community or house- ified by consumer-, fortificant-, or diet-related factors or
hold level. In some cases, it can be very challenging to the presence of other nutrients in the fortified food?
decide whether to categorize a particular intervention
trial as an example of food fortification or nutrient Conclusions
supplementation, particularly if the nutrient is added
to food at the point of consumption rather than at a The available studies clearly show that zinc fortifica-
central site. For the purpose of this review, we included tion can increase dietary zinc intake and total daily
data from intervention trials in which the added zinc absorption. Although fractional absorption of
micronutrients were consumed together with foods, zinc (i.e., the percentage of dietary zinc intake that
regardless of the site where the nutrients were added. is absorbed) decreases with increasing zinc intake,
For example, studies in which nutrients were added to the total amount of absorbed zinc (TAZ) increases
infant formula in the home just before each serving in relation to the amount of zinc consumed until it
were considered as food fortification trials along with approaches a plateau at higher levels of zinc intake.
other studies in which the nutrients were added to the Zinc seems to be absorbed equally well from foods
formula at the industrial production site. We consid- fortified with zinc oxide or zinc sulfate, the two cheap-
est sources of zinc that are generally recognized as safe
ered the case of home fortification with products like
for human consumption, although extension of this
multiple micronutrient (MMN) powders as a special
conclusion to infants and young preschool children
category of intervention, because the additional nutri-
remains to be confirmed. The presence of phytic acid
ents (typically provided in the form of a dry powder to
in food reduces zinc absorption from zinc-fortified
be mixed with foods at the time of serving the meal) are
foods, but a greater amount of zinc is absorbed when
usually administered just once each day, and the food these foods are fortified than would have occurred if
vehicle is generally not stipulated, making these aspects they were not fortified. There is little evidence of any
of the implementation strategy different from typical benefit of putative zinc absorption enhancers on zinc
fortification interventions. Other aspects of the house- absorption from zinc-fortified foods, but the number
hold fortification strategy, such as the potential for of available studies is small. There is insufficient infor-
interactions between the micronutrients and the food mation available to assess the effects of cofortification
components, are more akin to food fortification carried with other micronutrients on zinc absorption from
out during the course of the food production [2]. zinc-fortified foods.
This paper is divided into five sections, which
address the following sets of questions in relation to
Detailed review of evidence
zinc-fortified foods:
Overview
Section 1: Does zinc fortification affect total zinc
intake and fractional and total absorption of zinc? Are Background information on tracer studies to assess zinc
these effects modified by consumer-, fortificant-, or bioavailability. Nutrient absorption from foods can be
diet-related factors or the presence of other nutrients measured by using metabolic tracers. Such tracers are
in the fortified food? uniquely detectable substances (usually isotope-labeled
forms of the nutrient of interest) that presumably are
Section 2: Does zinc fortification affect biochemical metabolized in the same way as the nutrient that is
indicators of zinc status and zinc-related functions? naturally present in the food or added to it as a forti-
Section 3: Does household-level fortification with ficant. To assess the bioavailability of a fortificant, in
multiple micronutrients (MMN) (including zinc) affect particular, it is important that the tracer be present in
indicators of zinc status and zinc-related functions? food in the same chemical form and phase as the for-
tificant. In this section, we review the available studies
Section 4: Are there any adverse clinical effects of of zinc absorption from zinc-fortified foods, based on
zinc fortification or negative effects of zinc fortifica- studies that used radioisotopes or stable isotope tracers
tion on the utilization of other nutrients due to zinc of zinc. Some of this information has been reviewed
fortification? previously [3].
Zinc fortification S81
Bibliographic search. Data sets for inclusion in this low-zinc meals (38.8% to 40.6%) than from the high-
analysis were identified by using a computerized bib- zinc meals (22.7% to 26.7%), but the total absorbed zinc
liographic search (PubMed), with the following key (TAZ) was greater from the high-zinc meals (0.73 mg
words: 1) zinc; limiting for human, English, clinical of zinc/meal) than from the low-zinc meals (0.48 mg
trial, meta-analysis, randomized controlled trial; 2) zinc of zinc/meal), as presented in figure 2.
fortification; limiting for human, English; and 3) zinc In a second study, Swedish adults consumed meals
fortif*; and limiting for human, English. A total of 1,673 containing bread prepared from either whole-wheat
articles were identified. All titles or abstracts of every flour (1.3 mg of zinc/meal) or refined-wheat flour (72%
record identified were reviewed. Of these, only nine extraction rate; 0.4 mg of zinc/meal), with or without
articles specifically examined zinc bioavailability from additional zinc, as zinc chloride, at a final zinc content
zinc-fortified foods. These articles will be summarized of 3.5 to 3.6 mg/meal when the fortified breads were
in the following section. provided [5]. With both types of bread, FAZ from the
fortified breads was lower than that of the intrinsic
Results of tracer studies of zinc absorption from zinc- zinc contained in the unfortified food (fig. 1). Never-
fortified foods theless, TAZ was greater when the bread was fortified,
Zinc absorption from foods fortified with different levels although the increment was relatively small with the
of zinc. Three studies are available in which different whole-wheat product (fig. 2).
levels of zinc fortification (or the same foods with and A third study was conducted in Peruvian preschool
without fortification) were compared. One study in children who received two meals (breakfast and lunch)
Danish adults was designed to examine the effect of that contained a total of 100 g of wheat products (bis-
folic acid on zinc absorption from bread prepared with cuits and noodles) fortified with 3 mg of iron and 0, 3,
or without folic acid and either low or high zinc content or 9 mg of zinc, as zinc sulfate, per 100 g of wheat [6].
[4]. The final bread meal contained a total of either As with the findings of the above-mentioned studies,
1.2 mg of zinc/meal or 2.9 to 3.0 mg of zinc/meal as the mean TAZ was positively related to zinc intake
added zinc chloride solution (prepared from zinc oxide (fig. 2), despite the inverse relation between zinc
and dilute hydrochloric acid). Folic acid did not affect intake and FAZ (fig. 1). The findings of these three
zinc absorption from breads containing either amount studies indicate that increasing zinc intakes by adding
of zinc. As indicated in figure 1, the fractional absorp- greater amounts of zinc to food results in greater net
tion of zinc (FAZ) was significantly greater from the absorption of zinc, although the increments in TAZ
Hansen [4] a ,c Sandström [5] b,c López de Romaña [6] a

Whole Refined
60 wheat wheat
50
p < .001 p < .001 p < .001 p < .001
40
FAZ (%)
30
20
10
0
1.2 3.0 1.3 3.5 0.4 3.6 2.14 4.72 10.04
Total amount of zinc (mg)
FIG. 1. Effect of the level of zinc fortification of wheat products on fractional absorption of zinc (FAZ) in studies
of adults [4, 5] and preschool children [6]
a. Results are shown as mean ± SD.
b. Results are shown as mean (range).
c. For the purpose of this review, we calculated p values from the published results, using two-way ANOVA. For Hansen et al. [4],
the p value was calculated under the conservative assumption of zero within-subject correlation. For Sandström et al. [5], the
SD was estimated under the assumption that the range represents the mean ± 2 SD.
Reproduced with permission from Int J Vitam Nutr Res 2007;77(3):174–81 © Hogrefe & Huber Publishers [3]
Hansen [4] a ,c Sandström [5] b López de Romaña [6] a

Whole Refined
2.0 wheat wheat
1.5 p < .001 p < .05 p < .001 p < .001

TAZ (mg)
1.0
0.5
0
1.2 3.0 1.3 3.5 0.4 3.6 2.14 4.72 10.04
FIG. 2. Effect of the level of zinc fortification of wheat products on total absorption (TAZ) of zinc in studies of adults
[4, 5] and preschool children [6]
b. Results are shown as mean (range).
c. For the purpose of this review, we calculated p values from the published results using two-way ANOVA, under the conservative
assumption of zero within-subject correlation.
are progressively less as the level of zinc fortification common cereal staples. Whether this is also true for
or total zinc intake increases. A similar conclusion has infants and young preschool children is not known.
been reported from a recent study of varied doses of One other study is available in which zinc absorp-
zinc supplements [7]. tion was measured from a wheat-soy-milk porridge
Bioavailability of different chemical forms of zinc. that was fortified with a mixture of micronutrients
Several zinc compounds are generally regarded as safe that contained either zinc sulfate or zinc methionine
(GRAS) for human consumption and are, therefore, [11]. There were no significant differences in fractional
available for use in food fortification. Zinc oxide, which absorption of zinc from the respective diets, even
is the cheapest GRAS zinc compound, is insoluble at though the zinc methionine-containing fortificant
neutral pH, so concerns have been raised about its bio- mixture also contained other putative enhancers of zinc
availability from fortified foods. Three well-designed absorption, as described below.
tracer studies have been published in which zinc absorp- Effect of phytate on zinc absorption from zinc-fortified
tion was compared for foods fortified with either zinc foods. Inositol phosphate (phytic acid or phytate) is the
oxide or zinc sulfate in amounts to provide the same chemical form in which phosphorus is stored in plant
level of zinc fortification. There were no differences in seeds, such as cereal grains and legumes, that are used
zinc absorption by healthy US adults from the two zinc for human consumption. The phosphate groups in
compounds when they were provided in either a high- phytate can form strong and insoluble complexes with
phytate, whole-wheat porridge or a low-phytate bread divalent cations such as zinc, and because the gastroin-
prepared from yeast-treated, refined-wheat flour [8]. testinal tract of humans lacks any significant phytase
Similarly, in studies conducted in Indonesian school- activity, phytate-bound minerals are excreted in the
children, there were no differences in zinc absorption stool. Several studies have been conducted to measure
from zinc-fortified, refined-wheat dumplings when zinc absorption from high- or low-phytate foods that
the two chemical forms of zinc were compared [9], have been fortified with zinc. Although the phytate con-
and in a study carried out among Mexican women, tents were not necessarily the only differences between
there were no differences in zinc absorption from the high- and low-phytate meals in each of the studies
maize tortillas fortified with either form of zinc [10]. In described below, it seems likely that the phytate:zinc
summary, as shown by the results presented in figure molar ratio of the meals was the major factor account-
3, the available evidence suggests that there is no dif- ing for the observed differences in FAZ.
ference in zinc absorption by school children or adults In one study of Swedish adults, which was described
when either zinc oxide or zinc sulfate is used to fortify above, bread was prepared from either whole-wheat
López de Romaña [8] a ,d Hotz [10] b,d Herman [9] c,d

Whole Refined
40 wheat wheat
30
FAZ (%)
20
10
0
3.1–3.3 3.1–3.7 3.3 1.75
ZnSO4 ZnO
FIG. 3. Fractional absorption of zinc (FAZ) from foods fortified with ZnO or ZnSO4 in studies of adults [8, 10]
and children [9]
a. Results shown as geometric mean (95% CI).
b. Results shown as mean (95% CI).
c. Results shown as mean ± SD.
d. No significant differences by type of fortificant within study.
flour or 72% extraction refined wheat [5], and zinc in Peruvian schoolchildren [13]. One serving of the
chloride was added to each type of bread to produce beverage contained 3.75 mg of zinc as zinc gluconate,
a final zinc content of 3.5 to 3.6 mg/meal. As shown and the meal contained an additional 1.1 mg of zinc
in figure 4, FAZ from the zinc-fortified, low-phytate, and 13.6 mg of phytate (final phytate:zinc molar ratio
refined-wheat bread was significantly greater than that of the combined beverage and meal ~0.3:1). The mean
from the zinc-fortified, phytate-containing, whole- FAZ values from the beverage alone (22.8 ± 7.6%)
wheat bread (13.2% vs. 8.2%). Similarly, in the afore- and from the beverage and the meal together (24.5 ±
mentioned study by López de Romaña et al. [8], FAZ 10.7%) were not significantly different. These results
was approximately twice as great from a refined-wheat, are not unexpected because of the low phytate content
yeast-fermented bread than from a whole-wheat, of the meal and the fact that, for the purpose of the
unfermented porridge (fig. 4), even though their zinc absorption study, additional zinc was included in the
contents were in the same general range (3.1 to 3.7 mg beverage to ensure that the same total amount of zinc
of zinc per serving). In a recent study among healthy was consumed as when the beverage and the meal were
Swedish adults, Fredlund et al. [12] investigated the taken together.
impact of the addition of various amounts of phytate Potential enhancers of zinc bioavailability from for-
to white wheat bread rolls on zinc absorption. The zinc tified foods. Several studies indicate that the metal-
content was adjusted to 3.1 mg by adding zinc chloride. chelating compound ethylenediaminetetraacetic acid
The investigators found a progressive decrease in zinc (EDTA) facilitates iron absorption from individual
absorption with increasing amounts of added phytate. foods and mixed diets [14–17], and additional research
Thus, each of the foregoing studies found that phytate has been completed to determine whether EDTA also
inhibits zinc absorption from zinc-fortified foods, enhances zinc absorption from foods fortified with
although the total amount of zinc absorbed from both zinc and EDTA.
phytate-containing foods is greater when the foods are In a study of 24 Sri Lankan schoolchildren [18],
fortified with zinc than when they are not fortified. the mean FAZ was significantly greater (13.5 ± 6.0%
Zinc absorption from beverages fortified with zinc. vs. 8.8 ± 2.0%, p = .037) when 9.6 mg of Na2H2EDTA
The recent development of sweetened or flavored was added to a local food that was prepared from 25 g
beverages fortified with MMN provides an alternative of rice flour (containing 0.9 mg of iron and 0.9 mg of
method for delivering these nutrients. Zinc absorption zinc) and fortified with 1.5 mg of iron as ferrous sulfate
from one such beverage and the effect of simultane- and 1.5 mg of zinc as zinc oxide (EDTA:zinc molar
ous consumption of a low-phytate meal were studied ratio, ~1.4:1; phytate:zinc molar ratio ~1:1) than when
EDTA was not added, as shown in figure 5. fortificant mixture was designed in an attempt to
By contrast with the foregoing results, Hotz et al. maximize mineral uptakes from the porridge. The final
[10] found no difference in zinc absorption from meals porridges contained either a low (192 mg) or a high
containing several foods, including a 64-g maize tortilla (392 mg) amount of calcium. Except for a marginally
that was fortified with zinc oxide alone, zinc oxide plus significant, small negative impact of the higher calcium
Na2H2EDTA, or Na2ZnEDTA in a study of 42 adult intakes, there were no significant differences in zinc
Mexican women. The maize flour incorporated in absorption with any of the diets, possibly because the
the tortilla was fortified at a level of 40 mg of zinc/kg relatively high level of zinc intake reduced FAZ to such
to provide an additional 1 mg of zinc per tortilla, and an extent (~7% FAZ overall) that any smaller effects of
the whole meal (including beans, tomato sauce, milk, the other factors would no longer be detectable.
and coffee) contained a total of 3.3 mg of zinc, with a In summary, there is evidence of an enhancing effect
phytate:zinc molar ratio of ~17:1. The molar ratio of of EDTA on zinc absorption from zinc-fortified foods
EDTA:zinc in the tortilla was 0.5:1, although the final from just one of the three available studies, although
ratio in the meal was ~0.2:1. Possible explanations for the possible modifying effect of different molar ratios
the discrepant conclusions of the two studies are the of enhancer:zinc and phytate:zinc on the impact of
dissimilar ratios of EDTA:zinc that were used or the potential enhancers of zinc absorption needs further
differences in the phytate:zinc ratios of the respective study.
diets. Effect of cofortification with other micronutrients on
A third study of US adults [11], which was described zinc absorption from zinc-fortified foods. It is conceiv-
briefly above, compared zinc and iron absorption able that cofortification with other micronutrients
from a wheat-soy-milk porridge that was fortified could affect zinc absorption from the zinc-fortified
to provide an additional 10 mg of each mineral per food. However, with the exception of the study of
serving with either ferrous sulfate and zinc sulfate folic acid [4] and the study of calcium [11] that were
or Na2FeEDTA, zinc methionine, ascorbic acid, and described above, we could not locate any other reports
citric acid (EDTA:zinc molar ratio, ~1.2:1). The latter of research that specifically isolated the effect of other
Sandström [5] a,c López de Romaña [6] b

Whole Refined Whole Refined
30 wheat wheat wheat wheat
25 p < .001 p < .001
20
FAZ (%)
15
10
0
3.5 3.6 3.1–3.3 3.1–3.7
~8.6:1 ~0.4:1 ~12:1 ~0.5:1
Total zinc content (mg)

Phytate:zinc molar ratio
FIG. 4. Fractional absorption of zinc (FAZ) from whole-wheat and refined-wheat products
fortified with zinc [5, 8]
a. Results are shown as mean (range).
b. Results are shown as geometric means (95% CI).
c. For the purpose of this review, we calculated p values from the published results using two-way
ANOVA. The SD was estimated under the assumption that the range represents the mean ± 2
SD.
Reproduced with permission from Int J Vitam Nutr Res 2007;77(3):174–81 © Hogrefe & Huber
Publishers [3]
fortificants on zinc absorption from zinc-fortified Section 2

foods. Most studies investigating the impact of other
micronutrients on zinc absorption only considered the Does zinc fortification affect biochemical indicators of
effects of these other fortificants on the intrinsic zinc zinc status and other zinc-related functions?
content of the food.
Summary of tracer studies of zinc absorption from Conclusions
zinc-fortified foods. In summary, the results from the
available sets of tracer studies indicate that the addi- The effects of zinc fortification on indicators of zinc
tional zinc provided in fortified foods can contribute status or other potentially zinc-related functional
positively to the total amount of absorbed zinc. Mul- outcomes were analyzed according to the type of food
tiple studies indicate that FAZ declines progressively product that was fortified and whether or not zinc was
with increasing levels of added zinc. However, the the only nutrient that differed by study group. The
total amount of absorbed zinc increases in relation to results of seven available trials of milk products with
the amount consumed. There does not appear to be or without zinc fortification are inconsistent, although
any significant difference in zinc absorption from zinc the results suggest that zinc-fortified milk products
oxide or zinc sulfate, regardless of the phytate:zinc appear to boost the zinc status of infants and young
molar ratio of the meal. However, it appears that children and increase the growth of premature infants
high-phytate meals depress zinc absorption from and malnourished children.
zinc-fortified foods. With the possible exception of We identified five studies that assessed the impact
Na2H2EDTA, there is no other evidence that putative of zinc-fortified cereal products among preschool-
enhancers of zinc absorption confer major benefits for aged or school-aged children. The limited available
zinc absorption from zinc-fortified foods, although the information suggests that zinc fortification of cereal
amount of relevant available information is still quite products might have a positive impact on serum zinc
limited. There is also too little information regarding concentration among school-aged children, but not
the effects of fortification with other micronutrients among younger children. There is insufficient informa-
on zinc absorption from cofortified foods to permit tion to determine whether zinc fortification of cereal
informed judgments on this issue. products could enhance growth or reduce morbidity
among children at risk for zinc deficiency because of
the small number of available studies and the fact that
Hettiarachchi [18] a Hotz [10] a,c Mendoza [11] b,c

ZnO + ZnO + ZnSO4 + ZnMet +
2.0 ZnO Na 2H2EDTA ZnO Na 2H2EDTA Na 2ZnEDTA FeSO4 Na 2FeEDTA
p = .037
1.5
FAZ (%)
1.0
0.5
0
2.4 3.3 11.4
~1.4:1 ~0.2:1 ~1:1
Total zinc content (mg)

EDTA:zinc molar ratio
FIG. 5. Effect of EDTA on fractional absorption of zinc (FAZ) in studies of adults [10, 11] and children [18]
b. Results are shown as mean (95% CI)
c. Groupwise comparisons in studies by Hotz et al. [10] and Mendoza et al. [11] are not significantly different.
these studies sometimes enrolled children who were be due to inadequate consumption of the zinc-fortified
not growth-restricted or were too brief to detect such food, an insufficient level of fortification, poor bioavail-
changes. ability of the zinc fortificant, or simply the fact that
Eight studies were identified in which the impact the study population was not initially zinc-deficient.
of fortifying foods with MMN, including zinc, was Thus, it is worthwhile in such evaluations to include
assessed. Only one of the six studies that monitored a separate study group that receives adequate doses of
serum zinc concentration found a positive impact on a readily absorbable zinc supplement (rather than a
final serum zinc concentration. Only two of the eight zinc-fortified food) as a positive control.
studies of MMN-fortified foods found a positive impact Bibliographic search. Data sets were identified for
on growth, although these effects may not have been this section by using a computerized bibliographic
due exclusively to zinc. The lack of growth impact in search (PubMed with key words as described above).
the other studies may be due to methodologic inad- In addition to the PubMed search, we included three
equacies, including short study durations, small sample reports of fortification trials that were in press at the
sizes, and study populations that were not sufficiently time [22–24] and one fortification trial using MMN
growth-restricted to benefit from zinc. that had included measurements of serum zinc con-
In summary, a few studies of zinc fortification have centration [25]. Randomized, controlled intervention
found positive impacts of zinc-fortified foods on trials among prepubertal children were considered
serum zinc concentrations and growth, and one study acceptable for inclusion in the present review either
reported a reduction in morbidity from diarrhea; but if they involved a zinc-only fortification group or if
most studies failed to detect such beneficial effects. zinc was given together with other nutrients and zinc
The reason for the inconsistent results of these stud- status indicators were assessed (specifically, serum zinc
ies is uncertain but may be related to the choice of concentration, growth, or morbidity). We identified a
food vehicles, the age group and zinc status of the total of 7 trials using milk or infant formula in young
study populations, or particular aspects of the study children and 14 trials using cereal products, other
design. Thus, additional research is needed to assess beverages, or condiments in prepubertal children.
the impact of zinc-fortified products, with or without Most of the available information on the impact of
other micronutrients, in populations known to be at zinc fortification interventions was collected during
risk for zinc deficiency. efficacy trials. Very little information is available from
large-scale, programmatic interventions. Except for two
Detailed review of evidence trials in elderly subjects [26, 27], no fortification trials
were identified in which zinc- or MMN-fortified foods
Overview were compared with nonfortified foods in other age
Background information on intervention trials with zinc- groups and serum zinc concentration was measured.
fortified foods. Evaluations of zinc fortification trials are Because one of these latter trials was not masked, there
challenging because of the lack of adequate biomarkers is insufficient information to warrant inclusion of this
of individual zinc status. Nevertheless, recent analyses age group in the review.
indicate that the mean serum (or plasma) zinc con-
centration of a population responds consistently to Results of intervention studies with zinc-fortified foods
zinc supplementation, regardless of the population’s As indicated above, a total of 21 controlled studies
initial mean serum zinc concentration [19]. Thus, were identified in which a zinc-fortified product was
the change in serum zinc concentration following the compared with the same product that did not include
introduction of a zinc-fortified food should be able to additional zinc. These studies examined a variety of
serve as a useful indicator of the impact of fortifica- food vehicles. In 10 of these studies (5 studies of milk
tion interventions as well. In addition to the change products or infant formulas and 5 studies of cereal
in serum zinc concentration, it is possible to use other products), zinc was the only nutrient that differed
functional indicators of zinc responsiveness, such as between study groups. In two studies of infant formula,
increased physical growth [20] or morbidity reduc- both zinc and copper were added to the study-group
tion [21], to assess the efficacy of zinc-fortification formula but not the control-group formula. In the
interventions, although several caveats must be borne remaining nine studies, zinc was one component of a
in mind. First, zinc must be the only nutrient that dif- MMN fortification premix, and the foods under study
fers between the intervention group and the control were fortified with either all of the nutrients in the
group to be able to attribute any functional responses premix or none of them. Because zinc was just one
specifically to zinc fortification. Second, it is likely micronutrient among several others in these MMN-
that functional responses are useful only if the study fortified foods, it is not possible to attribute any con-
population is zinc deficient initially. Thus, the failure of sequences of these latter interventions specifically to
a zinc-fortification program to deliver positive results zinc. Nevertheless, we examined the impact of these
with regard to functional indicators of zinc status may MMN-fortified products on serum zinc concentration
and other potentially zinc-related outcomes, such as than did the control group [34].
growth and morbidity, when relevant information was In summary, the results of these trials are inconsist-
available. ent, although more than half of the studies found a
Efficacy trials of zinc-fortified infant formulas or other positive impact of zinc-fortified milks on serum zinc
milk products in which zinc was the only nutrient that concentration. Two of the three trials in premature
differed between intervention and control groups. Of the infants and the one trial in severely malnourished
seven trials of milk-based products, zinc was the only children found a positive effect of zinc-fortified milks
micronutrient that differed between groups in five on physical growth, but no growth impact was detected
studies, and in two studies the treatment group received among healthy term infants, probably because the con-
both additional zinc and copper [28, 29] (table 1). trol formulas had adequate zinc contents for the term
Three studies were completed in healthy, full-term infants, who had no prior growth restriction.
infants [28, 30, 31], three were conducted in preterm Efficacy trials of zinc-fortified cereal products in which
infants [29, 32, 33], and one was conducted in severely zinc was the only nutrient that differed between the
malnourished infants [34]. Apart from the study of intervention and control groups. Results are available
malnourished children, which took place in Chile, the from five controlled studies of zinc-fortified, cereal-
remaining studies were completed in North America, based food products [6, 18, 22, 35, 36] (table 2). These
Europe, or Japan. The sample sizes of these studies were five studies were carried out in both industrialized and
generally small, ranging from 20 to 68 total subjects, lower-income countries, and they included a broad
thereby limiting their power to detect significant dif- range of products. The first study compared the effects
ferences in growth outcomes. The zinc concentration of of providing a ready-to-eat breakfast cereal with or
the milk formula that was provided to the term infants without additional zinc (3.75 mg of zinc added per 28-g
ranged from 1.0 to 1.8 mg/L in the control groups and serving, as zinc oxide) to healthy US schoolchildren for
from 3.2 to 5.8 mg/L in the intervention groups; the a period of 9 months [35]. On the basis of the number
respective milk zinc concentrations were similar in of servings consumed per week and the size of the
one of the three studies of preterm infants but were serving, children in the group that received the zinc-
considerably greater in the fortified milks provided in fortified cereal were estimated to receive 2.57 mg of
the two other studies, as described below. additional zinc per day during the course of the study.
Four of the seven studies of zinc-fortified milk found Children in the zinc group had a significantly greater
a significantly greater increase in the final serum zinc (i.e., less negative) change in plasma zinc concentration
concentration among children who received the higher than those in the control group. There were no signifi-
zinc-containing products, but three of the studies did cant differences between study groups in any growth
not. These differences among studies could not be increments, but the children were not underweight
explained by the clinical characteristics of the subjects, or stunted initially, so they would not be expected to
the level of zinc fortification, or the presence of other exhibit a growth response to zinc.
micronutrients in the formulas. A study was conducted among 7- to 11-year-old
Two of the three studies of healthy, full-term infants Turkish schoolchildren with low initial serum zinc
found no significant changes in length or weight among concentrations (< 65 μg/dL) who were provided bread
the groups that received the zinc-fortified product with or without zinc fortification for a period of 13
compared with the control groups [28, 31], but the weeks [36]. The paper published from this study
children included in these trials had no preexisting describes the level of fortification as “2 mg/kg/day
evidence of growth restriction. Similarly, in one study elemental zinc acetate,” which seems to refer to the
of low-birthweight infants, in which the infants’ birth- zinc intake per kilogram body weight of the child,
weights ranged from 1,800 to 2,500 g and the milk zinc not to the zinc concentration of the bread. Because
concentrations were either 1.4 or 4.1 mg/L in the two the report provides no information on the children’s
study groups, there were no groupwise differences in body weights or the amounts of bread consumed, it is
the increments in length, weight, or head circumfer- not possible to calculate the level of zinc fortification
ence [32]. By contrast, the two studies of very-low- or the amount of additional zinc that was consumed.
birthweight premature infants who were provided with Children who received the zinc-fortified bread had sig-
zinc-fortified infant formulas containing either 6.7 or nificantly greater increases in serum zinc and alkaline
5 mg/L versus 10 or 11 mg/L of zinc, respectively, found phosphatase concentrations at the end of the interven-
a significantly greater change in height-for-age z-score tion period than their counterparts who received the
(HAZ) [33] or change in length [29] compared with nonfortified bread.
the control group. There were no groupwise differences Two studies were completed among Peruvian chil-
in the final growth increments of the malnourished dren to assess the effects of adding zinc to iron-fortified
Chilean children at 60 days, but those who received wheat flour. In the first study [6], which was described
the zinc-fortified preparation had significantly greater above in the section on tracer studies, children 3 or 4
height increments after 30 and 45 days of treatment years of age received two meals that contained wheat
TABLE 1. Efficacy trials of zinc fortification of infant formula: Impact on zinc-related indicators (serum zinc concentration, height,
weight, morbidity) and possible adverse effects on iron and copper status
Country, year Zinc content in

[reference] Inclusion Sample Duration infant formula
author Study design Age criteria sizea (mo) Group (mg/L)
USA, 1976 Double-blind, control- 6 days Full-term, 68 6 Zinc 5.8d
[30]c led. No statement on healthy
Walravenes randomization infants
Control 1.8
Japan, 1984 Placebo-controlled. 1 mo Full-term, 39 5 Zinc 3.2d

[31] Matsuda No statement on healthy
randomization and infants,
blinding normal birth
Control 1.0
weight
Finland, Randomized, control- 2–3.5 mo Full-term, 32 12 Zinc 5.1i

1994 [28] c led. healthy
Salmenperä No statement on infants
blinding
Control 1.1
Austria, Placebo controlled. 6 days Male, LBW, 20 3.8 Zinc 4.1

1985 [32] c Alternative assign- preterm
Haschke ment. No statement infants
on blinding
Control 1.4
Canada, 1993 Randomized, double- 1 mo prior Preterm, very 50 12l Zinc 11d
[33] Friel blind, controlled to home LBW infants
discharge
Control 6.7
Spain, 2003 Randomized, double- ~ 4 wk Preterm, very 36 6 Zinc 10i

[29]n Diaz- blind, controlled LBW infants
Gomez
Control 5
Chile, 1992 Double-blind, control- 7.6 ± 2.6 mo Malnourished 39 3.5 Zinc 15d
[34]n led. No statement on infants
Schlesinger randomization
Control 3.2
HAZ, height-for-age z-score; LBW, low-birthweight; NS, not significant; WAZ, weight-for-age z-score
a. Sample size includes total number in control and fortification groups.
b. Results are shown as mean ± SD for serum zinc, hemoglobin, and serum copper concentrations and as geometric mean (range of 1 SD)
for serum ferritin concentration.
c. Mean serum zinc concentration estimated from graph.
d. Infant formula was also fortified with iron (and in some cases other micronutrients) in both groups.
e. No significant differences in final serum zinc concentration between groups.
f. No significant difference in final concentrations.
g. Final concentrations are not adjusted for initial concentration.
h. P value for the difference in final serum zinc concentration between groups.
Initial serum Final serum Final hemo- Final serum Final serum
zinc concen- zinc concen- globin con- ferritin con- copper con-
tration tration Change in Change in centration centration centration
(μg/dL)b (μg/dL)b height weight Morbidity (g/L)b (μg/L)b (μg/dL)b
— 72 18.2 ± 0.5 4,282 ± 104 — — — 124.2 ± 5.8
(NS)e cm/6 mo (p g/6 mo (p < (NS)f,g
< .05) .03)
— 76 17.0 ± 0.5 3,867 ± 170 — — — 112.3 ± 6.1
cm/6 mo g/6 mo
73 ± 16 103 ± 17 1.0 ± 0.4 mm/ 26.4 ± 8.3 g/ — — — 111 ± 31
(p < .01)h day day (NS)f,g
(NS) (NS)
68 ± 12 78 ± 12 1.0 ± 0.4 mm/ 26.5 ± 7.0 g/ — — — 124 ± 21
day day
87 At 6 mo: No significant No significant — — — —
85 ± 3 difference difference
(p = .03)h in change in in change in
lengthj weightj
90 At 6 mo: — — — —
65 ± 4
110 106 1.2 ± 0.2 mm/ 27.7 ± 7.8 g/ — — — No significant
(p < .05) h day day difference
(NS) (NS) in serum
copper con-
120 75 1.2 ± 0.2 mm/ 31.0 ± 8.8 g/ — — —
centration at
day day
any timek
83 ± 26m 95 ± 17 0.087 ± 0.087 −0.06 ± 0.06 — — — —
(NS)e ΔHAZ/12 ΔWAZ/12
mo mo
(p = .004) (NS)
67 ± 14 93 ± 30 −0.027 ± 0.13 −0.11 ± 0.13 — — — —
ΔHAZ/12 ΔWAZ/12
mo mo
69 ± 20 119 ± 37 1.04 ± 0.07 180 ± 24 g/ — 122 ± 11 g/L — —
(p = .01)h mm/day wk (NS)o (NS)f,g
(p = .036)
78 ± 24 87 ± 30 0.99 ± 0.07 174 ± 32 g/ — 122 ± 9 g/L — —

mm/day wk
127 ± 36 122 ± 28 0.62 ± 0.23 24.9 ± 6.3 g/ Significantly 118 ± 7 (NS)f,g 15.6 (7.4– 155 ± 28
(NS)e mm/day day more 32.8) (NS) f,g
(NS) (NS) diarrhea epi- (NS)f,k
sodes in zinc
than in con-
153 ± 55 118 ± 38 0.58 ± 0.26 25.8 ± 10.2 g/ trol group 121 ± 8 12.6 149 ± 29
mm/day day (6.0–26.5)
i. Infant formula provided to zinc group was fortified with zinc and copper. Amounts of other micronutrients were equal for both groups.
j. Results not provided.
k. Results are presented in graph only.
l. Study duration was 12 months, but infants consumed infant formula for 6 months only.
m. Initial serum zinc concentrations were significantly different between groups (p = .041).
n. Additional information provided by author.
o. Final body weight was significantly different between groups (p = .01).
TABLE 2. Efficacy trials of zinc fortification of cereal products in which zinc was the only nutrient that differed between intervention
and control groups: Impact on zinc-related indicators (serum zinc concentration, height, weight, and morbidity) and possible adverse
effects on iron and copper status
Initial serum
Additional zinc concen-
Country, year Study Sample Duration Fortified zinc in zinc Study tration
[reference] author design Age sizea (mo) food group groups (μg/dL)b
USA, 1979 Randomized, 2–7 yr 96 9 RTE cereal 2.57 mg/day Zinc 80 ± 2c
[35] Hambidge double-blind,
controlled
Control
Turkey, 1998 Randomized, 7–11 yr 24 3 Bread 2 mg/kg/day Zinc 61 ± 4

[36] Kiliç double-blind,
controlled
Control 59 ± 3
Peru, 2005 Randomized, 3–4 yr 41 2 Biscuits/ 0, 3, 9 mg/ Zinc 9 mg/ 78 ± 16

[6] López de double-blind, noodlesj day day
Romaña controlled
Zinc 3 mg/ 77 ± 13
day
Control 71 ± 9
Peru, 2007 Randomized, 6–8 mo 178 6 Wheat ~3mg/day Zinc 78 ± 15

[22] Brown double-blind, porridgej
controlled
Control 79 ± 15
Sri Lanka 2004 Randomized, 6–10 yr 49 1 Rice flourj 1.5 mg/day Zinc 78 ± 13
[18]l Hettiarachchi controlled. No
statement on Control 79 ± 17
blinding
NS, not significant; RTE, ready-to-eat
a. Sample size includes the total number in the control and zinc fortification groups.
b. Results shown as means ± SD.
c. Mean serum zinc concentration at baseline for both groups combined.
d. P value for change in serum zinc concentration. Final mean and SD estimated from change in serum zinc concentration.
e. Not significant, but population not moderately or severely stunted or underweight.
f. No significant difference in change in concentration over study period.
g. P value for the difference in final serum zinc concentration between groups.
products (biscuits and noodles) fortified with iron and group had zinc added to their porridge (at a level to
0, 3, or 9 mg of zinc for a period of ~7 weeks. Despite provide ~3 mg additional zinc per day, as zinc sulfate),
the documented increase in their total zinc absorption, the second group received the same amount of zinc
as described above, there were no significant groupwise in the liquid vitamin supplement, and the third group
differences in their change in serum zinc concentration received no additional zinc. Mothers were instructed to
or physical growth, although the period of observation give the vitamin supplement apart from meals. Notably,
may have been too short to detect such changes. only the group that received additional zinc as a sup-
In the second study from Peru, infants who were plement demonstrated a significant increase in serum
initially 6 to 8 months of age were randomly assigned zinc concentration compared with the other groups,
to one of three groups, each of which received an iron- and there were no significant differences in growth or
fortified wheat-based porridge and a liquid multivi- morbidity between study groups, possibly because the
tamin supplement for 6 months to assess the effect of study population was not zinc-deficient.
the mode of delivery of additional zinc on their serum In the final study, Sri Lankan children who were
zinc concentrations, growth, and morbidity [22]. One enrolled in the aforementioned tracer study of the
Final serum Change in Change in Change in

zinc concen- hemoglobin serum ferritin serum copper
tration Change in Change in concentration concen-tration concen-tration
(μg/dL)b height weight Morbidity (g/L)b (μg/L)b (μg/dL)b
77 ± 14 0.176 ± 0.03 7.97 ± 2.33 g/ — — — −13.7 ± 5.5
(p < .05)d mm/day day (NS)e (NS)f
(NS)e
71 ± 14 0.170 ± 0.03 7.45 ± 1.67 g/ — — — −6.2 ± 4.8
mm/day day
82 ± 9 — — — No significant 30.4 ± 9.7 85 ± 16 (NS)h
(p = .0001)g difference in (NS)h
final concentra-
tion between
63 ± 3 — — — 24.1 ± 11.7 94 ± 16
groupsh,i
82 ± 10 (NS)k 2.9 ± 1.3 cm/2 1.0 ± 1.2 kg/2 — 14.2 ± 11.4 −4.3 ± 16.6 —
mo (NS) k mo (NS) k (NS) k (NS) k
76 ± 8 3.0 ± 1.4 cm/2 1.1 ± 1.1 kg/2 — 17.2 ± 17.8 −8.3 ± 26.3 —
mo mo
78 ± 12 2.9 ± 1.3 cm/2 1.5 ± 1.2 kg/2 — 23.4 ± 14.4 3.6 ± 37.6 —
mo mo
77 ± 11 (NS)g 6.9 ± 1.1 cm/6 1.3 ± 0.5 kg/6 No significant −0.12 ± 1.08 −15.3 ± 43.4 −5.1 ± 37.2
mo (NS) mo (NS) difference (NS)f (NS) f (NS) f
in diarrhea
prevalence
and incidence
76 ± 14 7.0 ± 1.1 cm/6 1.3 ± 0.5 kg/6 −0.26 ± 1.05 −14.3 ± 27.4 9.6 ± 33.7
mo mo
84 ± 17 (NS)m — — — 7.5 0.5 —
(NS)f,n (NS) f,n
80 ± 17 — — — 3.7 3.6 —
h. Authors provided final values only after 3 months of supplementation. Change in concentration could not be calculated. No significant
difference between groups.
i. Results for hemoglobin concentration were unclear.
j. The food vehicle was also fortified with iron (and in some cases other micronutrients) in both groups.
k. No significant groupwise difference in treatment effects.
l. Results combined for the two groups receiving zinc and the two control groups.
m. Change in serum zinc concentration tended to be greater in the group receiving zinc than in the group without zinc (p = .065).
n. Changes in hemoglobin and serum ferritin concentration are presented as means without SD.
effects of iron, Na2EDTA, and zinc fortification of rice of zinc in the product; but, interestingly, the two test
received the respective study meals for a total period products that were included in the positive studies did
of 30 days [18]. The children in the two groups who not contain added iron, whereas all of the products
received the zinc-fortified rice snack (which provided used in the studies without significant impact were
1.5 mg of additional zinc per day, as zinc oxide) had a cofortified with both iron and zinc. Thus, it is possible
marginally greater increase in their serum zinc concen- that the iron in these products interfered with the uti-
trations (p = .065) at the end of the study. lization of zinc. However, some of the aforementioned
In summary, the results of these controlled trials of studies of infant formulas that were cofortified with
zinc-fortified cereal products carried out in infants and both iron and zinc did find a positive impact on serum
young children are inconsistent. In some of the stud- zinc concentration, so this hypothesized effect of iron
ies, but not all, there were significantly greater changes fortification on zinc uptake was not apparent with the
in serum zinc concentration among the groups that milk-based products.
received the zinc-fortified product. The differences None of the three available studies of cereal-based
among studies could not be explained by the amount products in which growth outcomes were reported
(table 2) identified a positive impact of zinc fortifi- been due to high concentrations of zinc-absorption
cation on children’s growth. However, the children inhibitors in the food ingredients.
enrolled in these studies may not have been zinc-defi- Faber et al. [38] examined the effects of MMN-forti-
cient, and in some cases the period of observation was fied maize porridge among South African infants 6 to
too short to detect growth differences or the children 12 months of age. Families were given two sachets con-
were not sufficiently growth-restricted to expect such taining 25 g of finely ground dry maize meal per day
differences to occur. Thus, additional studies of zinc- for a period of 6 months, with the expectation that the
fortified cereal products should be carried out over a infants would consume up to 40 g per day. The fortified
sufficiently long duration in populations at risk for product contained a total of 3 mg of zinc as zinc sulfate
zinc deficiency. per 40 g dry weight, as well as iron, copper, selenium,
Efficacy trials of products fortified with MMN, includ- B vitamins, and vitamins A, C, and E. Children in the
ing zinc, in young preschool-aged children. Several stud- control group were given the same product without
ies have evaluated the effects of fortification of different added micronutrients. At the end of the intervention
food products with MMN, including zinc, among period, children in the fortification group had greater
young children [23, 25, 37–42] (table 3). Although it increases in hemoglobin and serum ferritin concen-
is not possible to attribute any consequences of these trations, but there were no differences in the change
interventions specifically to zinc, their impact on in serum zinc concentration compared with those in
serum zinc concentration or other potentially zinc- the control group. There were no fortification-related
related outcomes can be assessed. A total of seven differences in growth increments, but the population
studies of MMN-fortified foods were conducted among was not underweight or severely stunted.
children in Ghana [37], South Africa (two studies) [38, A study in China [39] enrolled children 6 to 13
42], Ecuador [23], Mexico [25], China [39], and India months of age who were randomly assigned by village
[40, 41]. This set of studies examined a broad range to receive 17-g of either a fortified or non-fortified
of MMN-fortified products, including cereal-based “weaning rusk” prepared from wheat flour, sugar,
complementary foods, milk, other beverages, and and vegetable oil once daily for a period of 3 months.
condiments. The first five of the cited studies included The fortified product contained 3 mg of zinc as zinc
measurements of serum zinc concentration as one of gluconate along with iron, calcium, B vitamins, and
the study outcomes, and the latter two measured either vitamins A and D. Children in the control villages
children’s growth [39] or growth and morbidity [40, 41] received the same rusk without added micronutrients.
as the only potentially zinc-related outcomes. Five of Children who consumed the fortified rusk main-
the studies [25, 37–41] compared the fortified product tained their hemoglobin levels, whereas hemoglobin
with the same product to which no micronutrients were concentrations declined among those in the control
added; one of the South African studies [42] and the group, indicating that children were able to utilize the
study in Ecuador [23] compared the MMN-fortified iron and other micronutrients in the MMN-fortified
product with the usual home diets provided by the product. Although serum zinc concentrations were not
family. determined, there were no differences in the growth
In the Ghana study [37], children were randomly increments of children in the respective study groups.
assigned to one of four treatment groups at 6 months However, the rates of stunting were very low in both
of age. For the current review, only the two groups groups.
that received the same cereal–legume preparation In a recently published study from Mexico, 10- to
(“Weanimix”), with or without added micronutrients, 30-month-old children from low-income families
were considered. Weanimix was prepared by mixing received 400 mL per day of whole cow’s milk fortified
roasted and ground maize, soybeans, and groundnuts; with vitamin A, vitamin C, folic acid, ferrous gluconate,
the MMN mixture added to the fortified product for and zinc oxide, which provided an additional 5.3 mg
younger children provided an additional 143 mg of zinc of zinc per day [25]. After 6 months of intervention,
as zinc oxide per kilogram dry weight, along with addi- anemia prevalence declined significantly from 41.4%
tional iron, copper, calcium, phosphorus, potassium, B to 12.1% among the children who received the forti-
vitamins, and vitamins A and C. Children who received fied milk, whereas there was no change in those who
the fortified product consumed ~7 mg more zinc per received the nonfortified milk (30% and 24% at base-
day than those in the comparison group. Surprisingly, line and 6 months). There was also an improvement in
there were no differences between groups with regard iron status, as reflected by significantly lower transfer-
to their final serum zinc concentrations, and there were rin receptor concentrations in the fortified group. In
no differences in growth increments, although the pop- contrast, there were no groupwise differences in mean
ulation was not severely stunted or underweight. The serum zinc concentrations.
lack of impact on serum zinc concentration may have In a community-based, double-blind, individually
randomized trial in a periurban setting in India, 1- In summary, serum zinc concentrations of young,
to 3-year-old children received either fortified milk preschool-aged children did not respond to zinc-con-
(n = 315) or control milk (n = 317) for 1 year. The taining, MMN-fortified, cereal-based complementary
fortified milk provided an additional 7.8 mg of zinc foods or MMN-fortified milk in the five studies that
and 9.6 mg of iron, along with other micronutrients. measured this outcome. Two of the studies (one of
Although final serum zinc concentrations were not which examined a milk product, the other of which
reported, two possibly zinc-related functional outcomes assessed a cereal-based porridge) detected a positive
were significantly affected. Children in the fortified- impact on growth. However, as discussed above, these
milk group had significantly greater changes in HAZ results may not be due specifically to zinc. Moreover,
(mean difference, 0.19; 95% CI, 0.12 to 0.26; p < .001) because the children in most of the other study popu-
and weight-for-age z-scores (WAZ) (mean difference, lations were not growth-restricted, it is not possible
0.24; 95% CI, 0.11 to 0.36; p < .001) [41]. Moreover, the to judge whether the inclusion of zinc in the MMN
children in the fortified-milk group had significantly premix could have had a positive impact on growth in
fewer episodes of diarrhea [40]. However, because of those cases. Only the study from India reported mor-
the study design, it is uncertain whether these results bidity outcomes, and the investigators found reduced
were due specifically to zinc. diarrhea rates among the children who received the
In the second study from South Africa [42], Oelofse fortified product.
et al. randomly assigned 46 children 6 to 12 months Efficacy studies with MMN-fortified foods in school-
of age to receive an infant cereal (60 g dry weight aged children. Two intervention trials provided MMN-
per day) fortified with 5.6 mg of zinc as zinc sulfate fortified foods, including zinc, to school-aged children.
and iron, calcium, phosphorus, potassium, iodine, B However, one of the studies was excluded from this
vitamins, and vitamins A, C, D, and E for a period of 6 analysis because a single serving of a flavored beverage
months or to a control group that was not provided any was provided apart from meals, so the study was not
complementary food. Preliminary dietary studies indi- considered a true fortification trial [43].
cated that most children in the study area consumed One study examined the impact of a zinc-containing,
a “comparable cereal” that apparently was fortified MMN-fortified seasoning powder among schoolchil-
only with iron and a smaller amount of vitamin A, but dren in northeastern Thailand [24, 44]. The powder
none of the other nutrients. There were no significant was added to rice or noodles during a midday meal
differences between study groups in the change in served at the schools. The fortified powder contained
hemoglobin or serum zinc concentrations or change 5 mg of zinc as zinc sulfate and iron, iodine, and vita-
in weight or length, although the children in the study min A. Control children received the same seasoning
population were not severely stunted or underweight powder without added nutrients. At baseline, there was
and the sample size was fairly small. no difference in the percentage of children with low
Lutter et al. [23] evaluated the impact of the Ecua- serum zinc concentrations (< 65 g/dL), but children
dorian national food and nutrition program, which who received the fortified product had higher mean
provided a fortified complementary food to selected serum zinc concentrations at the end of the study than
children 6 to 12 months of age living in poor commu- the control children, and the former were less likely
nities. In a longitudinal, quasi-experimental design, to have low serum zinc concentrations (< 65 g/dL)
children in the program communities were compared (27.5% vs. 34.7%; odds ratio, 0.63; 95% CI, 0.42 to
with their counterparts in control communities who 0.94; p = .024) [44]. Children in the fortification group
did not receive any complementary food from the also had higher final hemoglobin concentrations, after
program. The children in the program communities adjustment for baseline values, and greater urinary
received 6.5 mg of additional zinc along with other iodine concentrations. In addition, the intervention
micronutrients in each daily ration. After 13 months reduced the incidence of respiratory-related illnesses
of intervention, 14% of the children in the program (rate ratio, 0.83; 95% CI, 0.73 to 0.94; p = .004) and
group (n = 171) and 24% of the children in the control diarrhea (rate ratio, 0.38; 95% CI, 0.16 to 0.90; p = .027)
communities (n = 150) were underweight (p = .02). [24]. A positive impact of the fortified product was
Children in the program communities consumed sig- also found on short-term cognitive function assessed
nificantly more energy, protein, fat, iron, zinc, calcium, by visual recall test.
magnesium, phosphorus, B vitamins, and vitamins A, By contrast with the foregoing results obtained with
C, and E than children in the control communities, a MMN-fortified foods in young children, serum zinc
difference that was due to the fortified complementary concentrations increased significantly in the one avail-
food provided by the program. There was no signifi- able study of schoolchildren who received a MMN-
cant difference between study groups in the mean final fortified seasoning powder with meals. The reason
serum zinc concentration. for these differences between the two sets of studies
TABLE 3. Efficacy trials of zinc and micronutrient fortification in which serum zinc concentration or other potentially zinc-related
outcomes were measured
Country, year Additional
[reference] Sample Duration zinc Other micronutrients
author Study design Age sizea (mo) Fortified food (mg/day) added to fortified food
Ghana, 1999 [37] Randomized, con- 6–12 mo 104 6 Maize-soy- ~7 Iron, copper, calcium,
Lartey trolled. No state- ground-nut potassium, phosphorus,
ment on blinding porridge vitamins A, B1, B2, B3,
B6, and B12, folic acid,
vitamin C
South Africa, Randomized, single- 6–12 mo 361 6 Maize 3 Iron, copper, selenium,
2005 [38] Faber blinded, controlled porridge vitamins A, B2, B6, B12,
C, and E
China, 1993 [39] Cluster-rand- 6–13 mo 226 3 Wheat flour 3 Iron, calcium, vitamins A,
Liu omized, control- weaning B1, B2, B3, and B12, folic
led. No statement rusk acid, vitamin D
on blinding
Mexico, 2006 Randomized, 10–30 130 6 Milk 5.3 Iron, vitamin A, vitamin
[25] double-blind, mo C, folic acid
Villalpando controlled
India, 2007 [40] Randomized, 1–3 yr 633 12 Milk 7.8 Iron, copper, selenium,
Sazawal, [41] double-blind, vitamins A, C, and E
Juyal controlled
South Africa, Randomized, 6–12 mo 46 6 Cereal 5.6 Iron, calcium, phospho-

2003 [42] controlled with porridge rus, potassium, iodine,
Oelofse nonintervention vitamins A, B1, B2, B3,
group, not blinded B6, and B12, folic acid,
pantothenic acid, biotin,
vitamins C, D, and E
Ecuador, 2008 Quasi-experimental 6–12 mo 321 10–13d Cereal 6.5 Iron, calcium, magne-
[23] Lutter porridge sium, phosphorus,
vitamins A, B1, B2, B3,
B6, and B12, folic acid,
vitamins C and E
Thailand, 2006 Randomized, 5–13 yr 569 7.8 Seasoning 5 Iron, iodine, vitamin A
[44] Win- double-blind, powder
ichagoon [24] controlled
Manger
HAZ, height-for-age z-score; MMN, multiple micronutrients; NS, not significant; RR, rate ratio between study groups; WAZ, weight-for-age z-score
a. Sample size includes the total number in the control and fortification groups.
b. All control groups received the same food product as the intervention group without added micronutrients, except for the studies by
Oelofse et al. [42] and Lutter et al. [23], in which the control groups received the usual, self-selected home diet.
c. Plus–minus values are means ± SD.
d. Blood collection at 10 months. Duration of intervention trial was 13 months.
e. P-value for the difference in final serum zinc concentration between groups.
f. Children who received the fortified powder were less likely to have low serum zinc concentrations (< 65 µg–dL) at the end of the study
(27.5% vs. 34.7%, p = .024).
Initial serum zinc

concentration Final serum zinc con-
Groupb (μg/dL)c centration (μg/dL)c Change in heightc Change in weightc Morbidity
MMN 102 ± 28 94 ± 22 (NS) 7.0 ± 1.4 cm/6 mo 1.3 ± 0.5 kg/6 mo (NS) No significant differ-
(NS) ence in diarrhea,
Control 101 ± 21 101 ± 28 7.0 ± 1.2 cm/6 mo 1.2 ± 0.6 kg/6 mo fever, respiratory
infection
MMN 67 ± 13 69 ± 14 (NS) 7.0 cm/6 mo (95% CI, 1.6 kg/6 mo (95% CI, —
6.7 to 7.3) (NS) 1.4 to 1.7) (NS)
Control 67 ± 14 69 ± 12 7.1 cm/6 mo (95% CI, 1.6 kg/6 mo (95% CI, —
6.9 to 7.4) 1.4 to 1.7)
MMN — — 3.8 ± 0.2 cm/3 mo 0.71 ± 0.04 kg/3 mo —
(NS) (NS)
Control — — 3.3 ± 0.14 cm/3 mo 0.75 ± 0.06 kg/3 mo —
MMN — 83 ± 6 (NS) — — —
Control — 86 ± 7 — — —
MMN — — Significantly greater Significantly greater 4.46 ± 3.8 diarrhea

changes in HAZ changes in WAZ episodes per child
(mean difference, (mean difference, (p < .05)
Control — — 0.19; 95% CI, 0.12 to 0.24; 95% CI, 0.11 to 5.36 ± 4.1 diarrhea
0.26; p < .001) 0.36; p < .001) episodes per child
MMN 79 ± 12 85 ± 9 (NS) 10.0 ± 1.5 cm/6 mo 2.1 ± 0.9 kg/6 mo (NS) —

(NS)
Control 69 ± 16 74 ± 12 10.1 ± 2.1 cm/6 mo 2.1 ± 1.2 kg/6 mo —

MMN 72 ± 28 89 ± 27 (NS) No impact on linear 12% were underweight —
Control 68 ± 26 86 ± 23 growth retardation in fortified group —
and 24% in control
group (p < .001)
MMN 67 ± 15 73 ± 11 No impact on growth: No impact on weight Reduced incidence of

(p = .011) e,f difference in means, gain: difference in respiratory-related
Control 66 ± 15 0.10 cm; 95% CI, means, −0.06; 95% illnesses (RR, 0.83;
71 ± 11
−0.05 to 0.25 (NS) CI, −0.24 to 0.13 95% CI, 0.73 to
(NS) 0.94; p < 0.01) and
diarrhea (RR, 0.38;
95% CI, 0.16 to 0.90;
p < .05)
is uncertain, but they may be related to the age of the Detailed review of evidence
subjects or to the nature of the fortification vehicles. It
is possible, for example, that inhibitors of zinc absorp- Overview
tion in the cereal-based complementary foods limited Various approaches have been suggested for house-
zinc uptake from these products and thereby prevented hold-level fortification of complementary foods, by
changes in serum zinc concentration, although this which supplemental MMN are added to these foods
would not explain the lack of impact in the Mexico just prior to consumption, so as to enable infants and
study, in which zinc was provided in milk. It is difficult young children to achieve adequate micronutrient
to interpret the functional responses to these interven- intakes [45]. Available home fortification products
tions, for a variety of reasons, as discussed above, and may be MMN powders (e.g., Sprinkles®), crushable
additional studies of the functional impact of zinc for- tablets (e.g., Foodlets®), or lipid-based nutrient sup-
tification will be needed in populations that are known plements (e.g., Nutributter®), which are added to food
to have a high risk of zinc deficiency. at the household level, usually just prior to serving.
Although each of these products contains a broad array
of micronutrients, most studies have only investigated
Section 3 the potential of these formulations to control iron defi-
ciency and anemia [46, 47]; little is known about their
Does household-level fortification with MMN (includ- impact on the status of other micronutrients. We have
ing zinc) affect indicators of zinc status and zinc-related reviewed the available literature on the impact of home
functions? fortification products on zinc-related outcomes.
Bibliographic search
Conclusion
Data sets were identified by using a computerized
There is little available information on the impact of bibliographic search (PubMed), with key words: 1)
household-level food fortification with zinc-containing zinc, Sprinkles; limiting for human; and 2) home
supplements on serum zinc concentration and zinc- fortification, zinc; limiting for human trials. A total
related functional outcomes. There is just one avail- of 17 articles were identified, of which only 1 efficacy
able efficacy trial in which children received one of trial compared the impact of MMN powders with and
two different formulations of micronutrient powders, without zinc [48]. Three additional efficacy trials com-
which did or did not contain zinc; this trial did not pared a zinc-containing MMN powder formulation
find any significant differences between treatment with a placebo-control group [49, 50] or with treatment
groups in final serum zinc concentration or change groups that received other home fortification products
in height or weight. The other four available studies [51]. In these latter three studies, any impact on growth
compared zinc-containing MMN products with a pla- or morbidity cannot be attributed specifically to zinc.
cebo or another type of household-level fortification We have further included one efficacy trial in which
product, so it is not possible to attribute any results crushable tablets containing MMN were provided in
specifically to zinc. Only one of these studies, which food. This study was part of a multicountry evaluation,
provided MMN crushable tablets, found a significant the International Research on Infant Supplementation
impact on the final serum zinc concentration, and (IRIS) trials. Only in the study included here [52] was
none of the randomized, controlled trials found any it explicitly stated that the crushable tablets were mixed
impact on growth. Nevertheless, the results of the only with complementary foods, so this trial can be con-
available evaluation of a large-scale program showed a sidered as an evaluation of household fortification. In
significant reduction in the stunting rate after 2 years addition to the articles identified in the PubMed search,
of distribution of MMN powders to young Mongolian we included an evaluation of a large-scale program
children aged 6 to 35 months compared with baseline. that was part of a comprehensive nutrition program
Although it is not possible to attribute the reduction implemented in Mongolia by World Vision and the
in stunting prevalence to zinc alone, or even to the government of Mongolia [53].
intervention per se, zinc may have contributed to the
observed reduction in stunting. Results of intervention studies of zinc-containing
On the basis of the limited available evidence, it household-level fortification products that reported
is not possible to conclude whether zinc deficiency potentially zinc-related outcomes
can be controlled through current approaches using Table 4 provides a summary of randomized, control-
household-level fortification. There is a need for led household-fortification trials that included the
further evaluation of the efficacy and effectiveness of measurement of serum zinc concentration or other
home-fortification products in zinc-deficient popula- zinc-related outcomes. Results are available from just
tions, with regard to specific zinc-related outcomes. one randomized, double-blind efficacy trial using
MMN powders in which zinc was the only nutrient
that differed between groups. This community-based lactic acid bacteria (MMN-LAB group) versus a con-
study was completed to simulate the potential impact of trol group that received a placebo and no lactic acid
a home-based fortification program in 6- to 18-month- bacteria for 2 months [50]. Children who suffered
old, moderately anemic Ghanaian children to deter- from a diarrhea episode within the past 2 weeks were
mine the effect of adding zinc to an iron-containing randomly assigned to one of the three groups (n = 25
powder [48]. The final product, which was added to per group). Fieldworkers visited homes twice weekly to
the local maize-based porridge, contained 80 mg of collect information on the number of diarrheal stools,
iron and 50 mg of ascorbic acid, with or without 10 mg days of diarrhea, dysentery, and other morbidity symp-
of zinc as zinc gluconate. At the end of the 2-month toms. After adjustment for age, the mean longitudinal
period of intervention, most children in both groups diarrhea prevalence was significantly lower in the
had recovered from anemia, although there was a multiple-micronutrients group (15 ± 10% child-days)
slightly greater percentage of children in the zinc-plus- than in the placebo group (20 ± 19% child-days) (p <
iron group who remained anemic (37.1% vs. 25.2%, .007). Surprisingly, the MMN-LAB group had a higher
p < .05). Although there were no significant differences prevalence of diarrhea (26 ± 20% child-days), which
in final mean serum zinc concentrations, significantly was not significantly different from that in the placebo
fewer children in the zinc-plus-iron group had a serum group (p = .28). These results are puzzling because
zinc concentration < 70 μg/dL than in the iron group both MMN groups received zinc, so it appears either
(22.6% vs. 36.1%, p < .05). There was no significant that the addition of lactic acid bacteria eliminated any
groupwise difference in the children’s growth. A pre- beneficial effect of zinc (or the other micronutrients)
vious meta-analysis of the effect of zinc supplementa- on diarrhea prevalence or the morbidity results were
tion on children’s growth found positive response to unrelated to the MMN supplements. Neither serum
zinc only among those studies that enrolled children zinc concentration nor final anthropometrics were
whose initial mean HAZ was less than approximately evaluated.
−1.5 [20]. Because the mean HAZ values at baseline A randomized intervention trial compared the effi-
were −1.81 and −1.70 in the two groups in the Ghana cacy and acceptability of MMN powders, crushable
study, a beneficial impact of additional zinc on the tablets, and a lipid-based nutrient supplement added
children’s growth would have been expected, although to home-prepared complementary foods in Ghana
the 2-month study period may have been too short [51]. The products differed in their energy and nutri-
to detect such an effect. Moreover, the micronutrient ent contents, but all contained iron (powder, 12.5 mg;
powders provided in this study contained a very high tablet, 9 mg; lipid-based supplement, 9 mg) and zinc
dose of iron (80 mg/sachet), which might have inter- (powder, 5 mg; tablet, 4 mg; lipid-based supplement,
fered with zinc absorption. 4 mg). At the end of the intervention, infants with the
One recent study evaluated the impact of different same eligibility criteria for the trial, but not randomly
formulations of MMN powders in a double-blind, selected for the intervention, were examined as a non-
placebo-controlled trial in Cambodian children [49]. intervention group for a cross-sectional comparison.
At 6 months of age, infants were randomly assigned The prevalence of anemia (32%; hemoglobin < 100 g/L)
to receive either MMN powders with 12.5 mg of iron, and iron deficiency (28%; ferritin < 12 µg/L) were
5 mg of zinc, 50 mg of ascorbic acid, 300 μg of vita- significantly higher in the nonintervention group at
min A, 7.5 μg of vitamin D3, and 150 μg of folic acid; 12 months than in the three treatment groups (aver-
powders with only 12.5 mg of iron and 150 μg of folic age of 15% and 10% for anemia and iron deficiency,
acid; or placebo powders for a period of 12 months. respectively) [51]. However, there was no impact of
There was no impact of either of the two types of sup- any of the three treatments on serum zinc concentra-
plement on the children’s growth, but the sample size tion compared with the nonintervention group. At 12
was just marginally adequate (n = 65 per group) to be months, the weight-for-age z-score (WAZ) and length-
able to detect such differences. Moreover, the stunting for-age z-score (LAZ) adjusted for sex and maternal
prevalence at baseline was only 6%, indicating that height were significantly greater only in the group that
the study population may not have been sufficiently received the lipid-base nutrient supplements (WAZ =
growth-restricted to benefit from zinc supplementa- −0.40 ± 1.1, LAZ = –0.14 ± 1.0) compared with the
tion. However, the stunting rate increased to 27% over group that received crushable tablets (WAZ = −0.88 ±
the study period of 12 months, and the final mean HAZ 1.1, LAZ = −0.44 ± 1.0). There was no significant dif-
of all groups was −1.55 ± 0.74. Regrettably, serum zinc ference between the lipid-based nutrient-supplements
concentration was not measured. group, the MMN-powders group (WAZ = −0.53 ± 1.1,
Another three-cell study, which was carried out LAZ = −0.40 ± 1.0) and the nonintervention group
among 6- to 12-month-old Pakistani children, investi- (WAZ = −0.74 ± 1.1, LAZ = −0.40 ± 1.0). Since all three
gated the impact of a MMN powder (including 5 mg of treatment groups received zinc, the greater growth rate
zinc) (MMN group) or the same formulation of micro- in the lipid-based nutrient-supplements group cannot
nutrients (including 5 mg of zinc) and heat-inactivated be explained by zinc alone.
TABLE 4. Efficacy trials of home fortification products that included zinc and assessed the impact on zinc-related indicators (serum zinc concentration, change in height, change S98
in weight, morbidity)
Initial
Addi- serum zinc Final
Country, year Dura- tional concentra- serum zinc
[reference] Age Sample tion Product zinc tion concentration Change in Change in
author Study design (mo) sizea (mo) Group formulation (mg/day) (μg/dL)b (μg/dL)b heightb weightb Morbidity
Ghana, 2003 Randomized, 6–18 239 2 MMN pow- 80 mg iron, 10 10 94 ± 29 87 ± 25 −0.11 ± 0.61 −0.20 ± 0.55 NA
[48]c Zlotkin double-blind, ders with mg zinc, 50 (NS)d ΔHAZ/2 mo ΔWAZ/2 mo
controlled zinc mg ascorbic (NS) (NS)
acid
MMN pow- 80 mg iron, 50 0 92 ± 29 81 ± 22 −0.05 ± 0.31 −0.15 ± 0.79 NA
ders with- mg ascorbic ΔHAZ/2 mo ΔWAZ/2 mo
out zinc acid
Cambodia, Randomized, 6 191 12 MMN 12.5 mg iron, 5 5 NA NA −0.77 ± 1.50 −0.32 ± 0.62 NA
2006 [49]c double-blind, powders mg zinc, plus ΔHAZ/12 mo ΔWAZ/12
Giovannini controlled multivitamins (NS) mo (NS)
Iron, folic 12.5 mg iron, 0 NA NA −0.65 ± 1.25 −0.20 ± 0.39 NA
acid 150 µg folic ΔHAZ/12 mo ΔWAZ/12
powders acid mo
Placebo No micronutri- 0 NA NA −0.85 ± 1.61 −0.28 ± 0.53 NA
powders ents ΔHAZ/12 mo ΔWAZ/12
mo
Pakistan, 2006 Randomized, 6–12 75 2 MMN 30 mg iron, 5 NA NA NA NA 15% ± 10%
[50] Sharieff double-blind, powders 5 mg child-days
controlled zinc, plus of diarrhea
multivitamins (p = .007)
MMN-LAB 30 mg iron, 5 NA NA NA NA 26% ± 19%
powders 5 mg zinc, child-days of
plus multi- diarrhea
vitamins and
LAB
Placebo No micronutri- 0 NA NA NA NA 26% ± 20%
powders ents child-days of
diarrhea
S. Y. Hess and K. H. Brown
Ghana, 2007 Randomized, 6 313 6 MMN 12.5 mg iron, 5 5 75 ± 33 61 ± 16 7.9 cm/6 mo 1.39 kg/6 mo No significant
[51]e Adu- only research powders mg zinc, vita- (NS)f (NS/signifi- (NS/signifi- differences
Afarwuah assistant mins A and cant)g cant)g in morbid-
blinded, C, folic acid ity between
Zinc fortification
not 4 65 ± 19 64 ± 17 7.8 cm/6 mo 1.35 kg/6 mo groups

Crush- 9 mg iron, 4
controlled
able MMN mg zinc, plus
tablets additional
micronutri-
ents
Lipid-based 9 mg iron, 4 4 67 ± 18 61 ± 17 8.3 cm/6 mo 1.57 kg/6 mo
nutrient mg zinc, plus
supplement additional
micronu-
trients and
energy (108
kcal/day)
South Africa, Randomized, 6–12 107 6 Daily MMN 10 mg iron, 10 10 71 ± 14 78 ± 16 2.4 ± 0.6 56.6 ± 25.8 No significant
2005 [52]h double-blind, crushable mg zinc, plus (signifi- mm/wk (NS) g/wk (NS) differences
Smuts controlled tablet additional cant/NS)i in morbid-
micronutri- ity between
ents groups
Weekly 20 mg iron, 20 2.9 74 ± 15 78 ± 14 2.5 ± 0.5 50.0 ± 21.3
MMN mg zinc, plus mm/wk g/wk
crushable additional
tablet micronutri-
ents
Placebo No micronutri- 0 75 ± 14 73 ± 10 2.4 ± 0.6 51.6 ± 27.7
crushable ents mm/wk g/wk
tablet
HAZ, height-for-age z-score; LAB, heat-inactivated lactic acid bacteria (Lactobacillus acidophilus); MMN, multiple micronutrients; NA, not available; NS, not significant; WAZ, weight-for-age z-score
a. Sample size includes the total number in the control and fortification groups.
b. Plus–minus values are means ± SD.
c. Change in HAZ and WAZ per group derived from baseline and final values.
d. Not significant, but incidence of low plasma zinc concentration was significantly lower in zinc group than in control group (p < .05).
e. Study also included a nonintervention group. Eligible children not randomly selected for intervention were included in the nonintervention, control group for a cross-sectional study at 12 months of
age.
f. No significant difference in final serum zinc concentrations among the three treatment groups and a nonintervention group at 12 months of age.
g. Infants in the group receiving lipid-based nutrient supplements gained significantly more weight and length than those in the group receiving crushable tablets (p < .05); the difference from the group
receiving MMN powder was not significant.
h. Results for the fourth intervention group (daily iron alone) are not presented here. Total sample size is the total number for the three groups whose results are presented here.
i. Change in serum zinc concentration was significantly different between the group receiving daily MMN and the placebo group.
S99
Efficacy trials of the crushable tablets were carried powders and other household-fortification products.
out in 6- to 12-month-old infants in four countries The only study that directly investigated the impact
[54]. Only the results of the South African study are of adding zinc to micronutrient powders did not find
presented here, since it was the only study that used the any effects on children’s growth or final serum zinc
household-fortification approach, in which the mothers concentrations. However, this study employed a formu-
were instructed to crumble and mix the tablets with lation that had a particularly high iron content (80 mg
porridge [52]. We further limit the presentation of the of iron); thus, it is not known whether different results
results to just three of the four study groups, because might occur with the usual formulation of MMN
one group received only iron. One of the three groups powder, which generally contains 12.5 mg of iron per
that are included in the present comparison received a dose. All other studies compared zinc-containing MMN
daily MMN crushable tablet (daily MMN), which con- products with a placebo, so it is not possible to attribute
tained 13 nutrients at the level of one adequate intake any effect on zinc-related outcomes specifically to zinc.
[54]. Another group received a weekly MMN crushable None of the randomized, controlled studies found a
tablet containing the same 13 nutrients, but at double positive impact on growth, although the mean initial
the dose, followed by six daily placebo tablets during HAZ ranged from −0.2 to −1.8 across studies, and only
the remaining days of the week (weekly MMN). The one was less than −1.5. This is important to emphasize,
third group received only placebo tablets. There were because the previous meta-analyses of the effect of
no differences in growth or morbidity between the sup- zinc supplementation on children’s growth found posi-
plemented groups and the placebo group during the 6 tive responses to zinc only among those studies that
months of the study. However, the stunting prevalence enrolled children whose mean initial HAZ or WAZ was
at baseline was only 10.7%, indicating that the study less than approximately –1.5 [20]. Nevertheless, distri-
population may not have been sufficiently growth- bution of MMN powders along with other nutritional
restricted to benefit from supplementation. Serum zinc interventions may have contributed to a reduced stunt-
concentration increased significantly in the daily MMN ing rate that was observed in a large-scale national pro-
group compared with the placebo group. This increase gram in young children in Mongolia. Further research
was not seen in the weekly MMN group, suggesting is urgently needed, as several countries are considering
that a weekly dose of 20 mg of zinc may not have been the distribution of MMN powders to young children,
sufficient or that the serum zinc concentration returned both to prevent anemia and to address a broad range
to baseline levels during the interval between the last of other potential nutrient deficiencies.
weekly dose and the time of specimen collection.
The effectiveness of a large-scale distribution pro-
gram with MMN powders has been evaluated as part of Section 4
a comprehensive nutrition intervention implemented
in Mongolia by World Vision and the government of Are there any adverse clinical effects of zinc fortification
Mongolia [53]. The program included distribution of or negative effects of zinc fortification on the utilization
MMN powders (40 mg of iron, 10 mg of zinc, 400 IU of other nutrients due to zinc fortification?
of vitamin D, 600 IU of vitamin A, 50 mg of vitamin
C, and 150 μg of folic acid) free of charge to more than Conclusion
14,000 children 6 to 35 months of age. Other program
components were provision of iron syrup to anemic With very few exceptions, most of the available stud-
children 36 to 59 months of age and iron–folate tablets ies indicate that there are no adverse effects of zinc
to pregnant and lactating women, community-based fortification on iron absorption or indicators of iron
nutrition education, and social marketing. A survey and copper status.
was conducted at baseline and approximately 2 years
after program implementation. Eighty-eight percent Detailed review of evidence
of eligible children in the intervention areas received
MMN powders for a mean duration of 13 months. The Overview
prevalence of anemia in children 6 to 35 months of age Currently available evidence concerning adverse
decreased from 55% at baseline to 33% at follow-up. effects of zinc has been derived entirely from zinc
The prevalence of stunting was significantly reduced supplementation trials [55]. As described in detail in
from 23% at baseline to 18% at follow-up (p < .01), and the previous paper on preventive zinc supplementa-
the prevalence of wasting remained low (1% at baseline, tion [56], overt toxicity symptoms, such as nausea and
1.5% at follow-up). Because of the pre-post design of vomiting, occur in adults only at high zinc intake levels
this intervention, it is not possible to attribute the results of 150 mg/day or more. Similarly, the adverse effect of
to the intervention in general or specifically to zinc. zinc on copper metabolism, as measured by a decrease
In summary, there is very limited information avail- in erythrocyte superoxide dismutase activity, has been
able on the specific impact of zinc included in MMN found only at zinc intakes greater than 50 mg/day in
adults [55]. Some studies in young children that pro- Contrary to these latter results, López de Romaña et
vided 10 mg of supplemental zinc daily, with or with- al. (unpublished) found no significant differences in
out iron, found that zinc supplementation had a small iron absorption when a total of 0, 3, or 9 mg of zinc,
negative effect on iron status in some cases [57, 58], as zinc sulfate, was added to iron-fortified wheat flour
although no overall effect was found in a recent meta- products consumed during two meals. Mendoza et al.
analysis [56]. However, little is known about whether [11] also found no differences in iron absorption from
any adverse effects may occur with zinc fortification. a wheat-soy-milk porridge to which zinc was added as
The tolerable upper intake level (UL) is defined as either zinc sulfate or zinc methionine in adults. Thus,
the highest average level of daily intake that is likely except for the results of the Indonesian study, and only
to pose no risk of adverse health effects for almost all when zinc was provided as zinc sulfate in that study,
persons in the general population. The UL for zinc the other available results indicate no adverse effects of
has been discussed in detail previously [55, 59]. It is zinc fortification on iron absorption from cofortified
worth emphasizing that the currently recommended foods, regardless of the chemical form of the zinc used
UL is based on the amount of total zinc intake per day for fortification. In summary, the weight of current
and does not consider the bioavailability of the zinc. In evidence suggests that zinc fortification should not
unrefined, cereal-based diets, for example, the amount adversely affect iron absorption from foods cofortified
of zinc absorbed is estimated to be lower than that for with both zinc and iron.
more refined foods because of the inhibiting effects
of phytic acid present in the whole-grain products Intervention studies with zinc-fortified foods
[55]. Thus, the suggested ULs for zinc may need to be Effect of zinc fortification on indicators of iron status.
reevaluated, taking into consideration the bioavail- Only two of the six studies of zinc-fortified milk
ability of zinc from foods, including zinc-fortified products presented information on final hemoglobin
foods. Notably, Arsenault and Brown [60] reanalyzed concentration [29, 34], and one assessed serum ferritin
the data of 7,474 nonbreastfeeding preschool children concentration [34]. In both studies, the treatment group
included in the US Continuing Survey of Food Intakes received not only additional zinc, but also additional
by Individuals and found that overall 36% of children iron, as compared with the control group. Neither of
had zinc intakes that were in excess of the UL. Despite these studies found any difference in iron status-related
these high intakes, there have been no recent reports outcomes between the two groups (table 1).
of zinc toxicity in US children. Of the five trials of zinc-fortified cereal products
To explore for possible adverse effects of zinc for- (table 2), three reported measurements of hemoglobin
tification, we reviewed the data available from tracer concentration [6, 18, 22] and four provided results for
studies and intervention trials, as described above, serum ferritin concentration [6, 18, 22, 36]. None of
to evaluate the impact of zinc fortification on iron these studies found a significant difference in final
absorption and indicators of iron and copper status. hemoglobin concentration or serum ferritin concentra-
For the identification of possible adverse effects of zinc tion between the zinc-fortified group and the control
fortification on iron status, zinc must be the only factor group. With the exception of the study by Kilic et al.
that differs between the study groups, so the following [36], all of the fortified food products provided to both
review is limited to the subset of relevant studies. groups in these trials contained additional iron. On
the basis of the current evidence, it can be concluded
Tracer studies of the effect of zinc fortification on that the addition of zinc to food fortification formula-
absorption of nutrients tions does not have any adverse effect on iron status
Effect of zinc fortification on iron absorption. The effect indicators.
of zinc fortification on iron absorption from foods Effect of zinc fortification on indicators of copper
cofortified with both zinc and iron has been examined status. Four of the studies of zinc-fortified milks pre-
in several studies. In the study of Sri Lankan schoolchil- sented data on serum copper concentration [30–32, 34]
dren, there was no difference in iron absorption from (table 1). One study is not included in the following
the iron-fortified rice product when 1.5 mg of zinc summary because the treatment group received both
was added as zinc oxide [18]. Similarly, in the study of zinc and copper in infant formula [29]. None of the
Indonesian children who received iron-fortified wheat studies found a significant difference in final mean
dumplings with or without added zinc (1.5 mg of zinc serum copper concentrations. Three of the studies of
added to a 25-g portion of dumplings and tomato puree zinc-fortified cereal products provided information
that contained ~0.25 mg of zinc), there was no effect of on the mean final serum copper concentrations [22,
zinc fortification on iron absorption when the added 35, 36] (table 2). None of these studies detected any
zinc was provided as zinc oxide [9]. On the other hand, impact of zinc fortification on copper status. Thus,
there was a significant 28% decrease in iron absorption there is currently no evidence for an adverse effect of
when the zinc fortificant was provided as zinc sulfate. zinc fortification on copper status.
Section 5 fortification products can exert a positive impact on

zinc-related outcomes in young children also remains
What are the steps in implementing zinc fortification to be proven. Thus, additional research is needed
programs and what additional research is needed? to determine the efficacy and effectiveness of these
intervention strategies in different populations and
The foregoing review of available scientific evidence using graded dosages of zinc. When such fortification
indicates that zinc fortification programs have the programs are implemented, it is highly recommended
potential of increasing zinc intake and total zinc that their impact on children’s zinc nutrition and health
absorption in targeted individuals. Although there is should be rigorously evaluated.
still only limited information available regarding the Although there is very little relevant information,
efficacy and effectiveness of zinc fortification, several mass fortification has the potential to benefit other seg-
studies have found a positive impact on serum zinc ments of the population, such as school-aged children
concentration or potentially zinc-related functional [44], adolescents, women of reproductive age, and pos-
responses with particular zinc-fortified food products sibly the elderly. In any case, further research is needed
in selected age groups. Moreover, studies indicate that to evaluate the efficacy and effectiveness of fortification
there is negligible risk and relatively low cost associated programs in all age groups.
with zinc fortification, so such programs are recom-
mended for implementation in high-risk populations Assessing zinc intake and status in the target
that have appropriate conditions for successful food population
fortification programs.
The following section will review briefly some of the WHO recommends gathering data on micronutrient
critical steps required for a zinc fortification program deficiencies in a representative sample of the target
and summarize pending research issues. Program- population to justify the needs for a fortification pro-
related considerations have been reviewed previously gram [1]. WHO, the United Nations Children’s Fund
by WHO [1] and the International Zinc Nutrition (UNICEF), the International Atomic Energy Agency
Consultative Group (IZiNCG) [55], so these will only (IAEA), and IZiNCG jointly recommend three indica-
be summarized in this section. The major issues are: tors for assessing the population’s risk of zinc deficiency
» Defining the target population [62]. The concentration of zinc in blood plasma or
» Assessing zinc intake and status in the target serum is the best available biomarker of zinc status
population in populations. Chronic inadequate dietary intake of
» Selecting the food vehicle(s) zinc is the most likely cause of zinc deficiency. Hence,
» Selecting the zinc fortificant estimating the adequacy of zinc intakes through
» Determining the level of zinc fortificant quantitative dietary intake surveys is useful to evaluate
» Establishing the regulatory parameters the risk of zinc deficiency in populations. Moreover,
» Costs information on dietary intake is needed to determine
» Monitoring and evaluation the appropriate food vehicles and level of fortifica-
tion, as discussed below. A third possible indicator of
Identifying the target population population zinc status is height-for-age, a measure of
nutritional stunting usually applied among children
The most vulnerable population groups for zinc defi- less than 5 years of age. Stunting is the best known and
ciency are infants, young children, and pregnant and easiest to measure of the adverse outcomes associated
lactating women because of their elevated requirements with zinc deficiency, and relevant data are often already
for zinc [55]. All of these groups probably need special available from previous national health or nutritional
attention, and food fortification products should be status surveys [63]. Stunting prevalence is expressed as
chosen specifically to target these population groups the percentage of children less than 5 years of age with
and their age-specific requirements [61]. As there is height-for-age below the expected range of a reference
currently no information available on the impact of population (i.e., less than −2.0 SD with respect to the
zinc fortification among pregnant and lactating women reference median), such as the recently established
and their offspring, research is needed to assess such WHO Growth Standards [64].
programs. The risk of zinc deficiency is considered to be
In the case of young children, as described above, elevated and of public health concern when the preva-
studies show that zinc fortification can increase dietary lence of low serum zinc concentrations is greater than
intake and total absorption of zinc. However, studies 20% or the prevalence of inadequate intakes is greater
have not yet demonstrated that zinc-fortified comple- than 25%. A stunting prevalence greater than 20% is
mentary foods can positively affect indicators of young also considered to be indicative of an increased risk of
children’s zinc status, growth, or other potentially zinc- zinc deficiency. However, since zinc deficiency is not
related functional outcomes. Whether household-level the only factor affecting children’s growth, assessment
of serum zinc concentration and dietary zinc intake zinc sulfate or zinc oxide, as described above. Thus,
should be used to confirm the risk of zinc deficiency zinc fortification programs will generally rely on zinc
in countries with high rates of stunting. oxide, because of its lower cost.
Selecting the food vehicle(s) Determining the level of zinc fortificant
The selection of food vehicles for fortification pro- The dietary goal of fortification is formally defined by
grams depends on the target population groups, their WHO as the provision of most (97.5%) individuals in
dietary pattern, and locally available food-production the population group(s) at greatest risk of deficiency
capability. The assessment of food-intake data not only with an adequate intake of specific micronutrients,
can be used to assess the risk of zinc deficiency but also without causing a risk of excessive intakes in these or
is necessary to make an informed judgment regarding other groups [1]. The WHO guidelines on food forti-
which food(s) would be the most suitable vehicle(s) for fication describe a detailed procedure to estimate the
fortification [1]. In the case of young children, the most appropriate micronutrient level for mass fortification
suitable products could be processed complementary programs, using the Estimated Average Requirement
foods or household-level fortification products, such (EAR) cutpoint method and the concept of the Fea-
as powders (e.g., Sprinkles®), crushable tablets (e.g., sible Fortification Level (FFL) [1]. Recommendations
Foodlet®), and lipid-based nutrient supplements (e.g., for zinc fortification of cereal flour have recently
Nutributter®). If a fortification program is targeting been revised on the basis of the population’s usual
schoolchildren, other products, such as condiments, flour intakes, degree of milling (level of extraction),
snacks, or beverages, may be more suitable, depend- and intakes of zinc and phytate from other dietary
ing on the local food habits. For mass fortification, the sources [66]. A summary table from that document is
selected food vehicle should be consumed regularly by reproduced herein, for 85% extraction wheat flour and
a large segment of the general public in fairly constant different assumptions regarding the average amounts
amounts throughout the year. Typical products chosen of flour intake and zinc intake from other sources
for mass fortification are cereals, vegetable oils, fats, (table 5). For more details on other levels of extraction
milk, and condiments [1]. and the basis for these recommendations, the original
The main advantage of mass fortification in com- document should be reviewed [66].
parison with supplementation or targeted fortification The amount of a particular micronutrient that can be
is that it requires very little investment because it uses added to foods is limited by various safety, technologic,
already existing products and delivery mechanisms. and economic constraints, which are all considered in
To keep the fortification program cost-effective and the FFL. In other words, the FFL is the maximum con-
efficient, the fortification technology should be sup- tent of micronutrient that can be added and still remain
ported by formal, centralized industries. In general, compatible with the food matrix, while the increase in
the low cost of mass fortification can only be realized food price stays within an acceptable range. The FFL
in industrialized settings, where only a few technically provides the greatest number of at-risk individuals with
advanced factories produce the foods. The challenge in an adequate intake without causing an unacceptable
developing countries is to find a suitable food vehicle risk of excessive intake in the population at large. As a
that is industrially produced and widely consumed in consequence, the maximum content of micronutrients
reasonable amounts by the target population and, ide- in mass fortification programs is determined by those
ally, with a narrow range of inter- and intraindividual individuals who eat the food vehicles in the largest
variation in intake [65]. quantities (i.e., the upper 5% to 10%) [2].
In the case of zinc, experience has shown that the
Selecting the zinc fortificant addition of zinc compounds such as zinc oxide and
zinc sulfate to wheat and maize flour at recommended
Several zinc compounds are listed by the US Food and levels is technically feasible and does not alter the
Drug Administration as generally recognized as safe organoleptic qualities of the fortified foods [67]. The
(GRAS) and can be used for fortification. As described national wheat and maize flour fortification programs
previously by IZiNCG, there is no consensus as to in Mexico and South Africa are examples in which zinc
which of the GRAS compounds is the most appropri- is included in the fortification formulation.
ate for fortification programs [55]. Zinc oxide and zinc A more critical limiting factor may be safety con-
sulfate are the GRAS salts that are least expensive and straints with regard to zinc fortification. Most indi-
most commonly used by the food industry. Despite viduals receiving food products should keep the total
theoretical considerations that suggest that zinc may be intake (from the natural content of foods, fortified
better absorbed from water-soluble compounds such as foods, and supplements) of certain nutrients lower than
zinc sulfate, several studies indicate that zinc is equally the levels that are recognized as safe (Tolerable Upper
well absorbed from cereal products fortified with either Intake Levels, ULs) [55, 59]. In mass fortification, adult
TABLE 5. Amount of zinc fortification of wheat flour (mg of zinc/kg flour) that is necessary to ensure 3.84 mg of absorbed
zinc/day, considering different amounts of usual zinc and phytate intakes from sources other than wheat flour and the stated
amounts of flour consumption (80% extraction flour)a
Dietary phytate (mg/day)
Dietary zinc from sources other Dietary zinc from sources other Dietary zinc from sources other
Flour
than wheat: 3 mg/day than wheat: 5 mg/day than wheat: 7 mg/day
intake
(g/day) 0 500 100 0 500 1,000 0 500 1,000
50 134 229 325 94 189 285 54 149 245
75 96 160 223 69 133 197 42 106 170
100 77 124 172 57 104 152 37 84 132
200 48 72 96 38 62 86 28 52 76
300 38 54 70 32 48 64 25 41 57
400 34 46 58 29 41 53 24 36 48
500 31 40 50 27 36 46 23 32 42
600 29 37 45 26 34 42 22 30 38
700 28 34 41 25 32 38 22 29 36
800 27 33 39 24 30 36 22 28 34
Source: reproduced from Brown et al. [66].
a. Shaded values indicate fortification levels > 100 mg of zinc/kg flour, which is the upper range of zinc fortification for which sensory accept-
ability has been confirmed.
males, who tend to have the highest levels of consump- operating fortification process, which is ~12%. In this
tion of staple foods, have the greatest risk of excessive case, the minimum and maximum production param-
micronutrient intakes. To avoid excessive intakes in this eter for zinc results in a range of 30 to 50 mg of zinc
population group, the level of the fortificant may need per kilogram of flour. As the loss of zinc during storage
to be kept so low that the most vulnerable population and distribution is expected to be minimal, the LML is
groups, such as young children, may not benefit suf- equal to the minimum production factor (30 mg/kg),
ficiently from a mass fortification program. In that case, and the MTL is equal to the maximum production
fortification of several food vehicles, targeted fortifica- factor (50 mg/kg) [2].
tion, or zinc supplementation should be considered
for completing the dietary or nutritional gap of those Costs
individuals with low consumption of the fortified food
or larger nutritional needs. Estimating costs is an important step in planning a
food fortification program. Estimates must include
Establishing the regulatory parameters both the costs to industry (e.g., capital investment and
recurrent costs, such as the purchase of fortificant)
Regardless of whether the fortification intervention as well as the public sector costs (e.g., enforcement,
is voluntary or mandatory, suitable standards should monitoring, and evaluation). In the case of mass for-
be established to allow appropriate quality control by tification programs, which tend to rely on staples and
the producers and inspection and enforcement by the condiments as the food vehicles, cost is often the most
public sector. Specific levels of the added micronutrients significant limiting factor. Staples and condiments are
should be stipulated, and the government authorities consumed frequently and in large amounts, not only by
should check on compliance based on those levels [65]. the population directly, but also by the food industry.
Two regulatory parameters are essential: the legal mini- Even small variations in price can thus have profound
mum level (LML) for all nutrients and the maximum consequences, and experience has shown that mass
tolerable level (MTL) for those nutrients for which fortification in an open market economy is most likely
excess consumption is of concern (including zinc). to be successful when the increase in the price of the
If, for example, under the circumstances of a par- fortified product, relative to the unfortified one, does
ticular program in which low-extraction wheat flour not exceed 1% to 2% [1].
has an intrinsic zinc content of 10 mg/kg [2] and a However, the addition of zinc adds very little to the
zinc fortification level of 30 mg/kg is proposed, the fortification cost and hence, for this nutrient, cost is
average final zinc content in the flour is calculated as not a restrictive factor. In a recent review of the cur-
the sum of these two values, resulting in a value of 40 rent costs to supply the EAR of nutrients to women
mg/kg. An acceptable range of final zinc content is of reproductive age through food fortification, it
determined by subtracting and adding two times the was concluded that zinc (as zinc oxide) is one of the
expected coefficients of variation of a satisfactorily least expensive nutrients [2]. Taking the expected
micronutrient loss during production, distribution, is usually conducted by regulatory authorities as well
storage, and food preparation into account, it is esti- as by the producers themselves as part of their inter-
mated that a woman can receive her entire yearly nal quality-control activities. Additional monitoring
requirement of zinc through food fortification at an activities should also assess whether or not a program
annual cost of US$0.006 to US$0.013. is providing appropriately fortified food products in
sufficient amounts and at affordable prices to the target
Monitoring and evaluation population and whether these products are consumed
at the household level.
Monitoring and evaluation are essential components The term “evaluation” refers to the assessment of the
of any food fortification program, and detailed guide- impact of a program on the nutritional and health status
lines are provided elsewhere [1]. In the context of of the target population. During a program evaluation,
food fortification, the term “monitoring” refers to the zinc intake, serum zinc concentrations, and functional
continuous collection, review, and use of informa- outcomes, such as stunting prevalence, should be
tion on program-implementation activities, for the assessed in a representative sample of the target popula-
purpose of identifying problems and redirecting the tion [62]. WHO recommends that program evaluation
program as needed. The ultimate purpose of monitor- should only be undertaken once appropriate moni-
ing a fortification program is to ensure that the forti- toring has established that the fortification program
fied product, of the desired quality, is made available has been implemented as planned and is operating
and is accessible to consumers in sufficient amounts efficiently [1]. There is currently a lack of information
[1]. Regulatory monitoring includes all monitoring on the effectiveness of zinc fortification programs, and
activities at the production level, as well as monitoring program evaluation is highly encouraged.
at customs warehouses and retail stores. This activity
References
1. World Health Organization/Food and Agriculture 9. Herman S, Griffin IJ, Suwarti S, Ernawati F, Permaesih
Organization. Guidelines on food fortification with D, Pambudi D, Abrams SA. Cofortification of iron-
micronutrients. Allen L, de Benoist B, Dary O, Hurrell fortified flour with zinc sulfate, but not zinc oxide,
R, eds. Geneva: WHO, 2006. decreases iron absorption in Indonesian children. Am J
2. Dary O. The importance and limitations of food forti- Clin Nutr 2002;76:813–7.
fication for the management of nutritional anemias. In: 10. Hotz C, DeHaene J, Woodhouse LR, Villalpando S,
Kraemer K, Zimmermann MB, ed. Nutritional anemia. Rivera JA, King JC. Zinc absorption from zinc oxide,
Basel, Switzerland: Sight and Life Press, 2007:315–36. zinc sulfate, zinc oxide + EDTA, or sodium-zinc EDTA
3. Brown KH, Wessells KR, Hess SY. Zinc bioavailability does not differ when added as fortificants to maize
from zinc-fortified foods. Int J Vitam Nutr Res 2007; tortillas. J Nutr 2005;135:1102–5.
77:174–81. 11. Mendoza C, Peerson JM, Brown KH, Lonnerdal B. Effect
4. Hansen M, Samman S, Madsen LT, Jensen M, Sorensen of a micronutrient fortificant mixture and 2 amounts of
SS, Sandstrom B. Folic acid enrichment of bread does calcium on iron and zinc absorption from a processed
not appear to affect zinc absorption in young women. food supplement. Am J Clin Nutr 2004;79:244–50.
Am J Clin Nutr 2001;74:125–9. 12. Fredlund K, Isaksson M, Rossander-Hulthen L, Almgren
5. Sandström B, Arvidsson B, Cederblad A, Bjorn-Ras- A, Sandberg AS. Absorption of zinc and retention of
mussen E. Zinc absorption from composite meals. I. calcium: Dose-dependent inhibition by phytate. J Trace
The significance of wheat extraction rate, zinc, calcium, Elem Med Biol 2006;20:49–57.
and protein content in meals based on bread. Am J Clin 13. Avalos Mishaan AM, Zavaleta N, Griffin IJ, Hilmers DC,
Nutr 1980;33:739–45. Hawthorne KM, Abrams SA. Bioavailability of iron and
6. López de Romaña D, Salazar M, Hambidge KM, Penny zinc from a multiple micronutrient-fortified beverage. J
ME, Peerson JM, Krebs NF, Brown KH. Longitudinal Pediatr 2004;145:26–31.
measurements of zinc absorption in Peruvian children 14. Layrisse M, Martinez-Torres C. Fe(III)-EDTA complex
consuming wheat products fortified with iron only or as iron fortification. Am J Clin Nutr 1977;30:1166–74.
iron and 1 of 2 amounts of zinc. Am J Clin Nutr 2005; 15. Hurrell RF, Reddy MB, Burri J, Cook JD. An evaluation
81:637–47. of EDTA compounds for iron fortification of cereal-
7. Tran CD, Miller LV, Krebs NF, Lei S, Hambidge KM. based foods. Br J Nutr 2000;84:903–10.
Zinc absorption as a function of the dose of zinc sulfate 16. Davidsson L, Dimitriou T, Boy E, Walczyk T, Hurrell
in aqueous solution. Am J Clin Nutr 2004;80:1570–3. RF. Iron bioavailability from iron-fortified Guatemalan
8. López de Romaña D, Lonnerdal B, Brown KH. Absorp- meals based on corn tortillas and black bean paste. Am
tion of zinc from wheat products fortified with iron J Clin Nutr 2002;75:535–9.
and either zinc sulfate or zinc oxide. Am J Clin Nutr 17. Mendoza C, Viteri FE, Lönnerdal B, Raboy V, Young KA,
2003;78:279–83. Brown KH. Absorption of iron from unmodified maize
and genetically altered, low-phytate maize fortified with 31. Matsuda I, Higashi A, Ikeda T, Uehara I, Kuroki Y.
ferrous sulfate or sodium iron EDTA. Am J Clin Nutr Effects of zinc and copper content of formulas on growth
2001;73:80–5. and on the concentration of zinc and copper in serum
18. Hettiarachchi M, Hilmers DC, Liyanage C, Abrams SA. and hair. J Pediatr Gastroenterol Nutr 1984;3:421–5.
Na2EDTA enhances the absorption of iron and zinc 32. Haschke F, Singer P, Baumgartner D, Steffan I, Schilling
from fortified rice flour in Sri Lankan children. J Nutr R, Lothaller H. Growth, zinc and copper nutritional
2004;134:3031–6. status of male premature infants with different zinc
19. Hess SY, Peerson JM, King JC, Brown KH. Use of serum intake. Ann Nutr Metab 1985;29:95–102.
zinc concentration as an indicator of population zinc 33. Friel JK, Andrews WL, Matthew JD, Long DR, Cornel
status. Food Nutr Bull 2007;28:S403–29. AM, Cox M, McKim E, Zerbe GO. Zinc supplementa-
20. Brown KH, Peerson JM, Rivera J, Allen LH. Effect of tion in very-low-birth-weight infants. J Pediatr Gastro-
supplemental zinc on the growth and serum zinc con- enterol Nutr 1993;17:97–104.
centrations of prepubertal children: A meta-analysis 34. Schlesinger L, Arevalo M, Arredondo S, Diaz M,
of randomized controlled trials. Am J Clin Nutr 2002; Lonnerdal B, Stekel A. Effect of a zinc-fortified formula
75:1062–71. on immunocompetence and growth of malnourished
21. Zinc Investigators’ Collaborative Group, Bhutta ZA, infants. Am J Clin Nutr 1992;56:491–8.
Black RE, Brown KH, Gardner JM, Gore S, Hidayat A, 35. Hambidge KM, Chavez MN, Brown RM, Walravens PA.
Khatun F, Martorell R, Ninh NX, Penny ME, Rosado Zinc nutritional status of young middle-income children
JL, Roy SK, Ruel M, Sazawal S, Shankar A. Prevention and effects of consuming zinc-fortified breakfast cereals.
of diarrhea and pneumonia by zinc supplementation in Am J Clin Nutr 1979;32:2532–9.
children in developing countries: Pooled analysis of ran- 36. Kilic I, Ozalp I, Coskun T, Tokatli A, Emre S, Saldamli
domized controlled trials. J Pediatr 1999;135:689–97. I, Koksel H, Ozboy O. The effect of zinc-supplemented
22. Brown KH, López de Romaña D, Arsenault JE, Peer- bread consumption on school children with asymp-
son JM, Penny ME. Comparison of the effects of zinc tomatic zinc deficiency. J Pediatr Gastroenterol Nutr
delivered in a fortified food or a liquid supplement 1998;26:167–71.
on the growth, morbidity, and plasma zinc concen- 37. Lartey A, Manu A, Brown KH, Peerson JM, Dewey KG.
trations of young Peruvian children. Am J Clin Nutr A randomized, community-based trial of the effects of
2007;85:538–47. improved, centrally processed complementary foods
23. Lutter CK, Rodriguez A, Fuenmayor G, Avila L, Sem- on growth and micronutrient status of Ghanaian
pertegui F, Escobar J. Growth and micronutrient status infants from 6 to 12 mo of age. Am J Clin Nutr 1999;
in children receiving a fortified complementary food. J 70:391–404.
Nutr 2008;138:379–88. 38. Faber M, Kvalsvig JD, Lombard CJ, Benade AJ. Effect of
24. Manger MS, McKenzie JE, Winichagoon P, Gray A, a fortified maize-meal porridge on anemia, micronutri-
Chavasit V, Pongcharoen T, Gowachirapant S, Ryan B, ent status, and motor development of infants. Am J Clin
Wasantwisut E, Gibson RS. A micronutrient-fortified Nutr 2005;82:1032–9.
seasoning powder reduces morbidity and improves 39. Liu DS, Bates CJ, Yin TA, Wang XB, Lu CQ. Nutritional
short-term cognitive function, but has no effect on efficacy of a fortified weaning rusk in a rural area near
anthropometric measures in primary school children Beijing. Am J Clin Nutr 1993;57:506–11.
in northeast Thailand: A randomized controlled trial. 40. Sazawal S, Dhingra U, Dhingra P, Hiremath G, Kumar
Am J Clin Nutr 2008;87:1715–22. J, Sarkar A, Menon VP, Black RE. Effects of fortified
25. Villalpando S, Shamah T, Rivera JA, Lara Y, Monterrubio milk on morbidity in young children in north India:
E. Fortifying milk with ferrous gluconate and zinc oxide Community based, randomised, double masked placebo
in a public nutrition program reduced the prevalence of controlled trial. BMJ 2007;334:140–4.
anemia in toddlers. J Nutr 2006;136:2633–7. 41. Juyal R, Osmamy M, Black RE, Dhingra U, Sarkar A,
26. Krondl M, Coleman PH, Bradley CL, Lau D, Ryan N. Dhingra P, Verma P, Marwah D, Saxsena R, Menon VP,
Subjectively healthy elderly consuming a liquid nutrition Sazawal S. Efficacy of micronutrient fortification of
supplement maintained body mass index and improved milk on morbidity in pre-school children and growth
some nutritional parameters and perceived well-being. - a double blind randomised controlled trial. Asia Pac J
J Am Diet Assoc 1999;99:1542–8. Clin Nutr 2004;13:S44.
27. Bunker VW, Stansfield MF, Deacon-Smith R, Marzil RA, 42. Oelofse A, Van Raaij JM, Benade AJ, Dhansay MA,
Hounslow A, Clayton BE. Dietary supplementation and Tolboom JJ, Hautvast JG. The effect of a micronutrient-
immunocompetence in housebound elderly subjects. Br fortified complementary food on micronutrient status,
J Biomed Sci 1994;51:128–35. growth and development of 6- to 12-month-old disad-
28. Salmenperä L, Perheentupa J, Pakarinen P, Siimes MA. vantaged urban South African infants. Int J Food Sci
Zinc supplementation of infant formula. Am J Clin Nutr Nutr 2003;54:399–407.
1994;59:985–9. 43. Abrams SA, Mushi A, Hilmers DC, Griffin IJ, Davila P,
29. Diaz-Gomez NM, Domenech E, Barroso F, Castells S, Allen L. A multinutrient-fortified beverage enhances
Cortabarria C, Jimenez A. The effect of zinc supplemen- the nutritional status of children in Botswana. J Nutr
tation on linear growth, body composition, and growth 2003;133:1834–40.
factors in preterm infants. Pediatrics 2003;111:1002–9. 44. Winichagoon P, McKenzie JE, Chavasit V, Pongcharoen
30. Walravens PA, Hambidge KM. Growth of infants fed a T, Gowachirapant S, Boonpraderm A, Manger MS,
zinc supplemented formula. Am J Clin Nutr 1976;29: Bailey KB, Wasantwisut E, Gibson RS. A multimi-
1114–21. cronutrient-fortified seasoning powder enhances the
hemoglobin, zinc, and iodine status of primary school Wasantwisut E, Hotz C. Assessment of the risk of zinc
children in North East Thailand: A randomized control- deficiency in populations and options for its control.
led trial of efficacy. J Nutr 2006;136:1617–23. Food Nutr Bull 2004;25(1 suppl 2):S99–203.
45. Nestel P, Briend A, de Benoist B, Decker E, Ferguson 56. Brown KH, Peerson JM, Baker S, Hess SY. Preventive
E, Fontaine O, Micardi A, Nalubola R. Complementary zinc supplementation among infants, preschoolers,
food supplements to achieve micronutrient adequacy and older prepubertal children. Food Nutr Bull 2009;
for infants and young children. J Pediatr Gastroenterol 30:S12–40.
Nutr 2003;36:316–28. 57. Dijkhuizen MA, Wieringa FT, West CE, Martuti S,
46. Zlotkin SH, Christofides AL, Hyder SM, Schauer CS, Muhilal. Effects of iron and zinc supplementation in
Tondeur MC, Sharieff W. Controlling iron deficiency Indonesian infants on micronutrient status and growth.
anemia through the use of home-fortified complemen- J Nutr 2001;131:2860–5.
tary foods. Indian J Pediatr 2004;71:1015–9. 58. Lind T, Lonnerdal B, Stenlund H, Ismail D, Seswand-
47. Zlotkin SH, Schauer C, Christofides A, Sharieff W, hana R, Ekstrom EC, Persson LA. A community-based
Tondeur MC, Ziauddin Hyder SM. Micronutrient randomized controlled trial of iron and zinc supplemen-
sprinkles to control childhood anaemia. PLOS Medicine tation in Indonesian infants: Interactions between iron
2005;2:e10. and zinc. Am J Clin Nutr 2003;77:883–90.
48. Zlotkin S, Arthur P, Schauer C, Antwi KY, Yeung G, 59. US Institute of Medicine. Dietary reference intakes for
Piekarz A. Home fortification with iron and zinc sprin- vitamin A, vitamin K, arsenic, boron, chromium, iodine,
kles or iron sprinkles alone successfully treats anemia in iron, manganese, molybdenum, nickel, silicon, vana-
infants and young children. J Nutr 2003;133:1075–80. dium, and zinc, Washington, DC: National Academy
49. Giovannini M, Sala D, Usuelli M, Livio L, Francescato Press, 2001.
G, Braga M, Radaelli G, Riva E. Double-blind, placebo- 60. Arsenault JE, Brown KH. Zinc intake of US preschool
controlled trial comparing effects of supplementation children exceeds new dietary reference intakes. Am J
with two different combinations of micronutrients Clin Nutr 2003;78:1011–7.
delivered as Sprinkles on growth, anemia, and iron defi- 61. Dewey KG, Brown KH. Update on technical issues
ciency in Cambodian infants. J Pediatr Gastroenterol concerning complementary feeding of young children in
Nutr 2006;42:306–12. developing countries and implications for intervention
50. Sharieff W, Bhutta ZA, Schauer C, Tomlinson G, Zlotkin programs. Food Nutr Bull 2003;24:5–28.
S. Micronutrients (including zinc) reduce diarrhoea in 62. de Benoist B, Darnton-Hill I, Davidsson L, Fontaine O,
children: The Pakistan sprinkles diarrhoea study. Arch Hotz C. Conclusions of the joint WHO/UNICEF/IAEA/
Dis Child 2006;91:573–9. IZiNCG interagency meeting on zinc status indicators.
51. Adu-Afarwuah S, Lartey A, Brown KH, Zlotkin S, Briend Food Nutr Bull 2007;28:S480–S79.
A, Dewey KG. Randomized comparison of 3 types of 63. World Health Organization. WHO statistical informa-
micronutrient supplements for home fortification of tion system (WHOSIS). Includes data from “World
complementary foods in Ghana: Effects on growth and health statistics 2006” and “The world health report
motor development. Am J Clin Nutr 2007;86:412–20. 2006 edition.” Geneva: WHO, 2006.
52. Smuts CM, Dhansay MA, Faber M, van Stuijvenberg 64. World Health Organization. WHO growth standards:
ME, Swanevelder S, Gross R, Benade AJ. Efficacy of length/height-for-age, weight-for-age, weight-for-length,
multiple micronutrient supplementation for improv- weight-for-height and body mass index-for-age: meth-
ing anemia, micronutrient status, and growth in South ods and development. Geneva: WHO, 2006.
African infants. J Nutr 2005;135:653S–9S. 65. Dary O. Establishing safe and potentially efficacious
53. World Vision Mongolia. Effectiveness of home-based fortification contents for folic acid and vitamin B12.
fortification with Sprinkles in an integrated nutrition Food Nutr Bull 2008;29:S214–24.
program to address rickets and anemia. Ulaanbaatar: 66. Brown KH, Hambidge KM, Ranum P, Tyler V, and the
World Vision Mongolia, 2005. Zinc Fortification Working Group. Zinc fortification of
54. Gross R, Benade S, Lopez G. The International Research cereal flours: Current recommendations and research
on Infant Supplementation initiative. J Nutr 2005; needs. Food Nutr Bull 2009; in press.
135:628S–30S. 67. López de Romaña D, Brown KH, Guinard JX. Sensory
55. International Zinc Nutrition Consultative Group trials to assess the acceptability of zinc fortificants
(IZiNCG), Brown KH, Rivera JA, Bhutta Z, Gibson added to iron-fortified wheat products. J Food Sci 2002;
RS, King JC, Lönnerdal B, Ruel MT, Sandtröm B, 67:461–5.
A review of interventions based on dietary diversifica-
tion or modification strategies with the potential to
enhance intakes of total and absorbable zinc
Rosalind S. Gibson and Victoria P. Anderson
Abstract Helen Keller International is given, together with the

critical steps needed to scale up dietary diversification or
Dietary diversification or modification has the potential modification for programs and future research needs.
to prevent deficiencies of zinc and other coexisting limit-
ing micronutrients simultaneously, without risk of antag-
onistic interactions. In this review, we have addressed Key words: Dietary diversification, dietary modifica-
the following. The first section focuses on strategies with tion, zinc deficiency, zinc intake
the potential to enhance intake and/or bioavailability
of zinc, and includes interventions (with and without
nutrition education) based on agriculture, production Introduction
or promotion of animal-source foods through animal
husbandry or aquaculture, and commercial and house- The increasing recognition of the coexistence of MMN
hold processing strategies to enhance zinc absorption. deficiencies has highlighted the importance of dietary
Outcome indicators include intakes of foods or nutrients diversification or modification as a strategy to prevent
(although rarely zinc) and, in some cases, zinc status, or deficiencies of zinc and other coexisting limiting
zinc-related functional responses. The next two sections micronutrients simultaneously without risk of antago-
address whether dietary diversification or modification nistic interactions. Through a participatory research
can achieve increases in absorbable zinc that are suf- process that focuses on building relationships with the
ficient to enhance zinc status or zinc-related functional community and involving them in their design and
responses in breastfed infants and toddlers and in older implementation, dietary diversification or modification
children and women of reproductive age. Evidence for strategies have the potential to be culturally accept-
the impact of dietary diversification or modification on able, economically feasible, and sustainable, even in
behavior change and on nutritional status in the short poor resource settings. Further, such strategies have
and long term, and the possible role of modifying factors the added advantage of enhancing the micronutrient
(e.g., baseline nutritional status, socioeconomic status, adequacy of diets for the entire household and across
infection, sex, age, and life-stage group) is the emphasis generations. Several additional non-nutritional benefits
of the next section. The following section highlights the may also be achieved through the community-based
evidence for three potential adverse effects of dietary nature of dietary diversification or modification. These
diversification or modification: aflatoxin contamination may include empowerment of women in the commu-
from germinated cereals, loss of water-soluble nutrients, nity, training, and income generation. To be success-
and displacement of breastmilk. Finally, an example of a ful, however, a multidisciplinary team of specialists in
dietary diversification or modification program (Home- agriculture, nutrition, epidemiology, rural extension,
stead Food Production) developed and implemented by adult education, psychology, and community health
is essential to assist with the design, implementation,
monitoring, and evaluation of dietary diversification
Rosalind S. Gibson is affiliated with the Department of or modification strategies.
Human Nutrition, University of Otago, New Zealand. Victoria This paper provides a critical review of interven-
Anderson is affiliated with University College, University of tions employing dietary changes—diversification or
London. modification strategies at the community or household
Please direct queries to the corresponding author: Rosalind
S. Gibson, Department of Human Nutrition, University of level that have the potential to increase the intake
Otago, PO Box 56, Dunedin, New Zealand; e-mail: Rosalind. of total and/or absorbable zinc. Potential strategies
Gibson@stonebow.otago.ac.nz. at the level of agricultural production are addressed
S108 Food and Nutrition Bulletin, vol. 30, no. 1 © 2009, The United Nations University.
Dietary diversification or modification to enhance zinc intakes S109
elsewhere. Three types of intervention strategies the health and development of breastfed infants and
have been included: agricultural interventions, inter- young children?
ventions based on the production or promotion Section 3: Can foods nutritionally improved through
of animal-source foods, and strategies at the com- dietary diversification or modification have an impact
mercial or household level designed to enhance zinc on the zinc status and the health and development of
absorption. Only those interventions that included children and women of reproductive age?
an indicator of consumption, either at the household
or the individual level, are included, although very Section 4a: Do interventions that promote dietary
few of these measured intakes of total or bioavailable diversification or modification have an impact on
zinc. Depending on the evaluation design, outcome behavior change and on nutritional status in both the
indicators measured either changes or differences in short and the long term?
intakes of foods or nutrients, and in some cases, serum Section 4b: Is the impact modified by baseline nutri-
or plasma zinc, and functional health outcomes such tional status, socioeconomic status, infection, sex, age,
as anthropometry, morbidity, and cognitive develop- and life-stage group?
ment. Tables summarizing the interventions are also
provided in this review. Section 5: Are there any adverse effects of dietary
Details of the interventions reviewed were derived diversification or modification?
from the peer-reviewed literature, websites, and project
reports. A primary literature search was performed on
PubMed using the following key words: (agricult* OR Section 1a
farm OR garden OR rural development OR fish ponds)
AND (zinc* OR “animal source foods’’ OR “nutritional Can agricultural interventions that increase the produc-
status’’ OR anthropom* OR “diet intake’’ OR “nutrient tion, accessibility, and consumption of plant-based foods
intake’’ OR morbidity OR “child growth’’) and restricted enhance the intake and bioavailability of zinc?
mainly to human studies. This literature search yielded
45 articles. The gray literature search identified 15 addi- Conclusion
tional reports from conference proceedings from the
International Vitamin A Consultative Group (IVACG) Agricultural interventions can increase the production,
and others, and websites of the Food and Agriculture accessibility, and consumption of plant-based foods
Organization (www.fao.org), Helen Keller International by household members, provided they also contain a
(www.hki.org), the International Center for Research strong nutrition education component. Hence, they
on Women (www.icrw.org), the International Food have the potential to enhance intakes of zinc and other
Policy Research Institute (www.ifpri.org), the Interna- micronutrients, although the magnitude of the increase
tional Life Sciences Institute (www.ilsi.org), and World in bioavailable zinc that can be achieved depends on
Vision Canada (www.worldvision.ca). The important the type and amount of plant foods consumed by the
contributions made by Ruel [1] and Bierti et al. [2] household. Interventions that include cereals and leg-
through their earlier published reviews and the full umes have the potential to yield the greatest increase
report by Bierti et al. [3] kindly provided to the authors in zinc intake because of their higher zinc content
by PATH Canada are also acknowledged. compared with that of starchy roots or tubers, and
The review is divided into five sections, which fruits and vegetables. Nevertheless, the bioavailability
address the following questions in relation to interven- of zinc in cereals and legumes will be compromised by
tions based on dietary diversification or modification: high phytate levels unless phytate levels are reduced by
other strategies.
Section 1a: Can agricultural interventions that increase
Comprehensive agricultural strategies that also
the production, accessibility, and consumption of plant-
include a home gardening component with a focus
based foods enhance the intake and bioavailability of
on vitamin A–rich fruits and vegetables have several
zinc?
additional benefits. There is some indication that zinc
Section 1b: Can production or promotion of animal- absorption might be increased as a result of the interac-
source foods through animal husbandry or aquaculture tion between vitamin A and zinc. Further, substantial
enhance the intake of bioavailable zinc? increases in intakes of other important micronutrients,
including nonheme iron (although poorly absorbed),
Section 1c: Can processing strategies at the commer-
copper, vitamin C (an enhancer of nonheme iron
cial or household level enhance zinc absorption from
absorption), folate, thiamine, niacin, and dietary fiber,
plant-based diets?
will be provided by a more comprehensive agricultural
Section 2: Can complementary foods nutritionally intervention. Finally, because home gardening projects
improved through dietary diversification or modifica- often involve women, have the potential to generate
tion strategies have an impact on the zinc status and income, and can be readily integrated with nutrition
S110 R. S. Gibson and V. P. Anderson
education and behavior change strategies, they have all at the household level were used to estimate per capita
the attributes that have been associated with positive consumption, which limits the strength of the evidence.
nutrition outcomes. Eight of the agricultural interventions included a
To date, almost all of the interventions based exclu- nutrition education component [4–11, 14], and all
sively on agriculture have focused on improvements reported increased, or greater intakes of vegetables,
in vitamin A intakes or status from provitamin A–rich vitamin A–rich foods, or intakes of vitamin A, depend-
fruits and vegetables rather than zinc. Hence, the ing on the intervention design. However, the results of
data on increasing intakes of total and/or bioavail- three of these projects [6, 7, 13] are hampered by the
able zinc through agricultural interventions alone are absence of an appropriate control group. Furthermore,
negligible. it is not possible to distinguish the effect of the nutri-
tion education from the agricultural production or
Detailed review of evidence promotion effect in these projects, with the exception
of the project in Kenya [14]. In the Kenyan project,
Evidence for the first part of this conclusion is based differences in intakes of vitamin A–rich foods (i.e.,
on 10 agricultural interventions, chosen because they β-carotene–rich sweet potatoes), with and without
all included an indicator of consumption and, in some nutrition education, were investigated, and the results
cases, of production and accessibility. They have also clearly indicated a benefit of including nutrition educa-
been reviewed by Ruel [1] or Berti et al. [2] and are tion [14]. However, for the two projects listed in table 1
summarized in table 1. that did not include a nutrition education component
The focus of most of these projects was largely on [12, 13], the results were inconsistent. The project in
improving intakes of vitamin A–rich fruits and veg- Egypt [12] reported higher intakes of protein and iron
etables, with the exception of a project in Egypt that per capita in intervention households compared with
measured changes in the consumption of cereals and controls, whereas in the Nepal study [13], the results
legumes [12]. Increases in intake of provitamin A–rich are difficult to evaluate because no control group was
foods would result in some improvement in bioavail- included. Postintervention, vitamin A intakes were still
able zinc intakes, albeit small. The zinc content of fresh inadequate in mothers and children, despite a reported
fruits and green leafy vegetables is low (0.2 to 0.7 mg increase in the number of households producing
of zinc/100 g fresh weight), but zinc absorption might vegetables.
be enhanced by the concomitant increases in intakes
of provitamin A carotenoids, possibly as a result of the
interaction between vitamin A and zinc [15, 16]. By Section 1b
contrast, increased consumption of maize, wheat, and
peanuts, as reported in the study in Egypt [12], will lead Can production or promotion of animal-source foods
to greater increases in zinc intakes compared with the through animal husbandry or aquaculture enhance the
levels obtained through fruits and vegetables and will intake of bioavailable zinc?
be accompanied by more substantial increases in the
intakes of protein, iron, thiamine, niacin, folate, and Conclusion
dietary fiber. However, the bioavailability of zinc (and
iron) from these food sources will be poor because of Increasing the production of animal-source foods
their high phytate content. through animal husbandry or aquaculture in regions
In 6 of the 10 agricultural interventions [4, 5, 8, 9, 12, where such interventions are culturally appropriate
14], the evaluation design included both intervention can increase the consumption of animal-source foods
and control groups and, in three of these [4, 8,14], an in the household, especially when nutrition education
assessment of change from preintervention to postin- or behavior change is a component of the intervention.
tervention in both groups was reported. However, as Their impact on increasing intakes of bioavailable zinc
emphasized by Ruel [1], the intervention and control depends on the type and amount of animal-source
groups were sometimes in different geographic areas. foods consumed. Depending on the evaluation design,
Further, data on how the control group was selected increased consumption of red meat or liver can lead to
and its comparability to the intervention group at base- increased or higher intakes* of bioavailable zinc com-
line were not always provided, making it impossible pared with that obtained from fish or dairy products.
to establish whether any differences existed between This is because beef, pork, lamb, and liver have a
the two groups at baseline and to account for any such higher content of readily absorbed zinc (3.0 to 6.8 mg of
differences in the statistical analyses. For the remain- zinc/100 g) than poultry (~1.1 to 2.7 mg of zinc/100 g),
ing four interventions reviewed [6, 7, 10, 11, 13], the
evaluations were largely based on a comparison of pre- * “Increased intake” refers to situations when pre- and
intervention versus postintervention, with no control post-intervention data are available; “higher intakes” refer to
group. In two of the projects [4, 12], dietary intake data situations when only post-intervention data exist.)
TABLE 1. Agricultural interventions to increase household production, accessibility, or consumption of plant-based foods
Country, Intervention strategies Outcomes

year Nutrition
[reference] Production education Target groups Design Methods KAP + dietary intake Nutritional status
Bangladesh, Low-cost Nutrition Poor, near-landless Pre-postintervention Vegetable production, Increased size of plot culti- Decreased stunting in inter-
1998 home veg- education, women with Pilot study, 2-yr household dietary intake, vated, increased variety of vention (44%–34%) vs.
[4] Marsh etable gar- processing children < 5 yr, follow-up anthropometry, morbid- vegetables grown, increased control group (42%–35%)
dens, seed and cook- willing to engage Intervention group ity, clinical signs of VAD vegetable production, Decreased night-blindness in
distribution ing methods in farming (1,000 households) (night-blindness) year-round availability intervention (2.3%–1.2%)
to optimize activities vs. control group Anemia (no baseline and intake in intervention vs. control group (no
nutritional (200 households) measures) households (especially in change)
value of food; (separate village) young children) vs. control Maternal night-blindness
agricultural Increased decision-making lower in intervention (1.4%)
training by intervention womena than control group (3.4%)
No change in diarrhea preva-
lence during intervention
Anemia prevalence: interven-
tion (24.4%) vs. control
(33.2%)a
Tanzania, Home vegeta- Promotion of Mothers with No baseline Household KAP survey, Intervention mothers had No difference in serum
2000 ble gardens home produc- children aged 5-yr follow-up 7-day food-frequency better KAP than controls retinol (intervention vs.
[5] Kidala tion, con- 12–71 mo living Intervention house- recall intake of vitamin (p < .02). 67% of interven- control children) after
et al. sumption, and in VAD-prone holds (n = 75) vs A–rich foods tion households vs. 32% controlling for helminth
appropriate rural villages control households Serum retinol of control households had infection
storage of vita- (n = 71) Stool analysis (for home gardens (p = .001)
min A–rich helminths) 65% of intervention children
foods vs. 37% of control children
consumed > 7 vitamin
A–rich foods in past week
(p = .001)
India, 2000 Home vegeta- Nutrition Households with Pre-postintervention Production of vitamin Increased no. of varieties, Decreased clinical signs of
[6] ble gardens and health young children household survey A–rich foods, weekly production, and consump- VAD postintervention
Chakra- education (age group not No control group consumption of vitamin tion (> 2× baseline) of
varty specified) A–rich foods, clinical vegetables
signs of VAD Increased awareness of nutri-
tion needs and symptoms
of deficiency
continued
S111
TABLE 1. Agricultural interventions to increase household production, accessibility, or consumption of plant-based foods (continued) S112
Country, Intervention strategies Outcomes

year Nutrition
Vietnam, Home vegeta- Nutrition edu- Mothers in rural Pre-postintervention Maternal KAP, vegetable 26% increase in maternal Decreased clinical signs of
1995 ble gardens cation focused Vietnam design (2-yr production, dietary KAP scores VAD postintervention:
[7] Ngu on vitamin A follow-up) intake (by 24-h recall), Increased production (from night-blindness (from
et al. No control group clinical signs of VAD 41 to 233 g/person/day) 0.55% to 0.00%), Bitot’s
(night-blindness, Bitot’s Increased daily intakes of spots (from 0.40% to
spots, corneal scarring) energy (from 1,479 to 1,725 0.09%), corneal scarring
kcal), protein (from 39 to 48 (from 0.06 to 0.00%)a
g), and fat (from 8 to 14 g)
Increased vegetable intake by
children < 5 yr (from 13 to
32 g/day)
Philippines, Urban home Promotion Parents of chil- Pre-postintervention Vegetable production Increased consumption of > Not measured
1996 gardens and of vitamin dren < 14 yr, Intervention city vs. Consumption of vitamin 3 vegetables/wk
[8] Solon institutional A–rich veg- pregnant and control city A–rich vegetables by Increased percentage of
et al. gardens etables from lactating women, target group in previous children eating green leafy
3-mo media schoolchildren week (food-frequency vegetables daily (from 35%
campaign questionnaire) and day to 42%)
(e.g., radio (24-h recall) Vitamin A intake: 12%
gardening increase in intervention
lessons) children vs. 48% decrease
in control city children
No change in diet of children
< 1 yr
Guatemala, Home-gar- Promotion Families with No baseline Maternal KAP Maternal knowledge greater Control children at 3.5-fold
1996 dens (vita- of vitamin young children Intervention house- Dietary intake in intervention than in con- increased risk of VAD
[9] Phillips min A–rich A–rich plant holds (n = 300) vs. Percentage adopting trol groupa when home garden had no
et al. crops), foods control households gardens No. of days consuming vita- dark-green leafy vegetables
seeds, (n = 300) Cost-effectiveness of pro- min A–rich foods (both (controlled for SES and age)
technical Also compared with grams evaluated preformed vitamin A and
assistance VAC distribution vitamin A from garden
and fortification produce) greater in inter-
vention than in control
group (p < .01)
87% of intervention house-
holds had home garden
postintervention
R. S. Gibson and V. P. Anderson
Philippines, Home gar- Promotion Children in rural Pre-postintervention Intake of vitamin A, Increased vitamin A intake Improvement in expected
1979, dens (vita- of vitamin villages No control height, weight, serum in intervention group from weight (p < 0.05) in rela-
1980 min A–rich A–rich fruits 2-yr follow-up, inter- retinol baselinea tion to Harvard weight
[10] Solon fruits and and vegetables vention vs. VAC Clinical signs of VAD for-height in intervention
et al., [11] vegetables) Deworming, vs. vitamin A for- (xerophthalmia) group
Popkin medical care, tification of MSG Serum retinol: no change in
et al. pharmacy, (n≈400) intervention group (from
sanitation, 19.0 to 16.4 µg/dL) vs. MSG
immuniza- group (from 21.0 to 28.5
tion, health µg/dL)
and nutrition Xerophthalmia: no change in
education intervention group, signifi-
cant decrease in VAC and
MSG groups*
Egypt, 31 types of None Farmers and their Pre-postintervention Crop yield (maize, peanut, Energy intake similar in both No important differences in
1987 agricultural children 0–3 yr design wheat) groups. Higher per capita anthropometry, mortality,
[12] Galal interven- Self-selected Intervention house- Household dietary intake intakes of maize, peanuts, immunizations, or wean-
et al. tions to sample, owned holds (n = 227) vs. (24-h recall) wheat, and protein and iron ing age in subsurvey of
increase more land and control households Subsample: immuniza- in intervention than in con- 0–3-yr-oldsa
productiv- was therefore (n = 828) tions, anthropometry, trol householdsa
ity of local wealthier Subsurvey of chil- mortality, weaning age
crops and more dren 0–3 yr: inter-

literate than vention (n = 30) vs.
nonparticipants controls (n = 22)
Nepal, Home gar- None Rural households Pre-postintervention Anthropometry, “number Increased percentage of Deterioration of nutritional
1995 dens, water with children 3-yr follow-up of years since initiation households producing veg- status of children during
[13] CARE supply, irri- 6–60 mo No control group of kitchen garden” etables (from 51% to 75%)a study period
Nepal gation, seed Intake of vitamin A–rich Insufficient intake by moth-
distribution foods (food-frequency ers and children both pre-
questionnaire) (postin- and postintervention
tervention only)
Kenya, Introduc- Women’s Women and chil- Pre-postintervention Anthropometry, intake Increased frequency of intake Not reported
1999 tion of group meet- dren (n = 300) Intervention group of vitamin A (food-fre- of vitamin A–rich foods in
[14] β-carotene– ings: food aged 0–5 yr vs. control group quency questionnaire) intervention group (from
Hageni- rich sweet processing 4.2 to 5.8 food-frequency
mana potato techniques, questionnaire score) vs. con-
et al. vitamin trol group (from 4.3 to 3.0
A–rich foods food-frequency question-
naire score) when original
intake was low (p < .05)
KAP, knowledge, attitudes, and practices; MSG, monosodium glutamate; SES, socioeconomic status; VAC, vitamin A capsules; VAD, vitamin A deficiency
a. Not tested statistically or p-values not given.
S113
eggs (~1.0 to 1.3 mg of zinc/100 g), dairy products It is of interest that the project in northeastern Thai-
(~0.3 to 1.0 mg of zinc/100 g), or finfish (~0.3 to 0.7 mg land [21, 22] was initially a home gardening project
of zinc/100 g for flesh only; ~3.2 mg of zinc/100 g for that focused on ivy gourd, in combination with some
whole, soft-boned fish with bones) [17]. However, very nutrition and health education to enhance intake of
few of these interventions have quantified the changes provitamin A carotenoids. However, the project’s focus
in intakes of total or bioavailable zinc. on community participation led to the construction of
Animal-source foods can also contribute important fishponds and the introduction of a poultry-raising
and varying amounts (per 100 g or per MJ) of vitamin project to complement the home gardening efforts. As
B12, vitamin B2, readily available iron, calcium, and a result, consumption of chickens and eggs increased
preformed retinol, depending on the type of animal- among primary schoolchildren, who consumed them
source foods consumed. Nevertheless, because of in a school lunch program. Intakes of vitamin A, iron,
design limitations in the interventions reviewed, such or vitamin C also increased in certain life-stage groups,
as lack of baseline data that include intakes of total or but data on zinc intakes were not reported.
bioavailable zinc, and appropriate controls, the strength Four projects based exclusively on the production of
of the evidence for enhancing intake and bioavailability animal-source foods through small-animal husbandry
of dietary zinc through production or promotion of or aquaculture [27–31] are summarized in table 2B. In
animal-source foods is rated as moderate. three of the projects [27, 28, 31], increases in consump-
tion of animal-source foods were also reported. In the
Detailed review of evidence Ethiopian study, however, despite an increased intake
of animal-source foods overall, 37% of families still
Evidence for this conclusion is based on seven mixed consumed no meat, and in households with children,
interventions of agriculture combined with animal hus- only ~40% of eggs produced were consumed [27]. In
bandry or aquaculture (table 2A), four interventions the two Bangladesh studies [28–30], neither of which
based exclusively on the production of animal-source included nutrition education, the animal-source food
foods (table 2B), four in which the consumption of consumed by the treatment group was fish, especially
animal-source foods was promoted (table 2C), and small indigenous fish species. In the intervention of
four nonblinded randomized, controlled trials in which Nielsen et al. [28] (table 2B), the small indigenous fish
meat-based foods were supplied to the participants species were purchased by the intervention households
(table 2D). with income generated from the sale of the eggs or
Three of the mixed interventions (table 2A) [18–22] chickens produced, whereas in the study by Roos et al.
have been reviewed by Ruel [1] and Bierti et al. [2]. [29, 30], although fish was eaten with a high frequency,
Four additional mixed interventions not reported in the amount consumed was small. Indeed, total fish
these earlier reviews were also examined (table 2A) consumption did not differ between the fish-producing
[23–26]. and the non-fish-producing households (table 2B).
Nutrition education or behavior change was a com- Recent studies have highlighted the potential of cer-
ponent of five of the seven mixed interventions, tain indigenous fish species in Cambodia and Bangla-
although in two of the projects, consumption was not desh as rich sources of zinc (e.g., Esomus longimanus),
measured [25, 26], so that the impact of the interven- as well as of vitamin A and iron (e.g., Amblypharyngo-
tion on consumption of animal-source foods cannot don mola) [29, 30], so the impact of the consumption of
be assessed. In the three interventions in which both indigenous fish on intakes of bioavailable zinc warrants
nutrition education and consumption were measured attention. Moreover, improvements in intakes of sev-
[18, 19, 21, 22, 24], increases or higher intakes of ani- eral other important nutrients, such as animal protein
mal-source foods or iron were reported in the interven- and vitamin B12, as well as bioavailable calcium, are
tion as compared with the control groups, depending likely provided the small indigenous fish species are
on whether change from pre- to postintervention, or consumed whole with bones.
difference. In the Malawian project [31], increases in both the
It is noteworthy that in the two interventions without production and consumption of guinea fowl, chickens,
nutrition education [20, 23], no increases in consump- rabbits, eggs, and goat’s milk were observed in the
tion of fish were reported among the fishpond groups intervention households, as a result of a small-animal
compared with the control groups, even though in the revolving fund set up by World Vision. The use of these
Bangladesh study fish production increased [20]. The animal-source foods for household consumption rather
sample size of the mixed intervention in Thailand [23], than income generation was attributed to the nutrition
however, was very small (n = 30 children per group) education component of the World Vision Micronu-
and may not have been adequate to show an effect on trient and Health (MICAH) program, but because of
fish consumption. In the intervention in Iran [26], no flaws in the study design, this conclusion cannot be
measures of food production or consumption were confirmed [31].
reported. Of the total of 10 (6 in table 2A, 4 in table 2B)
TABLE 2A. Mixed interventions: Agriculture combined with small-animal husbandry or aquaculture to increase production or promotion of animal-source foods
Intervention strategies Outcomes
Country, year Nutrition
Vietnam, 1997, Home gardens, Targeted at Mothers and Intervention Anthropometry Intervention mothers had Decreased stunting (from
1998 fishponds, and mothers: iron-, their preschool group (n = 469) and morbidity better KAP than controlsa 50.3% to 41.7%, p = .0007)
[18] English et al. animals zinc-, vitamin children vs. control at baseline and Intervention group had higher in intervention vs. no
[19] English and Food production C-, protein-, (< 6 yr) group (n = follow-up production of fruit, veg- change in control group
Badcock and utilization and carotene- 251) from 1991 At follow-up: etables, fish, and eggs than (from 45.8% to 47.6%)
assessed at rich foods to 1993 maternal KAP, control group. Intervention Decreased incidence of ARI
baseline Only some food produc- children had higher intakes (from 49.5% to 11.2%, p <
measure- tion, and of vegetables, fruit, energy, .0001) in intervention group
ments taken at dietary intakes protein, vitamin C, vitamin vs. no change in control
baseline of households A (100 vs. 60 RE/day), and group
and young iron (8.9 vs. 5.4 mg/day) Decreased incidence of pneu-
children by 24 than controls (p < .01) monia complications (rapid
h recall breathing: 15.5% to 0.9%;
chest indrawing: 5.2% to
0.2%) in intervention group
vs. no change in controls
Decreased incidence of
diarrhea (from 18.3% to
5.1%, p < .0001) in interven-
tion group vs. no change in
controls
Bangladesh, Vegetable pro- No Women, their Pre-postinter- Focus group Increased production but no No effect on hemoglobin from
1998 duction, fish- households vention over interviews: increased consumption of either fishponds or vegetable
[20] IFPRI ponds, credit and children 2 yr (consumption fish among fishpond group production intervention
and agricul- Fishpond group, measures) Increased vegetable intake
tural training vegetable Anthropometry, among vegetable group
group, control hemoglobin compared with controls
group
Thailand, 1999 Seed distribu- Nutrition educa- Pregnant or lac- Quasi-experi- Maternal KAP, Intervention group: increased Intervention schoolgirls:
[21] Smitasiri tion, training tion and social tating women, mental design dietary intakes iron and vitamin A KAP increased serum retinol
and Dhana- of farmer marketing children Pre-postinter- by 24-h recall, at postintervention and (from 22.8 ± 7.0 to 33.7 ±
mitta, w women, targeted at 2–5 yr, and vention, 9-mo biochemical vs. controls (p < .01); 8.2 µg/dL) vs. no change in
[22] Smitasiri promotion women schoolgirls follow-up assessment (in increased vitamin A intake controls
et al. of gardens, Intervention schoolgirls) (all groups) (no change
fishponds, and group (n≈500) in fat intake); increased
chicken raising vs. control iron intake in 2–5-yr-olds,
group (n≈500) 10–13-yr-olds, and lactating
women; increased vitamin C
S115
intake in lactating womena

TABLE 2A. Mixed interventions: Agriculture combined with small-animal husbandry or aquaculture to increase production or promotion of animal-source foods (continued) S116
Intervention strategies Outcomes
Country, year Nutrition
Thailand, 2002 Home veg- No Children aged Quasi-experi- Interviews, No differences between No differences between
[23] Schipani etable gardens, 1–7 yr mental design anthropom- groups in mean intake of groups in mean hemo-
et al. fishponds, Gardening etry, 24-h recall energy, protein, fat, total globin, serum ferritin,
small-animal households (n over 3 seasons, iron, animal iron, plant iron, or retinol (for 3 seasons
husbandry, = 30) vs. ran- serum retinol, or vitamin C compared)
fruit orchards domly selected serum ferritin, Weight-for-height, weight-
nongardening hemoglobin for-age, and height-for-age
households z-scores tended to be higher
(control) (n in gardening group (NS)
= 30)
Seasonal study
(rainy, cool,
hot seasons)
Bangladesh, Home gardens Messages to Mothers and Pre-postinter- Food-frequency Increased access, availability, 12% decrease in anemia
2006 Poultry, fish, increase con- youngest child vention over questionnaire: and consumption of plant- among nonpregnant inter-
[24] Stallkamp milking cow sumption of < 5 yr 3 yr animal-source and animal-source foods vention group at postinter-
et al. animal-source Intervention foods in intervention vs. control vention (p < .05)
foods (eggs, households (n 24-h recall: red/ group
meat, liver, = 420) vs. con- orange/yellow Increased vitamin A intake in
milk) trol households fruits and veg- intervention children (6–59
(n = 420) etables, dark- mo) and mothers postinter-
green leafy vention (p < .05)
vegetables
Hemoglobin
China, 1994 Township com- Nutrition train- Poor rural, chil- Pre-postinter- Anthropometry, Changes in intake not Decreased prevalence of
[25] Ying et al. mittees encour- ing for health dren aged < 6 vention hemoglobin measured malnutrition: wasting
aged increased workers yr (n = 9,921), 4-yr follow-up –18%, stunting –2.6%,
production Media campaign: subsample No control group underweight –11.3% at
of green leafy advocacy for (n = 500) postinterventiona
vegetables, breastfeeding, Decreased prevalence of
poultry, and nutrition, and anemia at postintervention
small livestock health across all age groups (e.g.,
2–3-yr-olds: 37.7% to 27.5%,
p < .01)
Iran, 2004 Rural coopera- Nutrition, Children aged Pre-postinter- KAP: breast- Changes in KAP and food Decreased prevalence of
[26] Sheikhole- tives, income- health, and 6–35 mo vention feeding and intake not measured or underweight in interven-
slam et al. generating and literacy edu- Intervention Intervention vs. child feeding, reported tion group in Borazjan
loan schemes cation for group: n = control growth moni- (from 23% to 11%) and
Promotion of mothers 1,695 at base- Follow-up at 3 yr toring, family Bardsir (from 28% to 14%)
home gardens Promotion of line, n = 1,149 (anthropom- planning, sani- (p < .0001)
and animal breastfeeding, at endline etry only) tation, drink- Decreased prevalence of
husbandry complemen- Control group: ing water, food underweight in control
tary feeding, n = 1,631 production group in Bardsir (from 24%
consumption at baseline, Food-frequency to 16%, p < .0001)
of dairy prod- n = 1,029 at questionnaire Decreased prevalence of
ucts and fruits endline for household stunting across all districts
3 districts: Ilan, and child in intervention groups (Ilan:
Borazjan, and consumption from 25% to 12%; Borazjan:
Bardsir patterns from 41% to 13%; Bardsir:
Anthropometry from 31% to 19%; p < .0001)
Decreased prevalence of
stunting in Bardsir con-
trols (from 24% to 16%,
p < .0001)
Decreased wasting only in
Borazjan intervention group
(from 9% to 4%, p < .0001)
No statistical testing of differ-
ences between intervention
and control groups
*Not tested statistically or p values not given.
ARI, acute respiratory infection; KAP, knowledge, attitudes, and practices; NS, not significant; RE, retinol equivalents
S117
TABLE 2B. Animal-source food interventions: Small-animal husbandry or aquaculture to increase production or promotion of animal-source foods S118
Intervention strategy Outcome
Country, year Nutrition Nutritional
[reference] Production education Target groups Design Methods KAP + dietary intake status
Ethiopia, 2003 Dairy goats, Community- Poor, women- Pre-postintervention Production and con- Increased milk intake in children* No measures
[27] Ayele and poultry, based headed 2-yr follow-up sumption of goat’s Increased energy, animal protein, total pro-
Peacock home nutrition households No control group milk, goat meat, and tein, and fat from animal-source foodsa
vegetable education (n = 5,500) eggs (no. of eggs/ Some increase in meat intake, but 37% of
gardens to promote Children aged household/mo from families still had no meat
increased 6 mo to 6 yr food-frequency ques- In households with children, only ~40% of
consump- (n = 39) tionnaire and house- eggs produced were consumed
tion of hold survey)
goat’s milk
Bangladesh, Semiscav- No Women of Cross-sectional Production, utilization, More eggs per month produced and sold in No measures
2003 enging reproduc- comparative study and consumption adopter than in nonadopter households
[28] Nielsen poultry tive age and Adopters (women [n of poultry products NSD in egg or chicken consumption between
et al. production girls from = 35] and daugh- (24-h recall, qualitative the 2 groups
poor rural ters 5–12 yr [n = questionnaires) Consumption of small fish was higher in
households 35]) vs. nonadop- adopters than in nonadopters: 58 vs. 39 g
ters (women [n = fish/person/day (p < .08)
35] and daughters
5–12 yr [n = 35])
Bangladesh, Homestead No Poor rural Observational: fish- 7-mo fish production No difference in fish intake between groups No measures
2003 fishponds households producing group Household fish con- 98% of households consumed fish on at least
[29] Roos et al., (n = 59) vs. non- sumption (5-day recall 1 of the 5 days surveyed, of which 84% was
[30] Roos et al. fish-producing in 3 seasons) contributed by small wild indigenous fish
group (n = 25) Calcium, iron, and Vitamin A–rich “mola” fish contributed 18%
vitamin A intakes of the RDA for vitamin A at the household
determined level
Malawi, 2005 Small- Nutrition Subsistence Pre-postintervention Production, accessibility, Intervention group: increased KAP of No measures
[31] Radford animal education farming Intervention house- and consumption of animal-source foods to prevent anemia in reported
revolv- focused on households holds vs. control animal-source foods pregnant women
ing fund: importance households Increased prevalence of small-ani-
goats, of con- (sample size not mal husbandry (from 40% to 72%)
chickens, sumption reported) postintervention
guinea of animal- Increased percentage of households consum-
fowl, and source ing produce vs. control group (goat meat,
rabbits foods to 28% vs. 17%; poultry, 57% vs. 41%; rabbit,
prevent 66% vs. 41%; eggs, 68% vs. 51%)
anemia
KAP, knowledge, attitudes, and practices; NSD, no significant difference; RDA, recommended daily allowance
mixed agricultural or livestock production interven- complementary feeding [35].

tions reviewed, 6 included a control group (table 2A The results of two nonblinded, randomized, control-
[18–24]) and table 2B [31]). However, even these six led trials in which red meat–based foods were supplied
projects had limitations in their design and evaluation to infants (table 2D) [38] or schoolchildren [43] by the
that compromise the interpretation of the results. Of investigators also provide strong evidence that the con-
the remaining four, the two conducted in Bangladesh sumption of red meat–based foods can enhance intakes
included comparison groups, which were defined as of bioavailable zinc; these two trials are summarized in
adopters versus nonadopters [28] (table 2B) and fish- table 2D. In both of these trials, the group given red
producing families versus non-fish-producing families meat had higher [38] or increased [43] intakes of total
[29, 30], and two were evaluated by a comparison or available zinc (and iron) than controls. Furthermore,
between pre- and postintervention values, with no in the study on schoolchildren in Kenya, intakes of
control group [25, 27] (table 2A and B). vitamin B12 and vitamin A also increased in the group
The five interventions that promoted the consump- receiving red meat compared with the corresponding
tion (but not production) of animal-source foods changes in the control group postintervention. Intakes
through nutrition education and behavior change are of calcium, vitamin B12, vitamin A, and riboflavin were
summarized in table 2C. Of these, three were non- also significantly higher postintervention in the group
blinded, randomized, controlled trials [33, 35, 36], receiving milk than in the control group [43].
and two had a quasi-experimentally controlled design By contrast, no improvements in zinc (or iron)
with a nonequivalent control group [32, 34]. Three of intakes were observed in Danish infants [39] given
these interventions targeted women of childbearing a high-meat diet for 2 months; however, compliance
age [32, 33, 36]. The two in Peru [32, 33] involved the was not reported and the study was small (n = 41).
development and preparation of low-cost, heme-iron- Likewise, when fish powder was supplied to Ghanaian
rich meals (e.g., from liver, blood, spleen, and fish) in infants [37] or milk was given to Kenyan schoolchil-
community kitchens. Consumption of these meals led dren [43], no improvements in zinc (or iron) intakes as
to increases in intakes of bioavailable iron and vitamin compared with the control group were reported (except
C compared with preintervention levels and to the for the Ghanaian infants at 7 months of age). These
corresponding changes observed in the control groups. results indicate that red meat, but not milk or powdered
Improvements in intakes of bioavailable zinc (as well fish, can enhance intakes of readily available zinc when
as animal protein, vitamin B12, and preformed retinol) consumed by infants and schoolchildren.
probably also occurred but were not measured. In the
third, partially blinded, randomized, controlled trial
in New Zealand [36], women were encouraged to con- Section 1c
sume more animal-source foods (as well as to follow
other strategies to enhance nonheme iron absorption) Can processing strategies at the commercial or household
through intensive dietary counseling by a research level enhance zinc absorption from plant-based diets?
dietician. In this study, intakes of bioavailable zinc were
measured, but the data for zinc are not yet published Conclusion
(Gibson RS, personal communication).
The other two interventions summarized in table 2C There is abundant evidence, based on in vivo zinc
targeted infants rather than women of childbearing age radioactive or stable isotope studies, that high levels
[34, 35]. In the study of infants conducted in China of dietary phytate inhibit zinc absorption (table 3A)
[34], egg yolk was targeted for infant feeding, resulting and that by reducing the phytate content in cereal-
in the nutrition education and behavior change group based diets through commercially available exogenous
feeding more egg yolks (per day) than the comparison phytase enzymes, or phytases naturally occurring in
group; changes in nutrient intakes, including zinc, were whole-grain cereals, zinc absorption can be enhanced
not reported. In contrast, in a very successful effective- (table 3B). Whether zinc absorption can be enhanced
ness trial conducted in Peru, intakes of available zinc through household strategies designed to reduce the
(and iron) from complementary foods were measured phytate content of cereal-based diets (table 3B) has
and were significantly increased in children 6 to 18 not been investigated by isotopic measurements of
months old in the intervention compared with the con- zinc absorption. Phytate reductions of ~50% have been
trol group [35]. These improvements were attributed achieved by soaking pounded maize or maize flour or
to feeding more nutrient-dense, thick complementary fermenting maize porridges. Significant increases in
foods containing animal-source foods, mainly chicken zinc absorption have been achieved in men fed meals
liver (table 2C). This nonblinded trial employed a prepared from maize with a 60% reduction in phytate
cluster-randomized design and thus provides strong content compared with absorption from meals made
evidence for the role of animal-source foods, especially from wild-type maize (table 3C), a result suggesting
liver, in enhancing intakes of bioavailable zinc during that some improvement in zinc absorption is likely with
TABLE 2C. Animal-source food interventions: Nutrition education and behavior change to promote animal-source foods
Country, year Intervention strategy

[reference] Production Nutrition education Target group Design
Peru, 1998 Development of recipes Health and nutrition Periurban nonpregnant Quasi-experimental
Carrasco Sanez et and meals high in education in commu- women, aged 15–49 yr design with nonequiv-
al. [32] heme iron, vitamin A, nity kitchens alent controls
and vitamin C, some Capacity building of Preintervention (n =
of which were con- local women 310)/postintervention
sumed by the interven- (n = 189)
tion group 1-yr follow-up of inter-
vention group (n = 81)
vs. nonmembers as
controls (n = 108)
Peru, 2000 Development of iron- Behavior and nutrition Adolescent girls aged Cluster-randomized,
[33] Creed-Kana- rich and vitamin education campaign to 12–17.9 yr controlled trial, pre-
shiro et al. C–rich menus increase intake of iron postintervention
and vitamin C 9-mo follow-up
Promote local heme iron Intervention group (n =
sources (chicken liver 71) vs. nonequivalent
and blood, spleen, control group (n = 50)
beans) and vitamin C
China, 2000 NA Nutrition education and Infants aged 4–12 mo Quasi-experimental

[34] Guldan et al. counseling visits to from rural China design with nonequiv-
increase breastfeeding alent controls.
and quality and quan- Pre-postintervention
tity of complementary 1-yr follow-up
foods from 4 mo of Intervention group (n =
age (e.g., give egg yolk 250) vs. control group
daily after 46 mo) (n = 245)
Peru, 2005 NA Nutrition education to Birth cohort: infants Cluster-randomized
[35] Penny et al. increase intake of thick from a poor, periurban trial (nonblinded)
purees and animal- area, followed from 0 Pre-postintervention
source foods and to 18 mo Intervention (n = 187)
increase practice of vs. control (n = 190)
responsive feeding 18-mo follow-up
Demonstrations of
complementary food
preparation
Accreditation system
in government health
facilities
New Zealand 2001 NA Individual dietary coun- Premenopausal women Randomized, control-
[36] Heath et al. seling to diet group (n = 75), aged 18–40 led trial
only to increase intake yr with mild iron 3 groups: placebo, iron
and bioavailability of deficiencyb supplement (50 mg
dietary iron; also pro- iron/day), diet group
vided with a cast-iron (not blinded)
fry pan and fruit juice 16-wk follow-up
KAP, knowledge, attitudes, and practices; LAZ, length-for-age z-score; NA, not available; NSD, nonsignificant difference; SES, socioeconomic
status; TfR, serum transferrin receptor; WAZ, weight-for-age z-score
b. Mild iron deficiency is defined as serum ferritin < 20 µg/L and hemoglobin ≥120 g/L, in the absence of infection (i.e., elevated C-reactive
protein).
Outcomes
Methods KAP + dietary intake Nutritional status
Interviews Improved quality of meals: Decreased prevalence of anemia (from
Focus groups Increased iron, heme iron, vitamin A, and 49% to 41%) postintervention (p < .05)
24-h recall vitamin C content in intervention group
Hemoglobin Increased intake of iron and vitamin C–rich Note: Not compared with “control”
foods group of nonmembers
Increased intake of heme iron, bioavailable
iron, and vitamin C in intervention group
vs. controls
No increase in vitamin A in intervention
group vs. controls
KAP assessment: food, nutrition, Intervention vs. control: Change in anemia prevalence: from
health, and anemia Increased heme iron and anemia KAP 14.1% to 12.3% (NSD) in intervention
2 24-h recalls (pre- and Increased use of animal-source foods in group vs. from 14% to 37.5% in con-
postintervention) menus trol group (p < .01)
Hemoglobin, serum ferritin, height Increased intakes of total iron (from 7.75 Change in prevalence of iron deficiency
and weight measures ± 3.5 to 9.42 ± 5.0 mg/day, p < .01) vs. no (measured by serum ferritin): from
change in controls; heme iron (from 0.21 21.1% to 18.5% (NSD) in intervention
± 0.17 to 0.66 ± 1.35 mg/day, p < .01) vs. group vs. from 14% to 25% in control
no change in controls; vitamin C (from 44 group (NSD)
± 39.6 to 67 ± 45 mg/day, p < .05) in inter-
vention group vs. from 41 ± 34.6 to 40 ±
27.6 mg/day in controls [NSD]); absorb-
able iron (from 0.33 ± 0.16 to 0.43 ± 0.41
mg/day) vs. no change in controls (from
0.35 ± 0.13 to 0.37 ± 0.22 mg/day)
Weight and length Intervention group had greater nutri- NSD in growth between 2 groups before
Single 24-h recall; food-frequency tion knowledge, higher breastfeeding age of 12 mo; at 12 mo, WAZ in inter-
questionnaire rates (83% vs. 75%, p = .034), and better vention group was –1.17 vs. –1.93 for
KAP: Infant feeding and health- reported infant-feeding practices vs. con- controls (p = .004), HAZ was –1.32
related behaviors trols (p < .05) vs. –1.96 for controls (p = .022), and
Hemoglobin measures Intervention group had greater no. of eggs prevalence of anemia was 22% vs. 32%
fed per day to children 4–9 mo than for controls (p = .008)
controls
Home interviews and observations Intervention vs. control group: Intervention group had higher housing
for SES, hygiene, and feeding More caregivers received nutrition educa- and hygiene scores, education level,
practices tion (16/31 [52%] vs. 9/39 [24%]; p = .02) and body weight than controls at base-
Weight and length More infants were fed nutrient-dense thick line: analysis performed without and
24-h recalls: intake of complemen- purees at 6 mo (31% vs. 20%, p = .03) with adjustment
tary foods at 6, 9, 12, and 18 mo Higher intake of energy from animal-source Stunting at 18 mo: intervention 5% vs.
Morbidity (over past 24 h) at same foods at 15 mo (p = .082) and 18 mo (p = control 16% (p = .02; adjusted odds
visits .001) ratio, 3.04; 95% CI 1.21 to 7.64)
Fewer children failed to meet requirements Adjusted mean changes in WAZ and
for energy (8 and 12 mo), iron (8 and 9 LAZ better in intervention than in
mo), and zinc (9 mo) (p < .05) control group at 18 mo (p < .05)
Iron-specific validated food- Diet group: increased intake of flesh foods, Diet group: 26% increase in serum fer-
frequency questionnaire, serum heme iron, vitamin C, foods cooked in ritin vs. placebo group (NSD)
ferritin, hemoglobin, serum TfR, cast-iron cookware (p < .05); decreased 22% decrease in serum TfR:serum ferri-
C-reactive protein intake of phytate (p < .05) vs. placebo tin ratio vs. placebo group (NSD)
group
TABLE 2D. Nutritional improvement of home-prepared complementary foods or school lunches from inclusion of animal-source foods S122
Intervention strategy Outcome
Country, year Nutrition Target
[reference] Production education groups Design Methods KAP + dietary intake Nutritional status
Ghana, 1999 No No Healthy breast- Randomized, control- Monthly 24-h recalls for No differences in No significant differences among
[37] Lartey et al. fed infants ≥ led trial for 6 mo, 3 days (6–12 mo), 12-h intakes between W, 4 groups in morbidity outcomes,
2.5 kg at birth with 4 treatments— weighed records for 50% WF, and KF groups weight and length gain, differ-
Weanimix (W), W + of subjects (zinc intake except at 7 mo, when ences in head circumference,
vitamins and minerals from weighed records). zinc and iron intakes mid-upper-arm circumference,
(WVM), W + fish Morbidity, anthropome- were higher in KF skinfolds, arm fat area, arm
powder (WF), and try, hemoglobin, hemat- than W group (p < muscle area, plasma zinc, hemo-
koko with fish power ocrit, C-reactive protein, .05) globin, hematocrit, transferrin
(KF)—compared with plasma zinc, serum saturation, or RBC B-2 at any
nonintervention group retinol, serum ferritin, time.
(NI) (n = 464) TIBC, RBC B-2 WAZ and LAZ scores of NI group
were lower (p<0.05) than com-
bined intervention group at 6 and
9 mo of age.
Significant increase in percentage
with low ferritin between 6 and
12 mo in W, WF, and KF groups
but not in WVM group (p < 0.05)
Change in plasma retinol was sig-
nificantly greater in WM between
6 and 12 mo than in other 3
groups combined (0.14 ± 0.3 vs.
−0.04 ± 0.3 µmol/L, p = .003)
All 4 foods improved growth rela-
tive to nonintervention group (p
< .001)
USA, 2006 No No Exclusively Randomized, control- 2 3-day diet records/ Mean zinc intake Greater increase in head circumfer-
[38] Krebs et al. breastfed, led trial mo for 4 visits ( + zinc higher in meat than ence in meat than in cereal group
healthy, Pureed beef (n = 46) vs. intake) cereal group at 5 and Zinc and protein significant predic-
infants fol- iron-fortified infant Rating scale of infant’s 7 mo (p < .001) tors of head growth
lowed from ~6 rice cereal (n = 42) as acceptance of comple- Tolerance and accept- No biochemical differences
to 12 mo first complementary mentary food ance of beef and between groups
food at ~6 mo, plus Anthropometry, develop- cereal comparable Trend toward higher behavior
fruits and vegetables mental testing (Bayley index at 12 mo in meat than in
as desired scale), hemoglobin and cereal group (p = .08)
9-mo follow-up hematocrit, serum fer-
ritin, somatomedin,
plasma zinc
Denmark, 1998 No No Healthy, term, Randomized trial Food records: 24-h Despite higher intakes Change in hemoglobin: LMG –4.9
[39] Engelmann partially Low-meat group (LMG) weighed food record of meat and iron (−12.9 to −5.6 g/L) vs. HMG –0.6
et al. breastfed 10 g meat/day (n = 20) 1/wk from meat, differ- (−12.1 to −7.3 g/L) (p = .008)
infants, aged 8 vs. high-meat group Anthropometry, serum ences in intakes (mg/ NSD in change in serum ferritin,
mo (n = 41) (HMG) 27 g meat/day zinc, hemoglobin, serum day, range) of total TfR, or serum zinc
(n = 21) ferritin, TfR, morbidity iron in HMG (3.1, No differences in morbidity
2-mo follow-up (frequency of illness) 0.4–6.2) vs. LMG between LMG and HMG
(3.4, 1.4–6.1) or
zinc in HMG (3.3,
0.4–5.2) vs. LMG
(3.2, 1.4-5.2) were
not significant
Kenya, 2003, No No Undernour- Children randomized Physical examination and Intakes of calcium, In all 3 treatment groups, weight
2006 ished school- by school to 3 treat- health status vitamin A, vitamin gain was greater than in controls
[40, 41] Grillen- children aged ments—meat (n = Anthropometry B12, and vitamin B2 (p < .05)
berger et al. 5–14 yr (n = 134), milk (n = 144), 3 24-h recalls increased more in No effect of treatments on height,
[42] Murphy 554) or energy (n = 148) Cognitive, behavioral. and milk group than in HAZ, WHZ, or body fat
et al. supplement)—or con- activity measurements controls Gain in arm muscle area was
[43] Neumann trol (n = 129) group Morbidity Intakes of available greater in meat group than in
et al. 23-mo follow-up Biochemical measures: zinc, available iron, milk, energy, or control groups
[44] Siekmann zinc, iron, copper, reti- calcium, vitamin Among those with lower HAZ
et al. nol, folate, vitamin B2, B12, and vitamin B2 (≤ −1.4), those in milk group
[45] Whaley and vitamin B12 increased more in gained 1.3 cm more height than
et al. At baseline, across all meat group than in controls (p = .05) and 1 cm more
groups and independ- controls height than those in meat group
ently of intervention, (p = .09), but there were no effects
children with Entamoeba on change in HAZ
histolytica infection had Meat group had greater increase in
smaller increases in cognitive performance (especially
serum iron but greater on Raven’s Progressive Matrices)
increases in plasma reti- than all other groups
nol (p = .01), whereas No improvements in biochemical
those with enlarged zinc, iron, copper, folate, retinol,
spleens had smaller or vitamin B2 status
increases in plasma vita- Vitamin B12 status increased in
min B12 meat and milk groups compared
with energy and control groups
HAZ, height-for-age z-score; KAP, knowledge, attitudes, and practices; NSD, nonsignificant difference; RBC B-2, red blood cell vitamin B2 (riboflavin); TfR, transferrin receptor; TIBC, total iron-binding
capacity; WAZ, weight-for-age z-score; WHZ, weight-for-height z-score
S123
a 50% phytate reduction, the magnitude depending on ZIP4 is the principal zinc transporter responsible for
the composition of the habitual diet, zinc intake, and regulation of zinc absorption across the apical mem-
life-stage group. Reductions in the phytate content brane of the enterocyte [58].
of cereal- and legume-based diets will also increase In the three stable isotope studies summarized in
absorption of nonheme iron and calcium. Whether table 3B, the commercial strategies used completely
zinc absorption can be improved in vivo by the addition degraded the phytate content of the cereal-based test
of enhancers such as low-molecular-weight organic meals. However, when household strategies such as
acids produced during fermentation has not been soaking, germination, or fermentation are used, only
confirmed by isotope studies; only in vitro zinc dialyz- about 50% of the phytate content can be removed
ability has been measured. [59].
At present, there does not appear to be any evidence The level of phytate reduction required to yield a
of up-regulation of zinc absorption during infancy and marked improvement in fractional zinc absorption
childhood to help meet the additional zinc require- (FAZ) in population groups consuming cereal- and
ments for growth. Hence, for complementary diets legume-based diets will vary with the composition of
containing predominantly plant-based foods, phytate the cereal-legume blend, zinc intake, life-stage group,
reduction strategies without a concomitant increase in and health and nutritional status. The results of two
animal-source foods are unlikely to increase the intake short-term (single-day) stable isotope studies (table
of absorbed zinc to a level that meets the desired zinc 3C) [60, 61] reported a FAZ of 30% and 28% in adults
density for complementary foods set by the World fed 60% phytate-reduced maize and a FAZ of 38% in
Health Organization (WHO) [46] (table 3D). Even for adults fed 80% phytate-reduced maize, compared with
older children consuming predominantly plant-based levels ranging from 14% to 17% for the correspond-
diets, again a combination of strategies that includes ing wild-type maize with typical phytate content.
red meat or liver is probably needed to ensure their Increases in absorption of iron and calcium have also
estimated average requirements for zinc are met [47]. been reported among adults fed phytate-reduced maize
In both cases, however, in vivo zinc isotope absorption [63–65].
studies are needed to confirm these conclusions. Nevertheless, the results of a stable isotope study
of zinc absorption in Guatemalan schoolchildren fed
Detailed review of evidence normal or low-phytate maize (table 3C) [62] have
emphasized the difficulty of predicting the increase
The first part of this conclusion is based on a review of in FAZ that may occur when maize with a 50% reduc-
six tracer studies that measured and compared appar- tion in phytate content achieved through household
ent or fractional zinc absorption in low- and high- strategies is used. Although FAZ is known to be posi-
phytate diets (table 3A). Four of these studies were tively associated with the extent of phytate reduction
performed in adults [48–50, 52], one was performed in (table 3C) [61], maize is not the only source of phytate
young children [53], and one was performed in nurs- in household diets. Moreover, because FAZ is also
ing rhesus monkeys and nursing rat pups [51]. The dependent on the quantity of zinc ingested, declining
results have shown consistently that zinc absorption with increasing quantities of zinc consumed [57], this
is significantly lower from the high-phytate than from effect may counteract any increase in FAZ arising from
the low-phytate diets (table 3A). Further, the results a reduction in dietary phytate. Indeed, an unexpected
of two of the three stable isotope studies summarized increase in zinc intakes may have been responsible
in table 3B, in which dephytinization was achieved in part for the failure to observe any increase in FAZ
through the use of commercial phytase enzymes from among Guatemalan children fed low-phytate maize;
Aspergillus niger [54], or by naturally occurring phytase these results are also summarized in table 3C [62].
enzymes [56], showed increases in zinc absorption in Interventions employing household phytate reduc-
the dephytinized soybean isolate and wheat-soy blend tion strategies (and others) are summarized in table 3D
porridge compared with zinc absorption from soybean [66–73]. The Malawian intervention studies summa-
isolate and wheat-soy blend porridges with their native rized in table 3D [70–73] promoted a combination of
phytate content. In contrast, in a hospital-based study dietary strategies to enhance both the content and the
in school-aged children in Malawi [55], which also bioavailability of zinc in the maize-based Malawian
used a corn-plus-soy porridge dephytinized with A. diets. These included strategies to increase the content
niger, fractional zinc absorption was increased and of zinc-absorption enhancers such as organic acids
endogenous zinc losses decreased only in the children through fermentation, as well as the phytate-reducing
recovering from tuberculosis, but not in the apparently strategies noted earlier, in an effort to further enhance
well children. Restriction of this response to the chil- the content and bioavailability of zinc. Low-molecular-
dren experiencing catch-up growth suggested that zinc weight organic acids produced during fermentation
absorption may be up-regulated by ZIP4 in response to have the potential to chelate zinc (and iron), thereby
the high zinc requirements for catch-up growth [57]. making zinc unavailable for complex binding with
phytate. Moreover, the organic acids generate an opti- Conclusion

mal pH for the activity of any endogenous phytases
in cereal or legume flour [74] and may also lower Enhancing the complementary diets of breastfed
the gastric emptying rate [75]. The latter effect will infants and young children by the addition of red meat
thus theoretically increase the exposure of zinc to the or liver can have a positive impact on their growth
proximal intestinal epithelium, thereby potentially and some aspects of development, even though an
increasing zinc absorption. However, the magnitude improvement in biochemical zinc status per se may
of the enhancing effect, if any, of organic acids on zinc not always be evident. Strategies designed to enhance
absorption is unknown. Most of the evidence has been zinc absorption through phytate reduction or use of
based on in vitro dialyzability of zinc [76] and needs organic acids alone will not be efficacious in terms of
to be confirmed by in vivo zinc isotope absorption improving zinc nutriture in infants or young children
studies. unless a substantial amount of red meat or liver is also
Zinc absorption may also be improved by increasing included in their diets.
the amount of dietary protein, the magnitude of the
effect being enhanced when animal proteins from meat Detailed review of evidence
[77, 78] or cow’s milk or yogurt [79] are components of
the diet. Hence, in the Malawian studies [71–73], intake The first part of this conclusion is based on a review
of animal-source foods was also promoted, because of five interventions that either supplied or promoted
theoretically, increasing the content of animal-source the consumption of animal-source foods. Of these,
foods in predominantly high-phytate plant-based diets four were nonblinded, randomized, controlled trials
should not only increase the content of zinc ingested conducted in Ghana, the United States, Denmark (table
but also enhance FAZ. However, the magnitude of the 2D) [37–39], and Peru (table 2C) [35], whereas the
increase in FAZ will also depend on the adequacy of fifth was a study in China based on a quasi-experimen-
the baseline zinc intake, since zinc absorption is dose- tal design (table 2C) [34]. In the randomized, control-
dependent [80]. led trials in the United States [38] and Peru [35] and
In the studies in rural southern Malawi among the quasi-experimental study in China [34], significant
infants and young children, zinc absorption was not increases in growth were reported in the intervention
measured. A quasi-experimental design with a nonex- compared with the control group, although in the US
perimental control group was used, and the strategies study, the increase was in head circumference and not
were promoted through nutrition education and social length or weight gain, as in the other two studies. In
marketing (table 3D) [70–73]. Intakes of available the US [38] and Peru [35] studies, greater increases
zinc were higher (p < .001) in the intervention than in intakes of both animal-source foods and zinc in
the control groups postintervention, both in the study the intervention than in the control groups were also
on infants [73] and in the study on young children reported. Changes in nutrient intakes were not assessed
[70, 71] as a result of higher intakes of animal-source in the Chinese study [34].
foods (mainly small, whole, soft-boned fish), phytate It is of interest that in the two studies on infants in
reduction strategies, and use of germinated cereals to China and Peru (table 2C) [34, 35], in which animal-
enhance the energy and nutrient contents of the maize- source foods were not supplied but their consumption
based porridges. Nevertheless, the median intake of was vigorously promoted by a strong nutrition educa-
zinc in the intervention group for the weanling study tion component, marked increases in weight and length
did not meet the age-specific WHO estimated needs gain were reported in the intervention compared with
for infants [46], whereas among the young children, the control groups.
26% in the intervention group and 44% in the control These findings are in contrast to those of the rand-
group had inadequate intakes of zinc at the end of the omized, controlled trial conducted in Danish infants
intervention (p = .002) [71]. However, the results of (table 2D)[39], but the intervention period of this
the latter study are limited by the absence of baseline trial was short (2 months), the number of infants was
intake data due to unforeseen circumstances, so that no small (41 in total), and no details of compliance were
adjustments for baseline differences in intakes could reported. There were also no improvements in zinc
be made [71]. intake or serum zinc in the intervention compared with
the control group in the Danish trial. Likewise, in the
Ghanaian trial (table 2D) [37], no increase in infant
Section 2 growth was reported when fish powder was used to
enrich fermented maize-based or cereal-legume-based
Can complementary foods nutritionally improved complementary foods. In this study, although there was
through dietary diversification or modification strate- no control group, growth was significantly greater in
gies have an impact on the zinc status and the health and all the treatment groups than in the nonintervention
development of breastfed infants and young children? group (p < .001).
TABLE 3A. Isotope studies investigating the effect of phytate on zinc absorption S126
Country, year No. and age of Inclusion
[reference] subjects criteria Design Methods Results
Sweden,1984 58 subjects Healthy, zinc- Adults fed either soy isolate formula Radioisotope technique using whole- Apparent zinc absorption via 65Zn from
[48] Lönnerdal aged 20–30 replete adults (phytate:zinc, 6:1), soy flour (10:1), body counting of radioisotope 65Zn, 14 soy isolate: 14.0% ± 1.0%; soy flour:
et al. yr cow’s milk formula (0 phytate), cow’s days after intake of test meal 8.2% ± 1.5%; cow’s milk: 32.2% ± 1.4%;
milk + phytate level 1 (6:1), or cow’s cow’s milk + phytate level 1: 15.8%
milk + phytate level 2 (19:1) ± 0.8%; cow’s milk + phytate level 2:
9.4% ± 0.8%
USA, 1984 4 men aged Healthy young Crossover design. 63-day metabolic Zinc absorption determined by isotopic FAZ 34% in basal diet, vs. 33.8% with
[49] Turnlund 25–32 yr men study comparing basal diet, basal diet ratio of 67Zn/70Zn from complete fecal addition of α-cellulose (NSD) and
et al. + α-cellulose (34.4 g/day), or basal diet and urine excretion over 63 days 17.5% with addition of phytate (p < .02)
+ phytate (2.34 g phytate/day)
Sweden, 1985 42 adults (17 Healthy, zinc- Reduction of phytate content of bran by Bread dough fermented/leavened for 15 60% decrease in phytate with 2 h fer-
[50] Nävert M, 25 F) aged replete, leavening/fermentation of bread con- min, 45 min, 3 h, or 16 h; phytic acid mentation; maximum decrease (80%
and 20–52 yr nonpreg- taining bran concentration and phytate:zinc molar –85%) after 2 days of fermentation
Sandström nant adults Zinc absorption measured with different ratios determined After fermentation, phytate:zinc molar
without gas- phytic acid concentrations Apparent zinc absorption from bread ratio and FAZ were 16 and 9.6%,
trointestinal 10–14-day metabolic balance period measured by whole-body retention of respectively, after 15 min; 16 and
disorders radioisotope (65Zn) 10–14 days after 11.9% after 45 min; 8 and 14.7% after 3
intake of test meal h; and 4 and 19.8% after 16 h
Fermentation of bread containing bran
reduces its phytic acid content and
increases FAZ
USA, 1988 10 infant Nursing infant Animal models used to study the effect Formula dephytinized through extrac- FAZ in monkeys: 27% from regular soy
[51] Lönnerdal rhesus mon- monkeys and of phytate removal on zinc absorption tion and precipitation processa formula vs. 45% from dephytinized soy
et al. keys aged nursing rat from soy formula. Crossover design Zinc absorption determined by whole- formula. FAZ in rats: 16% from regular
1.5 mo and pups (for monkeys). Animals fed regular soy body counting; 14 days postintubation soy formula vs. 47% from dephytinized
30 rat pups, formula (0.621 mmol/L phytic acid) for monkeys, 6 h for rats soy
aged 14 days or dephytinized soy formula (0.067 Low bioavailability of zinc from soy
mmol/L) formula is a function of its phytate
14-day washout between test meals concentration and can be overcome by
removal of phytate
Denmark, 1989 33 adults (13 Healthy, non- Adults fed crispbread with normal or Crispbread: 20% bran, 60% starch, 10% Normal crispbread: phytate:zinc molar
[52] Kivistö M, 20 F) aged pregnant, reduced phytate content (part of larger gluten ratio, 22; FAZ = 6.2%
et al. 21–42 yr zinc-replete study on nutritional impacts of extru- Apparent zinc absorption from test meal Phytate-reduced crispbread: phytate:zinc
adults, with- sion methods). Meals served after measured by whole-body retention molar ratio, 2.8; FAZ = 17.6%
out known overnight fast. 10–14 days after test meal, oral and Higher FAZ with reduced phytate content
gastrointesti- 10–14-day metabolic balance period intravenous 65Zn of bran
nal disorders
USA, 2006 38 children Healthy, Test meal made from either beef or soy Zinc absorption calculated by multi- Mean ± SD FAZ from beef meal (13.7 ±
[53] Etch- (15 M, 23 F) non-allergy- protein tracer stable isotope techniques; ratios 6.0%) greater than that from soy meal
everry et al. aged 4–8 yr suffering 1 day test meal, urine collection at 48 h of oral (67Zn) and intravenous (70Zn) (10.1 ± 4.1%), p = .047. Even low-
children, not Note: Beef meal had 135 mg phytate per urinary excretion of zinc tracers 48 h phytate soy protein inhibits absorption
taking medi- serving, soy meal had 144 mg phytate after dosing in comparison to meat; 35% enhanced
cation or per serving FAZ from beef protein meal compared
supplements with low-phytate soy
at time of
study
FAZ, fractional absorption of zinc; NSD, nonsignificant difference
a. Puski G, Harman GH, Talbott RD, inventors; Mead Johnson and Co, assignee. Process for preparing low phytate soy protein isolate. US Patent 4,697,004. 1987 Sept 29.5p. Int Cl4 A23J1 1/14.
TABLE 3B. Isotope studies investigating the effect of phytate reduction by phytase enzymes on zinc absorption
Country, year
[reference] Subjects Inclusion criteria Design Methods Results
USA, 2004 9 infants (5 M, 4 Healthy, full-term Crossover design. Infants fed regu- Soybean protein isolate dephyti- Apparent zinc absorption measured
[54] Davidsson F) aged 69–191 infants, formula- lar soy formula (250–300 mg/kg nized with the use of phytase from by 70Zn isotope: 16.7% from regu-
et al. days fed since birth phytate) or dephytinized (< 6 mg/ Aspergillus niger lar vs. 22.6% from dephytinized
kg phytate) formula in random Apparent zinc absorption measured soy formula (p = .03) (95% CI,
order by stable isotope 70Zn via 72-h 0.9–10.8 )

Each study included a 10-day wash- fecal excretion of nonabsorbed
out period followed by a 3-day stable isotope and by chemical bal-
metabolic balance period ance technique
Malawi, 2000 23 children (11 Recruited at local Children randomly assigned to Corn-soy porridge dephytinized No significant effect of phytate
[55] Manary et al. M, 12 F) aged hospital (patients either reduced phytate corn-plus- with the use of phytase from A. reduction in well children (n = 9);
3–13 yr recovering from soy diet or standard high-phytate niger in children recovering from tuber-
tuberculosis [n = corn-plus-soy diet were served Oral and intravenous zinc stable culosis (n = 14), dietary phytate
14] or with minor 5 times per day for 37 days, after isotopes 67Zn and 70Zn and 7-day reduction resulted in higher FAZ
injuries, those which the children participated in urine and stool collections compared to high-phytate corn-
awaiting elective a 7-day zinc stable isotope study plus-soy diet (0.41 ± 0.14 vs. 0.24
surgery, or healthy ± 0.09, p < .05) and TAZ (169 ± 55
siblings (i.e., well vs. 100 ± 46 µg/kg/day, p < .05)
children [n = 9])
Switzerland,2004 9 adults (2 M, 7 Healthy, non- Crossover design. Adults were fed Wheat- and soy-based complemen- Significantly higher apparent zinc
[56] Egli et al. F) aged 22–34 pregnant adults, processed, dephytinized wheat- tary food dephytinized through use absorption from dephytinized
yr without gastroin- and soy-based complementary of native phytases vs. unprocessed complementary food than from
testinal disorders, food (< 0.03 mg/g phytate) or (native phytic acid concentration) unprocessed food with native
receiving no med- unprocessed wheat- and soy-based Apparent fractional zinc absorption phytic acid concentration: FAZ =
ication (other than complementary food (4 mg/g measured by stable isotope 70Zn 34.6 ± 8.0% vs. 22.8 ± 8.8% (p =
oral contraceptive phytate) from 7-day fecal excretion .005)
agents) 2–4-wk period between foods
S127
FAZ, fractional absorption of zinc; Phy:Zn, phytate to zinc molar ratio; TAZ, total absorption of zinc
S128
TABLE 3C. Isotope studies investigating the effect of phytate reduction by the use of low-phytate maize on zinc absorption
Country, year
[reference] Subjects Inclusion criteria Design Methods Results
USA, 2002 5 adults (2 Healthy adults, Crossover design. Adults fed regular Polenta prepared from low-phytate Phytate:zinc molar ratio 17:1 and FAZ
[60] Adams M, 3 F) not consuming maize polenta or low-phytate maize mutant maize (lpa1-l) (60% phytate 0.30 ± 0.13 for lpa1-1; 36:1 and 0.17 ±
et al. aged 24–39 supplements at polenta in random order as their only reduction) or wild-type, normal 0.11 for wild type maize (p < .005)
yr time of study food for 2 days (70Zn for day 1 and phytate-level maize Substitution of a low- phytate grain in a
67Zn for day 2) FAZ measured by stable isotope 70Zn maize-based diet is associated with a
and 67Zn from fecal samples collected 78% increase in zinc absorption
for 8–11 days (8 samples per subject)
USA, 2004 10 adults (6 Healthy adults Crossover design (different maize eaten Group A: lpal-1-LP (60% phytate Mean ( ± SD) FAZ values for tortillas
[61] Hambidge M, 4 F) habitually con- each day). Group A: 6 adults fed tor- reduction), and ND-LP (80% phytate prepared from ND-LP, lpal-l-LP, lpa1-
et al. aged 21–37 suming a non- tillas prepared from 1 of 2 types of reduction). Group B: corresponding 1-WT, and ND-WT were 0.38 ± 0.07,
yr vegetarian diet low-phytate mutant maize. Group B: wild-types; lpal-1-WT (wild type), and 0.28 ± 0.04, 0.15 ± 0.07, and 0.13 ±
4 adults fed 1 of 2 types of wild-type ND-WT 0.05, respectively. A negative correla-
maize. Comparison of overall FAZ for All test meals were extrinsically labeled tion existed between FAZ and dietary
each low-phytate and corresponding with 67Zn on day 1 and 70Zn on day phytate and phytate:zinc (p < .001)
wild-type in relation to their dietary 2. FAZ was measured by the dual- FAZ from maize tortillas is positively
phytate and phytate:zinc molar ratios isotope-tracer ratio technique. A third related to the extent of phytate reduc-
Each subject was fed 1 maize hybrid tracer (68Zn) was administered intra- tion achieved with low-phytate
for 1 day and then the other maize venously during the evening of day 1. hybrids
hybrid on the following day, the order Timed urine samples were collected
alternating between subjects for group twice daily from days 4 to 9, after
A and group B. The 2-day test meal administration of 1st tracer
study period was followed by a 9-day
metabolic balance period
Guatemala, 60 children Healthy children, Cross-sectional, free-living study. Chil- Test day: duplicate diets collected for Mean ( ± SD) FAZ values for lpa1-1,
2006 (29 M, 31 family willing dren randomly assigned to consume analysis of zinc and phytate content 0.32 ± 0.07, vs. WT (wild-type), 0.28 ±
[62] Maz- F) aged to consume either low-phytate maize (lpa1-1) Note: Low power because of high inter- 0.07, vs. local maize, 0.29 ± 0.06.(NS)
ariegos et al. 6–11 yr randomly (~60% phytate reduction) (n = 20), and intragroup differences in phytate FAZ was not increased by the long-term
assigned study wild-type isohybrid (WT) (n = 20), or intake use of low-phytate maize in children
maize provided locally grown maize (L) (n = 20) for 10 whose major dietary staple is maize,
wk. In final week, 1-day stable isotopic possibly because of unexpected
testing by dual isotope ratio technique increase in zinc intakes
as in Hambidge et al. [61]
FAZ, fractional absorption of zinc; lpal-1-LP, lpal-1-low-phytate; lpal-1-WT, lpal-1-wild-type; ND-LP, Nutridense low-phytate; ND-WT, Nutridense wild-type isohybrid; NS, not significant
Additional functional health outcomes that have in biochemical zinc status may not always be evident
been investigated include developmental outcomes in (tables 2D and 3D). Such strategies should simultane-
the US study (table 2D) [38] and morbidity in both ously enhance biochemical vitamin B12 status, and
the Danish [39] and the Ghanaian [37] studies. The possibly iron and hemoglobin status, depending on the
US infants fed pureed beef tended to have a higher baseline iron status of the study group. Among women
behavior index score than their counterparts fed forti- of reproductive age, specific data on zinc status are very
fied rice cereal [38], whereas no significant differences limited. Depending on their initial nutritional status,
in morbidity were observed between the groups in the increases in the content of animal-source foods in the
Danish [39] and Ghanaian [37] studies. diet may in some circumstances increase serum zinc,
It is of interest that improvements in growth, health, maintain or even enhance biochemical iron status,
or development were not reported in the intervention improve hemoglobin, and reduce anemia prevalence
as compared with the control groups in two rand- in young women of reproductive age.
omized, controlled trials of complementary feeding
in Sweden and Tanzania (table 3D), both of which Detailed review of evidence
were based on phytate reduction strategies without
any increase in the intake of animal-source foods. In The conclusion related to children is based on one
the Swedish trial (table 3D) [68, 69], consumption of cluster-randomized, controlled trial (not blinded)
phytate-reduced products (achieved by a commercial conducted on Kenyan schoolchildren (n = 554; median
hydrothermal process using endogenous phytases) by age, 7.4 years) (table 2D) [40–45] and one quasi-
Swedish infants from 6 to 12 months of age had no experimentally controlled study with a nonequivalent
effect on physical growth, development, morbidity, control group undertaken on rural Malawian children
or biochemical zinc status compared with the control (n = 281) aged 3 to 7 years (table 3D) [70–72]. In both
group. However, only 22% of the infants overall were studies [40, 72], improvement in muscle mass (based
zinc-deficient at baseline (serum zinc < 10.7 µmol/L), on anthropometry) was observed in the intervention
and the mean dietary phytate:zinc molar ratio, even group, which was reported to have a higher intake of
for the control group, was below that said to markedly cellular animal protein than the controls, although no
compromise zinc absorption (< 10:1). In the trial in improvement in biochemical zinc status was apparent
Tanzania [66, 67], the high phytate:zinc molar ratio in any group. In the Kenyan study (table 2D), improve-
of the complementary food in the treatment group, ments in body weight [40] (but not height) and in
even after processing, may be responsible, at least in certain domains of cognitive functioning [45] were
part, for the lack of improvement in growth [66, 67], also reported in the meat-based intervention group
hair zinc level [66], hemoglobin concentration, or zinc compared with the control group. In contrast, in a
protoporphyrin concentration [67] in the intervention subgroup of the Kenyan children with lower baseline
compared with the control group. The results of the height-for-age z-score (HAZ) (≤ −1.4) who received
Tanzanian study suggest that phytate reduction strate- the milk-based snack, an improvement in height was
gies alone are not sufficient to ensure that intakes of observed compared with the controls [40, 41]. An addi-
bioavailable zinc (and iron) meet the needs for optimal tional reported positive effect of these interventions
growth and prevention of anemia in infants fed pre- was improvement in biochemical vitamin B12 status
dominantly plant-based complementary diets. in Kenyan children receiving the meat- or milk-based
snack compared with controls [44]; furthermore, in
Malawian children, despite a fall in hemoglobin of 7 g/L
Section 3 in the control group, there was no significant change in
hemoglobin in the intervention group after 1 year, and
Can foods nutritionally improved through dietary diver- the incidence rates of anemia and common infections
sification or modification have an impact on the zinc were lower than in the control group, after adjustment
status and the health and development of children and for baseline differences [72]. Increases in intakes of
women of reproductive age? several other micronutrients in the intervention com-
pared with the control group were also apparent in
Conclusion these two studies.
The conclusion for women of reproductive age
Enhancing the diets of children in the household is based on two randomized, controlled trials (one
by the addition of animal-source foods to the diet cluster-randomized trial undertaken in Peru [33] and
or promotion of their use, with or without phytate another trial in New Zealand [36]) and a third trial with
reduction strategies, can have positive impacts on a quasi-experimental design, also conducted in Peru
body composition, growth, cognitive functioning, and (table 2C) [32]. All three interventions emphasized
intakes of bioavailable zinc, depending on the source nutrition education and behavior change to promote
of animal-source foods, even though improvements the consumption of animal-source foods. The Peruvian
TABLE 3D. Mixed interventions: Phytate reduction strategies with and without other household methods to enhance
content and bioavailability of dietary zinc
Intervention strategies
Country, year [reference] Production Nutrition education Target groups
Tanzania, 2006, 2004 Processed complementary food No Infants aged 6 mo at baseline
[66] Lachat et al., [67] Mamiro (PCF) with reduced phytate (n = 309)
et al. content: mean ( ± SD) 660 ± 20
mg/100 g DM prepared from
germinated millet, kidney beans,
roasted peanuts, mango puree;
Unprocessed complementary
food (U-PCF) with phytate con-
tent: 1150 ± 30 mg/100 g DM;
Phytate:zinc molar ratios: 25.8 for
PCF vs. 47.5 for U-PCF; moth-
ers added oil to PCF and U-PCF
Sweden 2003, 2004 Phytate-reduced (PR) milk-based No Infants aged 6-mo at baseline
[68, 69] Lind et al. cereal drink (MCD) (PRMCD) (n = 300)
and phytate-reduced porridge
(PRP)
Malawi, 2003, 2002, 2005 Distribution of sunflower, pigeon Social marketing Children aged 3–7 yr from
[70] Gibson et al. pea, soybean, and papaya Face-to-face communication subsistence farming
[71, 72] Yeudall et al. seedlings Nutrition education to households
Solar drying increase intakes of iron,
Agricultural training zinc, and vitamin A and
Food preparation and processing modify food-processing
to reduce phytate content techniques
Health education
Malawi, 2005 Food preparation and processing Nutrition education to Mothers and their breastfed
[73] Hotz and Gibson to increase energy and nutrient increase intakes of iron, children aged 6–23 mo
density and decrease phytate zinc, and vitamin A
content of complementary Food-processing techniques,
porridges health education
AMA, arm muscle area; MUAC, mid-upper-arm circumference; NA, not available
Outcomes
Design Methods KAP + dietary intake Nutritional status
Double-blind, randomized, Anthropometry, malaria No differences in intakes of No differences in growth,
controlled trial. blood smear, hemoglobin, energy, protein, fat, or total hemoglobin, zinc protopor-
Intervention (n = 157) vs. con- zinc protoporphyrin, hair iron. Intake of soluble zinc was phyrin, or hair zinc between
trols (n = 152) zinc, and 24-h recall (on a higher in infants consuming 2 groups at 12 mo of age
Infants supplied with PCF or subsample [n = 137]) PCF than in those consuming
U-PCF Mothers instructed to pre- U-PCF (0.164 vs. 0.0684 mg/
6-mo follow-up (at 12 mo pare similar amounts of CF day)
of age) every day (104 g DM/d)
Double-blind, randomized, 6- and 12-mo blood samples: No difference in energy intakes No differences in serum fer-
controlled trial until 12 mo. serum ferritin, serum zinc, among groups at 6–8 or 9–11 ritin or zinc between CC and
Followed to 18 mo hemoglobin mo PR groups
3 groups: commercial MCD Monthly dietary assessment Protein, iron, zinc (at 6 mo Hemoglobin lower (117
and porridge (CC group) (5-day food diaries) only), vitamin C, calcium, and vs. 120 g/L, p = .012) and
vs. PRMCD + PRP (PR Anthropometry: monthly phytate intakes differed in IF anemia prevalence higher
group), or milk-based from 6 to 12 mo; bimonthly and MCD groups (p < .05) (23% vs. 13%. p = .06) in IF
infant formula + porridge until 18 mo because of different nutrient than in PR group because
(IF group) Bayley Scales of Infant Devel- contents of IF and MCD (p < of differences in daily iron
opment. at 7, 13, and 18 mo .05) intake in 2 groups
Morbidity daily No differences between CC and
PR groups in growth, develop-
ment at any time, or morbidity
at 6–11 or 12–17 mo of age
Quasi-experimental design Focus groups, interviews, Intervention parents had greater No difference in growth in
with nonequivalent control observations, anthropom- knowledge of iron, vitamin A, intervention vs. controls at
group etry, 24-h recall over 2 sea- and food sources of iron (p < postintervention
2 intervention villages (n = sons, hemoglobin, hair zinc, .05); greater use of fermented MUAC and AMA z-score
200) vs. 1 control village (n malaria slides, morbidity maize flour (p < .001); greater higher in intervention
= 81) from picture calendars median intake of animal- than controls (p < .001)
12-mo follow-up source foods (especially fish), postintervention
energy, protein, fat, calcium, No change in hemoglobin in
zinc, and vitamin B12; lower intervention group, but sig-
intake of phytate (p < .05); and nificant decrease in hemo-
fewer children with inadequate globin in controls (p < .05)
intakes of protein, calcium, No differences in initial
zinc, vitamin B12, and folate or final hair zinc values
than controls between groups
Note: there was some leakage Incidence of common illnesses
of knowledge and practices to lower in intervention than in
controls control group postinterven-
tion (p < .001)
Quasi-experimental design Focus groups, counseling Adoption rates: 25% (enriched NA
with nonequivalent control visits, single interactive porridges) to 10% (soaked,
group 24-h recall pounded maize)
Intervention group: 3 villages Amount of complementary food
(n = 69) vs. 1 control village (g/day) and intakes of energy,
(n = 40) animal protein, vitamin B2,
niacin, calcium, iron, zinc, bio-
available zinc and iron were
greater in intervention group
than in control group (p < .05)
studies were conducted through local community feeding practices and nutrient adequacy (including
kitchens [32, 33]. Only the New Zealand study meas- zinc) of complementary foods, which in turn resulted
ured zinc intakes and serum and hair zinc concentra- in a dramatic improvement in length and weight
tions, but the results are not yet available. All three gain in the intervention compared with the control
interventions measured hemoglobin [32, 33, 36]; two group at 18 months of age. Moreover, this interven-
also measured serum ferritin [33, 36], although only tion enhanced the quality and coverage of the existing
the New Zealand study [36] accounted for infection. Of nutrition education offered, without providing material
the two interventions that measured serum ferritin [33, benefits to the centers, in an attempt to maximize the
36], the intervention in Peru [33] tended to prevent a sustainability of the program, although to date, its long-
decline in iron status compared with the control group, term sustainability has not been evaluated. Two other
whereas in New Zealand [36], the dietary counseling randomized, controlled trials conducted in Peru (table
program resulted in a nonsignificant increase (p = 2C) [33, 82] and New Zealand (table 2C) [36] that
.068) in serum ferritin in comparison with the placebo targeted adolescents and premenopausal women also
group. Only one of these three interventions reported showed a positive impact on behavior change, based on
an improvement in hemoglobin [32], but these changes an increased intake of animal-source foods, which in
were not compared with those in the “control group,” turn led to increases in intakes of bioavailable iron and
making it difficult to draw any conclusion. improved biochemical iron indices in the intervention
Functional health outcomes associated with zinc compared with the control group. In the New Zealand
deficiency were not assessed in the published studies. trial only, intakes and biochemical and zinc-responsive
functional indicators were also measured, and the data
are currently being analyzed.
Section 4a Further evidence supporting the impact of behavior
change on nutritional status stems from the results of
Do interventions that promote dietary diversification or interventions based on a quasi-experimental design that
modification have an impact on behavior change and on also included a nutrition education or social marketing
nutritional status in both the short and the long term? component. Some of these interventions promoted the
consumption of animal-source foods (n = 2) (table 2C)
Conclusion [32, 34], and others promoted their production and
consumption through small-animal husbandry or
Dietary diversification or modification can have an aquaculture, with (n = 3; table 2A) [18, 19, 21, 22,
impact on behavior change and on certain indicators 24] or without (n = 1; table 2B) [31] agriculture. Two
of nutritional status in the short term, most notably additional studies in Malawi also employed phytate
increases or greater intakes of animal-source foods and reduction strategies together with the promotion of
certain nutrients, including zinc, depending on the set- animal-source foods (table 3D) [70–73]. Of the eight
ting and the study design. Whether the interventions interventions, two interventions in table 2C [32, 34],
also have an impact on biochemical, anthropometric, two in table 2A [21, 22, 24], and one in table 2B [31]
or other functional health outcomes, even over the showed an increase in intake of animal-source foods or
short term, is less clear. The impact depends on the age total iron and heme iron, and in some cases, promising
of the participants, their baseline nutritional and health improvements in nutritional status based on reduc-
status, the duration and type of intervention strategies tions in night-blindness, morbidity, and anemia, and
employed, and the setting. The long-term impact was increases in serum retinol, serum ferritin, hemoglobin,
evaluated in so few of the interventions that a statement growth, and muscle mass in the intervention com-
on this issue cannot be made. pared with the control group. Certainly these findings
suggest some impact on behavior change and nutri-
Detailed review of evidence tional status. Further, theoretically, because a quasi-
experimental design was used, potential confounding
The most convincing evidence for the impact of factors and biases that might have been responsible for
behavior change on nutritional status, specifically zinc the observed effects can be taken into account in the
status, stems from the educational effectiveness trial analyses, allowing a plausibility evaluation to be made
conducted in Peru by Penny et al. [35]; details are sum- [81]. However, in most of these interventions, adjust-
marized in table 2C [35]. The design of this cluster- ments for baseline differences between the intervention
randomized, controlled trial involving six intervention and control groups were not performed, making it dif-
and six control health facilities meets the criteria for ficult to infer that the results observed were due to the
a probability assessment outlined by Habicht and intervention per se. Exceptions are the two studies in
coworkers [81]. The program had a significant effect Malawi (table 3D), which did adjust for most baseline
on behavior change, based on improvements in infant- differences, so that the improvements [73] or higher
intakes [71] of animal-source foods and available zinc Exceptions are the randomized, controlled trial in
achieved postintervention, as well as other indicators Kenya (table 2D) [43] and the cluster-randomized edu-
of nutritional status in the study on children [72], can cational effectiveness intervention conducted in Peru
be associated with the impact of the intervention on (table 2C) [35]. In the Kenyan schoolchildren, those
behavior change in the short term. with a lower baseline HAZ (≤ −1.4) who received the
Only two of these projects examined their long-term milk-based snack showed a significant improvement
impact on behavior change and nutritional status, in height compared with the controls [41]. In the Peru
although no indicators of zinc status were included trial, the socioeconomic characteristics of the inter-
(tables 1 and 2A) [5, 21, 22]. Both were participatory vention and control households differed at baseline
projects in which home gardening was promoted and and were subsequently shown to modify the effects on
agriculture education was included. Follow-up after both dietary intake and growth. As a result, differences
4 to 5 years revealed significantly greater knowledge between the intervention and control groups were
and practices related to vitamin A nutrition and a assessed after adjustment for these covariates [35].
positive effect on intakes of vitamin A–rich foods in Infection may modify the response to an intervention
the intervention group compared with the controls in through several mechanisms. These include impaired
both projects, demonstrating that dietary changes can appetite, and thus reduced dietary intakes, and altera-
be sustained over time. Nevertheless, these dietary tions in the integrity of the intestinal mucosa, causing
changes were not reflected by improvements in serum increases in intestinal permeability and reductions in
retinol values among the intervention group in the nutrient absorption [83]. However, very few of the
Tanzania study [5], and no biochemical or functional interventions promoting dietary modification or diver-
indicators of vitamin A status were measured in the sification (excluding the randomized, controlled trials)
Thai follow-up study [22]. measured infection by using a biochemical marker or
functional indicators of morbidity, as noted earlier.
Exceptions are the Vietnamese study (table 2A) [18,
Section 4b 19], in which morbidity due to acute respiratory infec-
tion was monitored; the Malawian study of the effect
Is the impact modified by baseline nutritional status, of dietary diversification or modification on children
socioeconomic status, infection, sex, age, and life-stage aged 3 to 7 years, in which malaria and morbidity due
group? to common illnesses were assessed (table 3D) [70, 72];
and the 5-year follow-up study in Tanzania outlined
Conclusion above, which examined stool helminth infection in
relation to serum retinol concentrations (table 1) [5].
It is likely that the impact of dietary diversification Unlike serum ferritin, infection reduces serum
or modification interventions is modified by baseline zinc and retinol concentrations [84]. In the 5-year
nutritional status (including birthweight), household follow-up of the food-based intervention in Tanzania
socioeconomic characteristics, age, infection, and pos- (table 1) [5], the mean serum retinol concentration
sibly sex. The evidence for this conclusion, however, is of children in the intervention area was significantly
not strong, because many of the interventions have not lower than that of children in the control areas, but
measured these factors at baseline, and hence whether after adjustment for helminth infection, the means in
they have a modifying effect has not been extensively the two areas were no longer significantly different.
investigated. Serum zinc was not measured in this intervention.
These findings highlight the importance of taking into
Detailed review of evidence account the existence of infection in the interpretation
of some biochemical indicators, notably serum ferritin,
Evidence from micronutrient supplementation studies retinol, and zinc. Indeed, this approach was adopted for
confirms that the magnitude of the response will be the randomized, controlled trial intervention studies
dependent on baseline nutritional status (including conducted on infants in Ghana (table 2D) [37] and the
zinc status). However, in most of the interventions United States (table 2D) [38] and on schoolchildren
reviewed here, with the exception of the randomized, in Kenya (table 2D) [44], but not in the randomized,
controlled trials, no comprehensive assessment of the controlled trial conducted on adolescents in Peru (table
nutritional status (including zinc) of the intervention 2C) [33]. Indeed, in the trial conducted on Kenyan
and control groups at baseline has been reported. Con- schoolchildren (table 2D), at baseline, across all groups
sequently, no baseline nutritional status variables could and independently of group assignment, children with
be used as covariates in the analysis. Hence, whether Entamoeba histolytica infection had a smaller increase
the impact of the intervention has been modified by in serum iron (p = .01) but a greater increase in plasma
baseline nutritional status is generally unknown. retinol (p = .01), whereas those with an enlarged spleen
had a smaller increase in plasma vitamin B12 [44]. all zones in West Africa is reported to be at risk for
Age can also modify the impact of dietary diversifica- aflatoxin contamination [86]. Germinated cereals
tion or modification. In the quasi-experimentally con- are vulnerable to aflatoxin contamination because
trolled study undertaken on Chinese infants (table 2C) of the extra moisture required for germination and
[34], age had a significant effect on the growth outcome. improper drying and storage conditions. Unfortunately,
Before 12 months, there was no difference in growth no household strategies exist to decontaminate plant
between the two groups, but at 12 months, infants in foods once they are contaminated with aflatoxins; heat-
the education group had better growth in weight and ing and fermentation do not destroy them. However,
length (p < .05) than the controls, and fewer had mod- research has consistently shown that if germinated
erate and severe malnutrition (weight-for-age z-score cereals are prepared appropriately, dried, and then
[WAZ] or HAZ < –2). Whether these differences in the stored in a covered container in the household to pre-
growth response were due to differences in age or the vent spoilage by insects, aflatoxin contamination can
length of the intervention is not clear. be avoided [87, 88].
Soaking maize and legume flours to remove water-
soluble sodium and potassium phytate by diffusion
Section 5 was one of the phytate reduction strategies used in
the Malawian interventions (table 3D) [70–72]. Soak-
Are there any adverse effects of dietary diversification ing maize flour may result in some loss of zinc and
or modification? water-soluble vitamins, such as thiamine, niacin, and
riboflavin (n = 6) [88, 89], but the losses are small and
Conclusion are not expected to exceed the amounts that are lost
during cooking or incurred when unrefined maize is
Existing evidence suggests that provided care is taken milled [59]. If microbiologically unsafe water is used
to ensure that germinated cereals are prepared with for soaking, there may be cause for concern, although
microbiologically safe water, free of any environmen- enteropathogenic microorganisms should be destroyed
tal contaminants, and stored appropriately, they can during the cooking of the cereal porridges. Care must
be used safely to enhance the energy and nutrient be taken, however, to avoid the use of water for soaking
density and reduce the phytate content of unrefined that may have been subject to environmental contami-
cereal-based porridges without risk of aflatoxin con- nation. There is also some evidence that soaking maize
tamination or environmental contamination from the flour during fermentation and removal of the excess
water. Further, although soaking cereal flours may water may result in loss of water-soluble antimicrobial
result in some loss of zinc and water-soluble vitamins, substances [90].
such losses are small and are unlikely to override the Care must be taken to ensure that the promotion
positive effect of increased zinc absorption arising of dietary diversification or modification to improve
from the loss of water-soluble phytate from unrefined complementary diets does not result in excessive
cereal-diets, provided care is taken to ensure that the displacement of breastmilk. Studies have reported an
soaking water is microbiologically safe and not subject inverse relationship between the intake of breastmilk
to environmental contamination. Care must be taken, and the intake of complementary foods [91, 92],
however, to ensure that when dietary diversification although for infants over 6 months of age the propor-
or modification strategies are promoted to enhance tion of breastmilk displaced by complementary foods is
complementary feeding, displacement of breastmilk is probably less than that for younger infants. Moreover,
minimized by sustaining optimal breastmilk intake. it appears that displacement of breastmilk is less with
those interventions that are designed to increase the
Detailed review of evidence energy density of complementary foods (e.g., by the
use of germinated cereals) rather than meal frequency.
The use of germinated cereals is often promoted as a Nevertheless, breastmilk is not an irreplaceable source
dietary modification strategy to enhance the energy of zinc for older infants, because of its low zinc content
and nutrient density and reduce the phytate content of from 6 months postpartum. The decline in breastmilk
unrefined cereal-based porridges used for infant and zinc content is independent of maternal zinc status [46]
young child feeding [47]. Concern has been expressed and is seldom balanced by an increase in breastmilk
over the susceptibility of germinated cereals to con- intake. Instead, in later infancy, complementary foods
tamination from aflatoxins produced by Aspergillus are responsible for the increase in total zinc intakes that
flavus, A. parasiticus, and A. nomius, because aflatoxins occur. Unless animal-source foods are also consumed,
are potent carcinogens. Plant foods most susceptible to however, the actual increases in absorbable zinc will be
aflatoxin contamination include peanuts, seeds, and small, because most complementary diets in develop-
cereal grains such as wheat and maize [85]. Indeed, ing countries are based predominantly on unrefined,
maize in the field and in traditional maize stores in high-phytate-containing cereals [93].
Scaling up dietary diversification or modification practices, and had a higher involvement of women in
interventions decision making than control households. However,
changes in zinc intakes and status per se were not
There are very few examples of dietary diversification or evaluated in this study. Comparable data from the
modification interventions that have been implemented Helen Keller International HFP programs in Nepal and
as programs. In 2002, Helen Keller International began Cambodia have not yet been published. A decline in
a pilot project on Homestead Food Production (HFP) the prevalence of anemia among nonpregnant women
in Bangladesh [94, 95]. This new program integrated and children from HFP households compared with
animal husbandry into the Helen Keller International controls was also reported in Nepal and Bangladesh,
home gardening and nutrition education programs but not in Cambodia. Whether this effect is due to
that had been ongoing in Bangladesh since 1990. The an increased consumption of animal-source foods
aim of the new integrated HFP program is to further among the target HFP households in Nepal and Bang-
enhance the production, availability, and consumption ladesh is uncertain, based on the analyses presented
of animal-source foods by women and children. Nutri- to date [24, 94].
tion education is an integral part of the program, and
nutrition messages are designed specifically to increase Implications for implementing dietary diversification
the consumption of eggs, meat, liver, and milk, as well or modification programs
as micronutrient-rich foods from plant sources. This
integrated HFP program has now been expanded to On the basis of the available scientific evidence
other areas within Bangladesh, as well as into Nepal, reviewed for Section 1a and 1b, dietary diversification
Cambodia, and, more recently, the Philippines. In all or modification strategies employing agricultural inter-
these countries, Helen Keller International works in ventions, animal husbandry, or aquaculture have the
close partnership with other nongovernmental organi- potential to increase intakes of total or absorbable zinc,
zations at the household level to provide overall guid- although the magnitude of the increase will depend
ance and training in the integrated HFP program. on the type of intervention, the habitual diet and life
To date, Helen Keller International has conducted stage of the target group, and whether nutrition edu-
preassessment and baseline surveys in all the program cation or behavior change is included as a component
sites. Cross-sectional data have been collected from of the intervention. To date, none of the agricultural
HFP program households and from nonprogram interventions (Section 1a) and very few of those based
households (controls) selected through multistage on animal husbandry or aquaculture, with or without
cluster sampling at baseline and for Bangladesh, Nepal, agriculture (Section 1b), have addressed zinc as a
and Cambodia at 3 to 4 years postintervention [94–96]. target nutrient, so that data on total or absorbable zinc
Data collected include sociodemographic status, home- intakes are very limited. This is unfortunate, because
stead food-production practices (including animal and such data could be obtained, together with intakes of
plant food production), household income generation food groups and other micronutrients, at no additional
and expenditure, and nutritional status and food con- cost. In several of the interventions reviewed, postin-
sumption of mothers and their children. Food con- tervention increases or higher intakes were reported of
sumption was assessed by a 7-day food recall [96] or animal-source foods, energy, total protein, and other
with the use of the Helen Keller International 24-hour micronutrients, notably vitamin A, iron, and vitamin
vitamin A semiquantitative (VASQ) method to assess C. Whether improvements in intakes of animal-source
vitamin A intakes [97]. In Bangladesh, an indicator of foods would be sufficient to meet the needs of absorb-
household vulnerability was created based on the exist- able zinc for infants and young children is questionable,
ence of anemia (hemoglobin < 120 g/L for nonpregnant given their small gastric capacity [46]. For this age
women and < 100 g/L for children < 5 years) and a low group, a combination of intervention strategies, includ-
intake of vitamin A (defined as intake of retinol activity ing home fortification of complementary foods with
equivalents less than the recommended nutrient intake) micronutrients, is probably more appropriate in most
of the mother, her youngest child, or both [95]. settings, provided their effectiveness and sustainability
The results of the evaluation of the HFP program have been confirmed and they are readily accessible to
in Bangladesh showed significantly higher intakes of poor-resource households.
animal-source foods, specifically eggs, poultry, other Increases in absorbable zinc resulting from agri-
meat, milk or milk products, and dark-green leafy cultural interventions alone in persons consuming
vegetables, at endline compared with baseline among predominantly unrefined or unleavened cereal-based
children aged 6 to 59 months from target households, diets are likely to be negligible (Section 1a), unless
whereas intakes in the control group were almost strategies that encompass phytate reduction (with or
unchanged [94, 95]. In addition, more of the target without the inclusion of absorption enhancers) are
households earned income from gardening and poultry included. The increase in absorbable zinc that might
activities, showed improvements in poultry-rearing be achieved through household phytate reduction
strategies in individuals consuming unrefined or provided or promoted through nutrition education and
unleavened cereal-based diets has not been measured social marketing, even though no positive response in
by in vivo stable isotope studies (Section 1c). However biochemical zinc status was observed. More rigorously
the level of household phytate reduction that can be designed efficacy trials in children are needed to con-
achieved approximates that achieved with the use firm these suggestions.
of low-phytate maize from plant breeding. Signifi- Whether dietary diversification or modification
cant increases in fractional zinc absorption from test results in a positive impact on biochemical zinc status
meals of low-phytate maize compared with that from and zinc-related functional outcomes for women
the corresponding isohybrid wild-type maize with of premenopausal age is unclear. Again, rigorously
a typical phytate content have been confirmed from designed efficacy trials with an appropriate design
stable isotope studies. The magnitude of the increase and sample size, which include measurements of both
in fractional zinc absorption depends on the level potential confounders and appropriate outcomes, are
of phytate reduction and the total zinc intake of the needed to ensure that results can be attributed to the
target group. intervention.
Interpretation of the results for the interventions Evidence for the sustainability (both short-term
reviewed in Sections 1a–c is hampered by limitations and long-term) of the behavior change or changes in
in both the intervention design and the quality of the nutritional status associated with the promotion of
data collected, as well as omission of several key meas- dietary diversification or modification was examined in
urements (e.g., data on zinc intakes), emphasizing the Section 4a. Although some positive effects on behavior
importance of conducting rigorously designed inter- change and nutritional status over the short term have
ventions based on agriculture, small-animal husbandry, been documented, the long-term impact has seldom
or aquaculture. been evaluated, so no conclusion can be drawn. Table 4
Sections 2 and 3 address whether dietary diversifica- provides a summary of the interventions reviewed and
tion or modification strategies can achieve increases in tabulated in tables 1 to 3D, together with a ranking
absorbable zinc that are sufficient to enhance serum compiled by the authors of the conclusions of each
zinc or zinc-related functional responses in breastfed study according to three levels of confidence (high,
infants and young children (Section 2) and in older medium, or low), based on the methods and data
children and women of reproductive age (Section 3). presented.
From the available scientific evidence on infants and
young children, inclusion of animal protein (especially How to achieve dietary diversification and
liver) in complementary diets appears to be associ- modification interventions
ated with a consistent positive response in growth (in
length, weight, or head circumference), which is pos- A detailed discussion on how to design and implement
sibly a zinc-related functional outcome. Whether there dietary diversification or modification strategies to
is also a response to serum zinc is uncertain, because increase intakes of micronutrients, including absorb-
there are fewer reports of this measurement. The posi- able zinc, is beyond the scope of this review. Several
tive effect on growth is consistent with the results of useful publications on food-based approaches are
a well-designed effectiveness trial in Peru involving available [97–100], and the reader is advised to consult
a complementary feeding educational intervention these sources for more details. A technical brief that
implemented through government health services in summarizes the dietary diversification and modifica-
areas where food availability was not limited. In this tion strategies for enhancing intakes and bioavailability
effectiveness trial, the two key nutrition education mes- of dietary zinc has also been prepared by the Interna-
sages were feeding infants thick purees that contained tional Zinc Nutrition Consultative Group (IZiNCG)
a source of animal protein, specifically chicken liver, [101]. Three of the more detailed publications, however,
egg, or fish, at each meal. Children in the intervention target mainly vitamin A deficiency and include lim-
group had significantly higher intakes of energy from ited information on how to enhance the production,
animal sources and of zinc (as well as iron and calcium) accessibility, or consumption of animal-source foods,
at 9, 12, and/or 18 months and a significant reduction a prerequisite for increasing intakes of absorbable zinc
in the prevalence of stunted growth at 18 months of to levels likely to be associated with zinc-related func-
age than those in the control group. Serum zinc was tional outcomes. Programs that have integrated animal
not measured in this trial. husbandry with home gardening include the integrated
For older children, evidence from a limited number HFP program developed by Helen Keller International,
of trials suggests that an improvement in a zinc-related as noted earlier, and the MICAH Program of World
functional outcome, body composition (i.e., muscle Vision Malawi. The latter emphasized the consump-
mass based on mid-upper-arm muscle area), may be tion of animal-source foods through the creation of a
achieved through dietary diversification or modifica- small-animal revolving fund for rabbits, poultry, guinea
tion strategies that include animal-source foods either fowl, and goats [31]. Community kitchens are another
TABLE 4. Summary of interventions reviewed

Baseline
Baseline Nutrition data for Data Reviewer’s
data for education nutritional collected Dietary Nutritional confidence
production or social status on control intake status in study
Author [reference] indicators marketing indicators population outcomes outcomes conclusions
Table 1. Agricultural interventions to increase household production, accessibility, or consumption of plant-based foods
[4] Marsh + + a+a + + + High
Zinc related
[5] Kidala et al. − + − + + + Medium
[6] Chakravarty + + a+b − + + Low
[7] Ngu et al. + + a+b − + + Low
[8] Solon et al. + + − + + − High
[9] Phillips et al. − + − + + + Medium
[10] Solon et al. , − + a+b − + + Medium
[11] Popkin et al.
[12] Galal et al. + − a+b + + + Low
[13] CARE Nepal − − a+b − + + Low
[14] Hagenimana et al. − + a+a + + + High
Table 2A. Mixed interventions: Agriculture combined with small-animal husbandry or aquaculture to increase
production or promotion of animal-source foods
[18] English et al., [19] − + a+a + + + High
English and Badcock Zinc related
[20] IFPRI + − a+a + + + Medium
[21] Smitasiri and + + a+a + + + High
Dhanamitta, [22]
Smitasiri et al.
[23] Schipani et al. + − a+a + + + High
Zinc related
[24] Stallkamp et al. + + a+a + + + High
[25] Ying et al. + + a+b − − + Low
Zinc related
[26] Sheikholeslam + + a+b + − + Medium
et al. Zinc related
Table 2B. Animal-source food interventions: Small-animal husbandry or aquaculture to increase production
of animal-source foods
[27] Ayele and Peacock + + − − + − Medium
[28] Nielsen et al. + − − + + − Medium
[29, 30] Roos et al. + − − + + − Medium
Table 2C. Animal-source food interventions: Nutrition education or behavior change to promote animal-source foods
[32] Carrasco Sanez − + a+b + + + Medium
et al.
[33] Creed-Kanshiro − + a+a + + + Medium
et al.
[34] Gulden et al. − + a+a + + + Medium
[35] Penny et al. − + a+a + + Zinc related High

[36] Heath et al. − + a+a + + + High
TABLE 4. Summary of interventions reviewed (continued)

Baseline
Baseline Nutrition data for Data Reviewer’s
data for education nutritional collected Dietary Nutritional confidence
production or social status on control intake status in study
Author [reference] indicators marketing indicators population outcomes outcomes conclusions
Table 2D. Nutritional improvement of home-prepared complementary foods or school lunches from inclusion
of animal-source foods
[37] Lartey et al. − − a+a + + + High
Zinc related
[38] Krebs et al. − − a+a + + + High
Zinc related
[39] Engelmann et al. − − +a + + + High
Zinc related
[40, 41] Grillenberger − − a+a + + + High
et al., [42] Murphy et Zinc related
al., [43] Neumann et
al., [44] Siekmann et
al., [45] Whaley et al.
Table 3D. Mixed interventions: Phytate reduction strategies with and without other household methods to enhance
content and bioavailability of dietary zinc
[66] Lachat et al., [67] − − a+a + + Zinc related Medium
Mamiro et al.
[68, 69] Lind et al. − − a+a + + Zinc related High
[70] Gibson et al., [71, − + +a + + Zinc related High
72] Yeudall et al.
[73] Hotz and Gibson − + a+a + + − High
+, present; −, absent
a. Baseline differences between intervention and control groups controlled for in analyses.
b Baseline differences between intervention and control groups not controlled for in analyses.
model that has been successful among women in Lima, [106], and the IZiNCG Technical Report [107]. Indica-
Peru, in increasing the consumption of iron-rich foods tors recommended by WHO, the United Nations Chil-
such as organ meats and thus probably increasing the dren’s Fund (UNICEF), the International Atomic Energy
intake of absorbable zinc; details are given in Carrasco- Authority (IAEA), and IZiNCG to monitor or evaluate
Sanez et al. [32]. risk of zinc deficiency include dietary zinc intake,
Regardless of the model adopted, all interventions plasma or serum zinc concentration, and the prevalence
should be designed, implemented, and evaluated by of stunting [107]; further details on the measurement
a formative research approach and should include of these indicators are available in IZiNCG Technical
appropriate information, education, and communica- Briefs 1, 2, and 3 [108–111].
tion strategies to ensure their adoption, sustainability,
and effectiveness, as described in Dickin et al. [102]. Future research needs
Details on how to train people in participatory methods
are described in Pretty et al. [103]. Methods aimed to In the future, measurements of intakes and major food
strengthen the leadership abilities of community women sources of zinc and phytate should be included in all
include the Cornell Modified Community-Based Nutri- dietary diversification or modification interventions,
tion Monitoring [104] and the Appreciation-Influence- together with energy, protein, animal protein, and
Control Process [105]. Finally, monitoring and evaluation other high-risk micronutrients (e.g., iron, calcium,
should be included as an integral component of any vitamin A, vitamin B6, vitamin B12, and riboflavin).
intervention program so that its performance and impact These measurements can be included at no additional
on the target group can be established. Selection of the cost, provided local food-composition data for zinc and
appropriate indicators and design for the monitoring phytate are available.
and evaluation to meet the program objectives is critical; More rigorously designed multicenter efficacy trials
details can be found in Habicht et al. [81], Levinson et al. in infants, children, and premenopausal women known
to be at high risk for zinc deficiency at baseline (based zinc deficiency to examine the long-term impact of
on a prevalence of stunting greater than 20% and either dietary diversification or modification on behavior
a prevalence of low serum zinc greater than 20% or a change and on zinc-related functional outcomes. These
prevalence of inadequate zinc intake of greater than trials should include the measurement of all confound-
25%) are needed to confirm whether improvements in ing factors that may have the potential to modify their
both serum zinc and zinc-related functional responses, impact and subsequent sustainability.
such as growth, body composition, and morbidity, can Effectiveness trials with designs similar to that of the
be achieved by including animal-source foods, specifi- trial undertaken in Peru should be undertaken in other
cally red meat or liver, in their diets. The animal-source settings where stunting rates are greater than 20% to
foods can be supplied to the participants in the trials or confirm whether nutrition education messages to pro-
their use can be promoted through nutrition education mote the inclusion of animal-source foods (appropriate
and behavior change. Care must be taken to collect and to the setting) in complementary diets, and delivered
analyze all blood samples collected in efficacy trials through the local health services within a country, can
according to IZiNCG recommended techniques and increase intakes of absorbable zinc and have a positive
to include measurements of all potential confounders impact on serum zinc concentrations or stunting and
(including α1-glycoprotein and C-reactive protein as other potentially zinc-related functional outcomes (e.g.,
indicators of infection). morbidity) in young children in countries where access
Rigorously designed efficacy trials should be con- to animal-source foods is not a limiting factor.
ducted in target groups known to be at high risk for
References
1. Ruel MT. Can food-based strategies help reduce vitamin 10. Solon F, Fernandez TL, Latham MC, Popkin BM. An
A and iron deficiencies? A review of recent evidence. evaluation of strategies to control vitamin A deficiencies
Washington, DC: International Food Policy Research in the Philippines. Am J Clin Nutr 1979; 32:1445–53.
Institute, 2001. 11. Popkin BM, Solon FS, Fernandez T, Latham MC.
2. Bierti PR, Krasevec J, Fitzgerald S. A review of the effec- Benefit-cost analysis in the nutrition area: A project in
tiveness of agriculture interventions in improving nutri- the Philippines. Soc Sci Med 1980;14C:207–16.
tion outcomes. Public Health Nutr 2004;7:599–609. 12. Galal OM, Harrison GG, Abdou AI, Zein el Abedin A.
3. Bierti PR, Krasevec J, Fitzgerald S. Effectiveness of The impact of a small-scale agricultural intervention on
small-scale rural agriculture interventions. Part 1. Nutri- socio-economic and health status. Food Nutr (Roma)
tion outcomes—literature review and critical analysis. 1987;13:35–43.
Ottawa: Programme for Appropriate Technology in 13. CARE Nepal. A study of the evaluation of home garden-
Health, 2002. ing program in Bajura and Mahottari districts. Project
4. Marsh R. Building on traditional gardening to improve Report. Kathmandu, Nepal: CARE, 1995.
household food security. Food Nutr Agric 1998; 14. Hagenimana V, Oyunga MA, Low J, Njoge SM, Gichuki
22:4–14. ST, Kabira J. The effects of women farmer’s adoption of
5. Kidala D, Greiner T, Gebre-Medhin M. Five-year follow- orange-fleshed sweet potatoes: Raising vitamin A intake
up of a food-based vitamin A intervention in Tanzania. in Kenya. Research Reports Series No. 3. Washington,
Public Health Nutr 2000;3:425–431. DC: International Center for Research on Women,
6. Chakravarty I. Food based strategies to control vitamin 1999.
A deficiency. Food Nutr Bull 2000;21:135–143. 15. Berzin NI, Bauman VK. Vitamin-A-dependent zinc-
7. Ngu T, Quang ND, Ha PH, Giay T, Badcock JC, FitzGer- binding protein and intestinal absorption of zinc in
ald S. A food based approach to nutrition improvement chicks. Br J Nutr 1987;57:255–68.
through household food security in Vietnam, with spe- 16. Skalan D, Halevy O, Donoghue S. The effect of dif-
cial reference to vitamin A deficiency. In: Proceedings ferent dietary levels of vitamin A on metabolism of
of 6th IVACG Meeting, Chiang Rai, Thailand, 1994. copper, iron, and zinc in the chick. Int J Vitam Nutr Res
Washington, DC: International Life Sciences Institute, 1987;57:11–8.
1995:7. 17. Holland B, Welch AA, Unwin ID, Buss DH, Paul AA,
8. Solon F, Briones H, Fernandez JR, Shafritz LB. Moving Southgate DAT. McCance and Widdowson’s the compo-
to a long-term strategy: Increasing vegetable garden- sition of foods. 5th revised and extended edition. Cam-
ing and consumption in the Philippines. In: Seidal RE, bridge, UK: Royal Society of Chemistry and Ministry of
ed. Strategies for promoting vitamin A production, Agriculture, Fisheries and Food, 1991.
consumption and supplementation. Four case studies. 18. English RM, Badcock JC, Giay T, Ngu T, Waters AM,
Washington, DC: Academy for Educational Develop- Bennett SA. Effect of nutrition improvement project on
ment, 1996. morbidity from infectious diseases in preschool children
9. Phillips M, Sanghvi T, Suarez R, McKigney J, Fiedler J. in Vietnam: Comparison with control commune. Br
The costs and effectiveness of three vitamin A interven- Med J 1997;315:1122–5.
tions in Guatemala. Soc Sci Med 1996;42:1661–8. 19. English R, Badcock J. A community nutrition project in
Viet Nam: Effects on child morbidity. Food Nutr Agric Mississauga, Ontario, Canada: World Vision Canada,
1998;22:15–21. 2005.
20. International Food Policy Research Institute (IFPRI)/ 32. Carrasco Sanez NC, de Ubillas RMD, Guillen IS, Fer-
Bangladesh Institute of Development Studies, Institute reira SM. Increasing women’s involvement in commu-
of Nutrition and Food Science, Data Analysis and Tech- nity decision-making: A means to improve iron status.
nical Assistance and Research Department of Human Research Report No. 1. Washington, DC: International
Nutrition, Royal Veterinary and Agricultural University. Center for Research on Women/Opportunities for
Commercial vegetable and poly-culture fish production Micronutrient interventions, 1998.
in Bangladesh: Their impacts on income, household 33. Creed-Kanashiro HM, Uribe TG, Bartolini RM,
resource allocation, and nutrition. Vol 1. Washington, Fukumoto MN, Lopez TT, Zavaleta NM, Bentley ME.
DC: IFPRI, 1998. Improving dietary intake to prevent anemia in adoles-
21. Smitasiri S, Dhanamitta S. Sustaining behaviour change cent girls through community kitchens in a periurban
to enhance micronutrient status: Community and population of Lima, Peru. J Nutr 2000;130:459S–61S.
women-based interventions in Thailand. Research 34. Guldan GS, Fan H, Ma X, Ni Z, Xiang X, Tang M. Cul-
Report no. 2. Washington, DC: International Center for turally appropriate nutrition education improves infant
Research on Women/Opportunities for Micronutrient feeding and growth in rural Sichuan China. J Nutr
interventions, 1999. 2000;130:1204–11.
22. Smitasiri S, Sangobwarchar K, Kongpunya P, Subsuwan 35. Penny ME, Creed-Kanashiro HM, Robert RC, Narrow
C, Banjong O, Chitchumroonechokchai C, Rusami- MR, Caulfield LE, Black RE. Effectiveness of an educa-
Sopaporn W, Veeravong S, Dhanamitta S. Sustaining tional intervention delivered through health services
behavioural change to enhance micronutrient status to improve nutrition in young children: A cluster-
through community and women-based interventions randomised controlled trial. Lancet 2005;365:1863–72.
in northeast Thailand: Vitamin A. Food Nutr Bull 1999; 36. Heath ALM, Skeaff M, O’Brien S, Williams SM, Wil-
20:243–51. liams BS, Gibson RS. Can dietary treatment of non-
23. Schipani S, van der Haar F, Sinawat S, Maleevong K. anemic iron deficiency improve iron status? J Am Coll
Dietary intake and nutritional status of young children Nutr 2001;20:477–84.
in families practicing mixed home gardening in north- 37. Lartey A, Mau A, Brown KH, Peerson KM, Dewey KG.
east Thailand. Food Nutr Bull 2002;23:175–80. A randomized, community-based trial of the effects of
24. Stallkamp G, Karim R, Jinnatunnessa, Habib A, Baten improved, centrally processed complementary foods on
A, Uddin A, Talukder A, Panagides D, de Pee S. growth and micronutrient status of Ghanaian infants from
Homestead food production reduces the prevalence of 6 to 12 mo of age. Am J Clin Nutr 1999;70:391–404.
anemia among non-pregnant women. In: Kalambaheti 38. Krebs NF, Westcott JE, Butler N, Robinson C, Bell M,
Y, Charoenkiatkl S, eds. Optimizing dietary strategies Hambidge KM. Meat as a first complementary food for
for better health in developing countries. Bioavailability breastfed infants: Feasibility and impact on zinc intake and
2006. Abstract O3.3. Chiang Mai, Thailand: Institute of status. J Pediatr Gastroenterol Nutr 2006;42:207–14.
Nutrition, Mahidol University, 2006:67. 39. Engelmann MDM, Sandström B, Michaelsen KF. Meat
25. Ying C, Fengying Z, Wenjun L, Keyou G, Daxun J, de intake and iron status in late infancy: An intervention
Onis M. Nutritional status of preschool children in study. J Pediatr Gastroenterol Nutr 1998;26:26–33.
poor rural areas of China. Bull World Health Organ 40. Grillenberger M, Neumann CG, Murphy SP, Bwibo NO,
1994;72:105–12. van’t Veer P, Hautvast JG, West CE. Food supplements
26. Sheikholeslam R, Kimiagar M, Siasi F, Abdollahi Z, have a positive impact on weight gain and the addition
Jazayeri A, Keyghobadi K, Ghaffarpoor M, Noroozi of animal source foods increases lean body mass of
F, Kalantari M, Minaei N, Eslami F, Hormozdyari H. Kenyan school children. J Nutr 2003;133:3957S–64S.
Multidisciplinary intervention for reducing malnutri- 41. Grillenberger M, Neumann CG, Murphy SP, Bwibo
tion among children in the Islamic Republic of Iran. NO, Weiss RE, Jiang L, Hauvast JG, West CE. Intake of
East Mediterr Health J 2004;10:844–52. micronutrients high in animal-source foods is associated
27. Ayele Z, Peacock C. Improving access to and con- with better growth in rural Kenyan school children. Br
sumption of animal source foods in rural households: J Nutr 2006;95:379–90.
The experience of a women-focused goat develop- 42. Murphy SP, Gewa C, Liang LL, Grillenberger M, Bwibo
ment program in the highlands of Ethiopia. J Nutr NO, Neumann C. School snacks containing animal
2003;133:3981S–6S. source foods improve dietary quality for children in
28. Nielsen H, Roos N, Thilsted SH. The impact of semi- rural Kenya. J Nutr 2003;133:3950S–7S.
scavenging poultry production on the consumption of 43. Neumann CG, Bwibo NO, Murphy SP, Sigman M,
animal source foods by women and girls in Bangladesh. Whaley S, Allen LH, Guthrie D, Weiss RE, Demment
J Nutr 2003;133:4027S–30S. MW. Animal source foods improve dietary quality,
29. Roos N, Islam MM, Thilsted SH. Small fish is an impor- micronutrient status, growth and cognitive function in
tant dietary source of vitamin A and calcium in rural Kenyan school children: Background, study design and
Bangladesh. Int J Food Sci Nutr 2003;54:329–39. baseline findings. J Nutr 2003;133:3941S–9S.
30. Roos N, Islam MM, Thilsted SH. Indigenous fish spe- 44. Siekmann JH, Allen LH, Bwido NO, Demment W,
cies in Bangladesh: Contribution to vitamin A and iron Murphy SP, Neumann CG. Kenyan school children
intakes. J Nutr 2003;133:4021S–6S. have multiple micronutrient deficiencies, but increased
31. Radford KB. World Vision Malawi’s Micronutrient and plasma vitamin B-12 is the only detectable micronutri-
Health (MICAH) Progam. Small animal revolving fund. ent response to meat or milk supplementation. J Nutr
2003;133:3972S–80S. absorption from low-phytate hybrid of maize and their

45. Whaley SE, Sigman M, Neumann C, Bwibo N, Guthrie wild-type isohybrids. Am J Clin Nutr 2004;79:1053–9.
D, Weiss RE, Alber S, Murphy SP. The impact of dietary 62. Mazariegos M, Hambdige KM, Krebs NF, Westcott JE,
intervention on the cognitive development of Kenyan Lei S, Grunwald GW, Campos R, Barahona B, Raboy V,
school children. J Nutr 2003;133:3965S–71S. Solomons NW. Zinc absorption in Guatemalan school-
46. World Health Organization/UNICEF. Complementary children fed normal or low-phytate maize. Am J Clin
feeding of young children in developing countries: A review Nutr 2006;83:59–64.
of current scientific knowledge. Geneva: WHO, 1998. 63. Mendoza C, Viteri FE, Lönnerdal B, Young KA, Raboy
47. Gibson RS, Hotz C. Dietary diversification/modifica- V, Brown KH. Effect of genetically modified, low-phytic
tion strategies to enhance micronutrient content and acid maize on absorption of iron from tortillas. Am J
bioavailability of diets in developing countries. Br J Nutr Clin Nutr 1998;68:1123–7. Erratum in: Am J Clin Nutr
2001;85:S159–66. 69:743.
48. Lönnerdal B, Cederblad Å, Davidsson L, Sandström B. 64. Hurrell RF, Reddy MB, Juillerat M-A, Cook JD. Deg-
The effect of individual components of soy formula and radation of phytic acid in cereal porridges improves
cows’ milk formula on zinc bioavailability. Am J Clin iron absorption by human subjects. Am J Clin Nutr
Nutr 1984;40:1064–70. 2003;77:1213–9.
49. Turnlund JR, King JC, Keyes WR, Gong B, Michel MC. 65. Hambidge KM, Krebs NF, Westcott JL, Sian L, Miller
A stable isotope study of zinc absorption in young LV, Peterson KL, Raboy V. Absorption of calcium from
men: Effects of phytate and α-cellulose. Am J Clin Nutr tortilla meals prepared from low-phytate maize. Am J
1984;40:1071–7. Clin Nutr 2005;82:84–7.
50. Nävert B, Sandström B. Reduction of the phytate con- 66. Lachat CK, Van Camp JH, Mamiro PS, Wayua Obuoro
tent of bran by leavening in bread and its effect on zinc F, Opsomer AS, Roberfroid DA, Kolsteren PW. Process-
absorption in man. Br J Nutr 1985;53:47–53. ing of complementary foods does not increase hair zinc
51. Lönnerdal B, Bell JG, Hendrickx AG, Burns RA, Keen levels and growth of infants in Kilosa district, rural
CL. Effect of phytate removal on zinc absorption from Tanzania. Br J Nutr 2006;95:174–80.
soy formula. Am J Clin Nutr 1988;48:1301–6. 67. Mamiro PS, Kolsteren PW, van Camp JH, Roberfroid
52. Kivistö B, Cederblad Å, Davidsson L, Sandberg A-S, DA, Tatala S, Opsomer AS. Processed complementary
Sandström B. Effect of meal composition and phytate food does not improve growth or hemoglobin status
content on zinc absorption in humans from an extruded of rural Tanzanian infants from 6–12 months of age in
bran product. J Cereal Sci 1989;10:189–97. Kilosa District, Tanzania. J Nutr 2004;134:1084–90.
53. Etcheverry P, Hawthorne KM, Liang LK, Abrams SA, 68. Lind T, Lönnerdal B, Persson L-Å, Stenlund H, Tenne-
Griffin IJ. Effect of beef and soy proteins on the absorp- fors C, Hernell O. Effects of weaning cereals with dif-
tion of non-heme iron and inorganic zinc in children. J ferent phytate content on hemoglobin, iron stores, and
Am Coll Nutr 2006;25:34–40. serum zinc: A randomized intervention in infants from
54. Davidsson L, Ziegler EE, Kastenmayer P, van Dael P, 6 to 12 mo of age. Am J Clin Nutr 2003;78:168–75.
Barclay D. Dephytinisation of soybean protein isolate 69. Lind T, Persson L-Å, Lönnerdal B, Stenlund H,
with low native phytic acid content has limited impact Hernell O. Effects of weaning cereals with different
on mineral and trace element absorption in healthy phytate content on growth, development and morbidity:
infants. Br J Nutr 2004;91:287–93. Randomized intervention trial in infants from 6 to 12
55. Manary MJ, Hotz C, Krebs NF, Gibson RS, Westcott months of age. Acta Paediatr 2004;93:1575–82.
JE, Arnold T, Broadhead RL, Hambidge KM. Dietary 70. Gibson RS, Yeudall F, Drost N, Mtitimuni BM, Cullinan
phytate reduction improves zinc absorption in Malawian TR. Experiences of a community-based dietary inter-
children recovering from tuberculosis but not in well vention to enhance micronutrient adequacy of diets low
children. J Nutr 2000;130:2959–64. in animal source foods and high in phytate: A case study
56. Egli I, Davidsson L, Zeder C, Walczyk T, Hurrell R. in rural Malawian children. J Nutr 2003;133:3992S–9S.
Dephytinization of a complementary food based on 71. Yeudall F, Gibson RS, Cullinan TR, Mtimuni B. Efficacy
wheat and soy increases zinc, but not copper, apparent of a community-based dietary intervention to enhance
absorption in adults. J Nutr 2004;134:1077–80. micronutrient adequacy of high-phytate maize-based
57. Hambidge KM, Miller LV, Tran CD, Krebs NF. Measure- diets of rural Malawian children. Public Health Nutr
ments of zinc absorption: Application and interpretation 2005;8:826–36.
in research designed to improve human zinc nutriture. 72. Yeudall F, Kayira C, Umar E, Gibson RS. Impact of a
Int J Vitam Nutr Res 2005;75:385–93. community-based dietary intervention on selected
58. Liuzzi JP, Cousins RJ. Mammalian zinc transporter. biochemical and functional outcomes in rural Malawian
Annu Rev Nutr 2004;24:151–72. children. Eur J Clin Nutr 2002;56:1176–85.
59. Hotz C, Gibson RS. Assessment of home-based process- 73. Hotz C, Gibson RS. A participatory nutrition education
ing methods to reduce phytate content and phytate/zinc intervention improves the adequacy of complementary
molar ratios of white maize (Zea mays). J Agric Food diets of rural Malawian children: A pilot study. Eur J
Chem 2001;49:692–8. Clin Nutr 2005;59:226–37.
60. Adams CL, Hambidge M, Raboy V, Dorsch JA, Sian 74. Porres JM, Etcheverry P, Miller DD. Phytate and citric
L, Westcott JL, Krebs NF. Zinc absorption from a low acid supplementation in whole-wheat bread improves
phytic-acid maize. Am J Clin Nutr 2002;76:556–9. phytate-phosphorus release and iron dialyzability. J
61. Hambidge KM, Huffer JW, Raboy V, Grunwald GK, Food Sci 2001;66:614–9.
Westcott JL, Sian L, Miller LV, Dorsch JA, Krebs NF. Zinc 75. Liljeberg H, Björck I. Delayed gastric emptying rate
may explain improved glycaemia in healthy subjects Imong S, Jackson D, Woolridge M. Nursing frequency
to a starchy meal with added vinegar. Eur J Clin Nutr and the energy intake from breast milk and supplemen-
1998;52:368–71. tary food in a rural Thai population: A longitudinal
76. Rimbach WA, Most E, Pallauf J. Effect of citric acid sup- study. Eur J Clin Nutr 1993;47:880–91.
plements to a maize-soya diet on the in vitro availability 92. Kimmons JE, Dewey KG, Haque E, Chakraborty J,
of minerals, trace elements, and heavy metals. Zentralbl Osendarp SJ, Brown KH. Low nutrient intakes among
Veterinarmed A 1998;45:517–24. infants in rural Bangladesh are attributable to low intake
77. Hunt JR, Gallagher SK, Johnson LK, Lykken GI. High- and low micronutrient density of complementary foods.
versus low-meat diets: Effects on zinc absorption, iron J Nutr 2005;135:444–51.
status, and calcium, copper, iron, magnesium, manga- 93. Paul AA, Bates CJ, Prentice A, Day KC, Tsuchiya H. Zinc
nese, nitrogen, phosphorus, and zinc balance in post- and phytate intake of rural Gambian infants: Contribu-
menopausal women. Am J Clin Nutr 1995;62:621–32. tions from breastmilk and weaning foods. Int Food Sci
78. Jalla S, Westcott J, Steirn M, Miller LV, Bell M, Krebs Nutr 1998;49:41–154.
NF. Zinc absorption and exchangeable zinc pool size in 94. Talukder A, Stallkamp G, Sapkota G, Kroeun H, Karim
breast-fed infants fed meat or cereal as first complemen- R, Witten C, Haselow N, de Pee S, Moench-Pfanner R,
tary food. J Pediatr Gastroenterol Nutr 2002;34:35–41. Bloem MW. Homestead food production and its impact
79. Rosado JL, Diaz M, Gonzalez K, Griffin I, Abrams SA, on the prevalence of anemia among non-pregnant
Preciado R. The addition of milk or yoghurt to a plant- women and children in Asia (Bangladesh, Nepal and
based diet increases zinc bioavailability but does not affect Cambodia). In: Consequences and control of micro-
iron bioavailability in women. J Nutr 2005;135:465–8. nutrient deficiencies: science, policy, and programs—
80. Sandström B. Dose dependence of zinc and manganese defining the issues. Micronutrient Forum. Istanbul,
absorption in man. Proc Nutr Soc 1992;51:211–8. Turkey, 2007. Abstract W18, p 102. http://www.micro-
81. Habicht JP, Victora CG, Vaughan JP. Evaluation designs nutrientforum.org/Meeting2007/posterPresentations.
for adequacy, plausibility and probability of public cfm.
health programme performance and impact. Int J 95. Stallkamp G, Akhter N, Karim R, Habib A, Baten A,
Epidemiol1999;28:10–8. Uddin A, Talukder A, de Pee S. Homestead food pro-
82. Creed-Kanashiro HM, Bartolini RM, Fukumoto MN, duction improves micronutrient status in mothers and
Uribe TG, Robert RC, Bentley ME. Formative research young children in Bangladesh. In: Consequences and
to develop a nutrition education intervention to improve control of micronutrient deficiencies: science, policy,
dietary iron intake among women and adolescent girls and programs—defining the issues. Micronutrient
through community kitchens in Lima, Peru. J Nutr 2003; Forum. Istanbul, Turkey, 2007. Abstract W15, p 102.
133:3987S–91S. http://www.micronutrientforum.org/Meeting2007/
83. Bjarnson I, Macpherson A, Hollander D. Intestinal posterPresentations.cfm.
permeability: An overview. Gastroenterology 1995; 96. Helen Keller International. Homestead food produc-
108:1566–81. tion—an effective integrated approach to improve
84. Thurnham DI, Mburu ASW, Mwaniki DL, De Wagt A. food security among the vulnerable char dwellers in
Micronutrients in childhood and the influence of sub- northern Bangladesh. Homestead Food Prod Prog Bull
clinical inflammation. Proc Nutr Soc 2005;64:502–9. 2006;4:1–4.
85. Quist CF, Howerth EW, Fischer JR, Wyatt RD, Miller DM, 97. Food and Agriculture Organization/International Life
Nettles VF. Evaluation of low-level aflatoxin in the diet Sciences Institute. Preventing micronutrient malnutri-
of white-tailed deer. J Wildlife Dis 1997;33:112–21. tion: A guide to food-based approaches. Washington,
86. Hell K, Cradwell KF, Setamou M, Poehling HM. The DC: International Life Sciences Institute, 1997.
influence of storage practices on aflatoxin contamina- 98. Kuhnlein HV, Pelto GH, eds. Culture, environment, and
tion in maize in four agroecological zones of Benin, food to prevent vitamin A deficiency. Boston: Interna-
West Africa. J Stored Prod Res 2000;36:365–82. tional Nutrition Foundation for Developing Countries,
87. Moss MO. Recent studies of mycotoxins. J Appl Micro- 1997.
biol Suppl 1998;84:62S–76S. 99. Gibson RS, Hotz C, Temple L, Yeudall F, Mtitimuni B,
88. Temple L. Household processing methods for improving Ferguson EL. Dietary strategies to combat deficiencies
the nutrient adequacy of complementary foods in rural of iron, zinc and vitamin A in developing countries:
Malawi: Their impact on content and relative bioavail- Development, implementation, monitoring and evalu-
ability of calcium, iron, and zinc. MSc thesis, University ation. Food Nutr Bull 2000;21:219–31.
of Otago, New Zealand, 2001. 100. Faber M, Laurie S, Venter S. Home-gardens to address
89. Temple L, Gibson RS, Hotz C. Use of soaking and vitamin A deficiency in South Africa: A food-based
enrichment for improving the content and bioavail- approach. Pretoria, South Africa: ARC-Roodeplaat
ability of calcium, iron, and zinc in complementary Vegetable and Ornamental Plant Institute, 2007.
foods and diets of rural Malawian weanlings. J Food Sci 101. International Zinc Nutrition Consultative Group.
2002;67:1926–32. Preventing zinc deficiency through diet diversification
90. Mensah P, Tomkins A. Household-level technolo- and modification. IZiNCG Technical Brief No. 5. Davis,
gies to improve the availability and preparation of University of California, 2007.
adequate and safe complementary foods. Food Nutr Bull 102. Dickin K, Griffiths M, Piwoz E. Designing by dialogue.
2003;24:104–25. A program planner’s guide to consultative research
91. Drewett R, Amatayakul K, Wongsawasdii L, Mangkla- for improving young child feeding. Washington, DC:
bruks A, Ruckpaopunt S, Ruangyuttikarn C, Baum D, Support for Analysis and Research in Africa/Academy
for Educational Development/Health and Human 107. de Benoist B, Darnton-Hill I, Davidsson L, Fontaine O,
Resources Analysis for Africa/Bureau for Africa, Office eds. Report of a WHO/UNICEF/IAEA/IZiNCG Intera-
of Sustainable Development, 1997. gency meeting on zinc status indicators. Food Nutr Bull
103. Pretty JN, Guijt I, Thompson J, Scoones I. Training in 2007;28:S399–S486.
participant methods in the workshop. In: Participatory 108. International Zinc Nutrition Consultative Group
learning and action: a trainer’s guide. London: Sustain- (IZiNCG), Brown KH, Rivera JA, Bhutta Z, Gibson
able Agriculture Programme, International Institute for RS, King JC, Lönnerdal B, Ruel MT, Sandtröm B,
Enviroment and Development, 1995:72–89. Wasantwisut E, Hotz C. Assessment of the risk of zinc
104. Pelletier DL, Kraak V, Ferri-Morris M. The CBNM deficiency in populations and options for its control.
problem solving model: An approach for improving Food Nutr Bull 2004;25(1 suppl 2):S99–203.
nutrition research decision-making in the community. 109. International Zinc Nutrition Consultative Group. Quan-
A manual prepared for the Community-Based Nutrition tifying the risk of zinc deficiency: Recommended indica-
Monitoring (CBNM) strategy workshop. Ithaca, NY, tors. IZiNCG Technical Brief No. 01. Davis, University
USA: Division of Nutrition Science, Cornell University, of California, 2007.
1994. 110. International Zinc Nutrition Consultative Group.
105. Smith WE, Landais FL. A new approach to organizing Assessing population zinc status with serum zinc
and planning. Dev Southern Afr 1991;8:223–38. concentration. IZiNCG Technical Brief No. 02. Davis,
106. Levinson FJ, Rogers BL, Hicks KM, Schaetzel T, Troy University of California, 2007.
L, Young C. Monitoring and evaluation. A guidebook 111. International Zinc Nutrition Consultative Group.
for nutrition project managers in developing countries. Determining the risk of zinc deficiency: Assessment
Washington, DC: Human Development Network/World of dietary zinc intake. IZiNCG Technical Brief No. 03.
Bank, 1999. Davis, University of California, 2007.
Dietary intervention strategies to enhance zinc
nutrition: Promotion and support of breastfeeding
for infants and young children
Kenneth H. Brown, Reina Engle-Stone, Nancy F. Krebs, and Janet M. Peerson
Abstract Background
Breastmilk is the only dietary source of zinc for exclu- Promotion and support of appropriate breastfeeding
sively breastfed young infants, and it remains a poten- practices should be considered among the recom-
tially important source of zinc for older infants and mended dietary strategies to enhance the zinc status
young children who continue breastfeeding beyond early of infants and young children, for two main reasons:
infancy. Therefore, we examined available informa- breastmilk is an important potential source of bioavail-
tion on breastmilk zinc concentration and total milk able zinc, and breastfeeding protects against diarrhea
consumption to develop estimates of the amount of zinc [1–3], which causes excessive zinc losses [4]. This
transferred in breastmilk to children of different ages. paper will provide a summary of available information
Breastmilk zinc concentration declines rapidly during on the amount of zinc that can be provided to infants
the first few months postpartum and more slowly there- and young children through breastfeeding. The paper
after. Breastmilk supplies all of the theoretical zinc needs is divided into three sections, which address the fol-
for at least the first several months of life, although the lowing questions with regard to breastfeeding as a
period during which breastmilk alone remains sufficient component of dietary intervention strategies to prevent
is uncertain. Breastmilk continues to provide more than zinc deficiency:
half of children’s estimated zinc requirements after the
Section 1: How much zinc is transferred in breastmilk
introduction of complementary foods, even into the
to exclusively breastfed infants less than 6 months of
second year of life. Public health programs to promote
age in relation to their physiological requirements for
and support breastfeeding should be included among
absorbed zinc?
the strategies to ensure adequate zinc status of young
children. Section 2: How much zinc is transferred in breastmilk
to infants and young children less than 24 months
of age who are consuming breastmilk in addition to
Key words: Breastfeeding, breastmilk, infant, zinc, complementary foods?
zinc absorption, zinc requirements
Section 3: What are the programmatic implications of
the answers to the foregoing questions, and what are
the remaining research needs?
Section 1
Kenneth H. Brown, Reina Engle-Stone, and Janet M.
Peerson are affiliated with the Department of Nutrition and How much zinc is transferred in breastmilk to exclusively
the Program in International and Community Nutrition,
University of California, Davis, California, USA; Kenneth breastfed infants less than 6 months of age in relation to
H. Brown is also affiliated with Helen Keller International, their physiological requirements for absorbed zinc?
Dakar, Senegal. Nancy F. Krebs is affiliated with the Depart-
ment of Pediatrics, University of Colorado School of Medi-
cine, Denver, Colorado, USA. Conclusions
H. Brown, Department of Nutrition, University of California, The mean amount of zinc transferred in breastmilk
One Shields Ave., Davis, CA 95616, USA; e-mail: khbrown@ to exclusively breastfed infants declines rapidly from
ucdavis.edu. ~4 mg/day during the first few days of life to ~1.75
Zinc intake through breastmilk S145
mg/day by 1 month. Zinc transfer declines more slowly Studies of milk zinc concentration in relation to
thereafter to ~0.7 mg/day at 6 months. The adequacy of time of day have yielded inconsistent results. Four
breastmilk as the sole source of zinc during this period studies found no relation between the time of day and
is difficult to determine because of uncertainty in the milk zinc concentration [5–8], but three studies found
fractional absorption of zinc (FAZ) from breastmilk that specimens obtained early in the morning had
and doubts regarding the true physiological require- ~15% greater milk zinc concentration than specimens
ments for absorbed zinc during this period of life. obtained in the afternoon or evening [9–11]. Thus, it
Nevertheless, using different assumptions for likely might be advisable to obtain more than one sample
FAZ, based on available studies that suggest a mean during the day to gather more representative infor-
FAZ of ~0.50 and current estimates of zinc require- mation on the total amount of zinc transferred, even
ments, breastmilk alone is probably adequate for though many studies suggest that just one sample might
exclusively breastfed term infants for at least 3 months be sufficient.
and possibly for as long as 6 months, depending on Two studies examined milk zinc concentrations in
which sets of estimates are applied and whether zinc different portions of the feeding, and neither found
accrued by the fetus during gestation is available for any significant differences in the zinc concentrations
infant metabolism. of foremilk and hindmilk [5, 12]. Only one study
There is little information from zinc supplementa- compared zinc concentrations in milk obtained from
tion trials conducted among exclusively breastfed term the left versus the right breast, and no consistent dif-
infants less than 6 months of age, and the results of ferences were detected [12].
supplementation trials carried out among a broader age Ten studies are available that compared the zinc
range of breastfed infants are inconsistent. In general, concentration in the milk of mothers who delivered
three such trials carried out in industrialized countries at term with that of mothers who delivered prema-
failed to demonstrate a consistent functional advan- turely [13–22]. There were no consistent differences
tage of providing additional zinc to infants less than 6 in the zinc concentrations of milk obtained from the
months of age, so the amount transferred in breastmilk respective sets of mothers. Other studies of the rela-
is presumably adequate during this period. Breastmilk tions between maternal age and parity and milk zinc
is clearly an important source of highly bioavailable concentrations are difficult to interpret because of the
zinc during this period of life and may be adequate as failure to control adequately for infant age. Among the
the sole source of zinc for exclusively breastfed term subset of studies that did compare results at specific,
infants until ~6 months. narrowly defined time periods postpartum, there are
no consistent effects of maternal age or parity on milk
Detailed review of evidence zinc concentrations [7, 23–26].
As reviewed by Hess and King in this issue [27],
Overview of methodologic issues maternal zinc status and maternal zinc supplementa-
Milk sampling. Accurate measurement of the amount tion do not appear to affect milk zinc concentration.
of zinc transferred through breastmilk from moth- On the other hand, several longitudinal studies have
ers to their nursing children requires first, collection found greater variability in milk zinc concentration
of representative samples of milk for analysis of zinc between women than within women [10, 28], indicat-
concentration and second, assessment of the total ing that individual women have characteristic levels of
amount of milk consumed. The total amount of zinc zinc in their milk, possibly because of genetic factors
transferred through breastmilk can then be determined influencing zinc transport to or within the mammary
by multiplying the milk zinc concentration by the total gland. The public health importance of these interin-
amount of milk consumed. To develop an appropri- dividual differences in breastmilk zinc concentration
ate milk sampling protocol for assessing milk zinc is unknown, because no relevant studies are available
concentration, several investigators have examined relating milk zinc concentration to infant growth.
whether milk zinc concentration varies by the time of In summary, there are few extraneous factors, other
day, portion of the feeding, left versus right breast, the than time postpartum, that consistently affect breast-
gestational age at birth and postnatal age of the child, milk zinc concentration. Thus, it is possible to collect
and the age, parity, and zinc status of the mother. The milk samples for analysis of zinc concentration from
results of these studies are summarized in the following either breast, from any portion of the feeding, and at
paragraphs. The single factor that appears to have the just one time or a few times per day, regardless of the
greatest impact on milk zinc concentration is the time age, parity, and nutritional status of the mother.
postpartum. Thus, assessments of other factors affect-
ing milk zinc concentration are interpretable only if the Pattern of distribution of data regarding milk zinc
age of the child is held constant. The specific relation concentration and milk volume. To calculate the range
between time postpartum and milk zinc concentration of total zinc consumption from breastmilk by chil-
is discussed in more detail in the next section. dren of different ages, information is needed on the
distribution of data concerning milk zinc concentra- were identified through a computerized database
tion and total milk intakes during each age interval search (PubMed, accessed September 12, 2007) using
of interest. Few studies have reported information on the keywords “breast,” “milk,” “zinc,” and “infant” and
the distribution of milk zinc concentration at different no additional limits, which yielded 320 bibliographic
ages, so we reanalyzed data from a previously published citations. The titles were screened for relevance, and the
study that collected milk samples from 71 infants at 2 full texts of the articles were obtained when possible.
weeks of age and again monthly from 1 through 7 9 Additional data sets were located from the lists of cita-
months of age [28]. As shown in figure 1, these data tions included in the articles found during this initial
indicate that the distribution of milk zinc concentra- search, through papers cited in a previously published
tion is skewed somewhat to higher values at each age, review [35], and by contacting other experts in the
although the data can be normalized by logarithmic field. Using this search strategy, we identified a total of
transformations. 69 publications from 64 different studies that reported
We also examined the results of seven previously data on breastmilk zinc concentration.
completed studies of breastmilk intake to assess the From the set of 69 publications that were retrieved,
age-specific distribution of milk volumes [28–34] and we selected only those studies that presented data
concluded that milk intake was generally normally from mothers of healthy, term infants and provided
distributed in all studies (Shapiro-Wilk statistic > 0.98 information on both milk zinc concentration and time
in all cases). postpartum. Despite the conclusions of the methodo-
logic review presented above, for the sake of caution
Studies of breastmilk zinc transfer to exclusively breastfed we excluded studies that enrolled preterm infants
infants under 6 months of age or children with specific illnesses, unless data were
Identification and selection of studies. Data sets con- reported separately for unaffected children. We also
taining information on breastmilk zinc concentration excluded studies of mothers who were receiving zinc
A. B.
Breastmilk zinc concentration (mg/L)
8 8
7 7
6 6
5 5
4 4
3 3
2 2
1 1
0 0
0 2 4 6 8 10 –1.0 –0.5 0.0 0.5 1.0 1.5 2.0 2.5
Age (mo) ln(age) (mo)
C. D.
ln (breastmilk zinc concentration) (mg/L)
ln (breastmilk zinc concentration) (mg/L)
2.0 2.0
1.5 1.5
1.0 1.0
0.5 0.5
0.0 0.0
–0.5 –0.5
–1.0 –1.0
–1.5 –1.5
–2.0 –2.0
0 2 4 6 8 10 –1.0 –0.5 0.0 0.5 1.0 1.5 2.0 2.5
Age (mo) ln(age) (mo)
FIG. 1. Distribution of breastmilk zinc concentration according to time postpartum, and effects of different mathematical
transformations of data. Data provided by Krebs et al. [28]
supplements, unless the authors specifically stated that

12
there was no effect of supplementation on milk zinc
concentration or data from nonsupplemented women 10
were presented separately. We also excluded studies of
women who were receiving other mineral-containing 8
supplements or other medicines. These exclusions
6
resulted in a final set of 33 studies that were included
in the combined analysis, as shown in table 1 [5, 7, 4
13–18, 21, 23, 26, 28, 36–61]. The studies that were
excluded from the analysis are shown in table 2 [6, 2
8–11, 19, 22, 24, 25, 62–86], along with the reason(s)
0
for the exclusion. 0 5 10 15 20
Milk zinc concentration according to infant age. To Age (mo)

express the data on milk zinc concentration from dif-
ferent studies in consistent units, we applied conversion FIG. 2. Breastmilk zinc concentration according to time
factors of 65.4 g of zinc per mole and 1.03 g of breast- postpartum, all studies. Ln(Zn concentration) = 0.975 −
milk per milliliter [35], and all results are presented 0.501*ln(age) − 0.063*ln(age)2
as milligrams of zinc per liter of milk. The data avail-
able from all 33 acceptable studies are summarized in
infant age and the means and standard deviations
figure 2 and table 3. Each point in the figure represents
of log(breastmilk zinc concentration) and of breast-
the mean milk zinc concentration reported for the mid-
milk intake were estimated with regression analysis.
point of a particular age interval in a given study. The
Finally, we generated 1,000,000 simulated data points,
data across studies are remarkably consistent, showing
assuming that age was uniformly distributed between
a relatively high concentration of zinc in colostrum
0 and 6 months for exclusively breastfed children
and early transitional milk, a rapid fall in milk zinc
and between 0 and 18 months for partially breastfed
concentration during the first 1 to 2 months postpar-
children, and that log(breastmilk zinc concentration)
tum, and a slower decline thereafter. Tests of various
+ log(breastmilk zinc concentration)2 and breastmilk
transformations of breastmilk zinc concentration and
volume were normally and independently distributed
age indicated that a log–log relationship including a
at each age. We then multiplied individual simulated
quadratic term for log(age) provided the best fit for the
values for breastmilk intake and milk zinc concentra-
relationship between milk zinc concentration and time
tion to estimate total zinc transfer.
postpartum (r2 = 0.88, p < .0001).
The results of the simulations of the total daily
Amount of milk consumed according to infant age. amount of zinc transferred in breastmilk according to
Information on the amount of milk consumed by child age are shown in figure 3 for exclusively breastfed
children of different ages was extracted from the pub- infants less than 6 months of age. The mean amount
lication “Complementary feeding of young children in of zinc transferred in breastmilk declines from ~4 mg/
developing countries” by the World Health Organiza- day during the first few days of life to ~1.75 mg/day
tion (WHO) and the United Nations Children’s Fund
(UNICEF) [87]. This document provided separate
information for exclusively breastfed infants less than
Zinc transferred in breastmilk (mg/day)
8
6 months of age and for non-exclusively breastfed chil-
dren less than 24 months of age, as shown in table 4. 7
6
Total zinc transfer to exclusively breastfed infants less than
6 months of age and relation to zinc requirements 5
To estimate the total amount of zinc transferred in 4

breastmilk, we completed the following simulation 3
exercise. Breastmilk zinc concentration was assumed 2
to follow a log‑normal distribution, based on the
1
information described above [28]. Because only sample
means and standard deviations were available from 0
0 1 2 3 4 5 6
the individual studies used to estimate milk zinc
Age (mo)
concentrations, the means and standard deviations
of log(breastmilk zinc concentration) were estimated FIG. 3. Simulated mean and 95% prediction interval of daily
from the untransformed summary data by using the zinc transfer in breastmilk to exclusively breastfed infants
method of moments [88]. The relationships between according to infant age
TABLE 1. Publications included in the data analysis S148
Zinc concentration in milk
Country, year Health and zinc status of according to stage of lactation Growth or zinc status
[reference] author Study group mothers (mg/L) Description of milk sample of infant
Argentina, 2001 n = 20 mothers with term Apparently healthy Mean ± SEM, term infants Full breast sample collected by Infant growth and zinc
[14] Ronayne de milk (mean gestational age only: manual or pump expression status not reported
Ferrer of infants, 39.5 wk; range, Days 2–5: 6.96 ± 0.69 between 10 and 12 am
38–41 wk; mean birthweight, Days 6–10: 4.27 ± 0.51
3,303 g) and n = 24 middle- Days 11–15: 3.36 ± 0.31
class mothers with preterm Days 16–30: 2.44 ± 0.11
milk (mean gestational age
of infants, 30.9 wk; range,
27–35 wk; mean birthweight,
1,440 g)
Brazil, 1989a n = 93 mothers of low SES Mothers received iron, vita- Mean ± SEM, exclusively 5–10 mL collected by manual Infant growth and zinc
[41] Donangelo with normal and uncom- min B12, and/or folate during breastfeeding mothers: expression between 9 and 10 status not reported
(excluded data plicated gestations, both pregnancy (but not lacta- Days 1–5: 5.94 ± 0.55 (n = 17) am before the infant was fed
from 31 to 280 partially breastfeeding and tion). 11% of mothers had Days 6–30: 2.84 ± 0.37 (n = 13)
days because of exclusively breastfeeding. serum zinc below cutoff for Days 31–280: 1.65 ± 0.38 (n
wide range of Mean maternal age, 25.9 low zinc for adult women = 10)
infant ages) ± 5.6 (SD) yr; mean birth- Mean ± SEM, partially breast-
weight, 3,277 ± 476 (SD) g feeding mothers:
Days 31–280: 1.26 ± 0.19 (n
= 21)
Brazil, 1998 n = 28 mothers with term milk All low SES; none received Mean ± SD (range), term 5–10 mL of milk collected by Infant weights were
[15] Trugo (mean birthweight, 3,380 ± supplements. No informa- infants only: manual expression between adequate for gesta-
53 [SD] g) and n = 7 moth- tion on maternal health and Days 1–5: 7.26 ± 3.03 (1.3– 9 and 10 am before the tional age. Infant zinc
ers with preterm milk (mean zinc status 13.7) (n = 22) infant was due to be fed status not reported
birthweight, 1,810 ± 47 [SD] Days 6–36: 2.79 ± 1.14 (1.2–
g) at 1–5 days postpartum; n 4.2) (n = 8)
= 9 mothers with term milk
(mean birthweight, 3,160 ±
29 [SD] g) and n = 15 moth-
ers with preterm milk (mean
birthweight, 1,700 ± 39 [SD]
g at 6–36 days postpartum).
All of low SES
Brazil, 2003 n = 50 mothers of term infants Maternal health and zinc Mean ± SD: Manual expression of 1.5 mL Infant growth and zinc
[21] da Costa (37–41 wk gestation) > 20 status not reported Days 1–7: 6.97 ± 2.82 (n = 48) of colostrum from both status not reported
yr of age, with at least 2 yr breasts between the 1st and
interparturition period; no 7th days of lactation
K. H. Brown et al.
supplements
Burundi, 1995 n = 5 exclusively breastfeeding Healthy; zinc status not meas- Mean ± SD: Breastmilk samples from 5 Infant growth and zinc
[53] Robberecht middle-class women (mean ured specifically Days 2–4 (colostrum): 3.76 ± lactating women [54] status not reported
age, 32 yr) with infants aged 0.51 (n = 1, 8 samples)
0–10 mo. Uncomplicated Days 5–14: 3.08 ± 0.62 (n = 1,
pregnancies 10 samples)
1 mo: 2.47 ± 1.17 (n = 3, 12
samples)
2 mo: 1.64 ± 0.06 (n = 2, 6
samples)
4 mo: 1.30 ± 0.06 (n = 2, 5
samples)
10 mo: 0.75 ± 0.02 (n = 2, 6
samples)
Canada, 1982 n = 10 mothers of term infants Normal pregnancy; maternal Mean ± SD, term infants only: Complete 24-h expression Infant growth and zinc
[18] Mendelson (mean gestational age, 39.0 health and zinc status not Days 3–5: 5.35 ± 1.2 (n = 8) with breast pump status not reported
wk; range, 38–40 wk; mean reported Days 8–10: 4.1 ± 0.65 (n = 8)
± SD birthweight, 1,143 ± Days 15–17: 3.37 ± 0.6 (n = 7)
305 g) and n = 14 mothers of Days 28–30: 2.6 ± 0.65 (n = 6)
preterm infants (mean ges-
tational age, 28.7 wk; range,
26–33 wk; mean ± SD birth-
weight, 3,361 ± 550 g)
All appropriate for gestational
age
Canada, 1999 n = 19 mothers 20–35 yr of Healthy, nonvegetarian; zinc Mean ± SD: Full breast expression by elec- Infant growth and zinc
[20] Friel age with full-term infants not specifically measured Days 2–3: 4.58 ± 0.52 (n = 17) tric pump during a regular status not measured
Day 8: 3.99 ± 0.88 (n = 16) feeding of the child between
Day 15: 3.54 ± 0.74 (n = 16) 10 am and 2 pm; a 15-mL
Day 22: 3.37 ± 2.06 (n = 15) aliquot was taken
Day 28: 2.5 ± 0.94 (n = 15)
Day 36: 2.32 ± 1.41 (n = 15)
Day 43: 1.89 ± 0.65 (n = 15)
Day 50: 1.56 ± 0.96 (n = 15)
Day 61: 1.14 ± 0.55 (n = 12)
China, 2007 n = 41 term, exclusively Mean maternal zinc intake, 7.5 Mean ± SD: 5-mL milk samples were Weights and lengths
[60] Xiang breastfeeding mother-infant mg/day; mothers apparently Group I (1 mo): 3.24 ± 0.23 expressed manually at 5 were in normal
pairs in rural China. Group healthy Group II (3 mo): 1.94 ± 0.17 time intervals (8–10 am, 10 ranges at birth, 1 mo,
I: n = 18 at 1 mo (mean am–12 pm, 12–2 pm, 2–4 and 3 mo. Infant zinc
maternal age, 26.6 yr). pm, and 4–6 pm) after the status not reported
Group II: n = 23 at 3 mo breast had been suckled by
(mean maternal age, 27.4 yr) the infant for 2 min
continued
S149
TABLE 1. Publications included in the data analysis (continued) S150
Finland, 1979 n = 27 mothers aged 20–35 yr Healthy, well-nourished, all Mean: Fore- and hindmilk sample Infant growth (meas-
[57, 58] Vuori (mean, 28 yr) were primiparous mothers of Wk 2: 4.0 (n = 20) from each feeding (5–7/day ured by height and
term infants Wk 3: 3.0 (n = 23) in first 3 mo; 3–4/day there- weight) was com-
Wk 4: 2.5 (n = 22) after) at 1–2-wk intervals up parable to reference
Wk 5: 2.4 (n = 25) to 2 mo and 3–4-wk inter- norms for Finnish
Wk 6: 2.1 (n = 17) vals thereafter infants. Infant zinc
Wk 7: 1.9 (n = 18) status not reported
Wk 8–10: 1.3 (n = 22)
Wk 11–13: 1.1 (n = 18)
Wk 14–16: 0.95 (n = 11)
Wk 17–19: 0.78 (n = 12)
Wk 20–22: 0.75 (n = 13)
Wk 23–25: 0.49 (n = 10)
Wk 26–28: 0.52 (n = 9)
Wk 29–31: 0.44 (n = 6)
Wk 32–34: 0.42 (n = 1)
Wk 35–37: 0.48 (n = 2)
Germany, 1992 n = 10 healthy mothers aged Healthy; zinc not measured Median (10th, 90th At each feed 2 mL of fore- or Negative zinc bal-
[55] Sievers 26–35 yr with term infants specifically percentile): hindmilk from nursed breast ance observed in
Day 17: 3.6 (2.4, 4.9) (n = 2,207 specimens) 11 of 33 breastfed
Day 35: 2.6 (1.6, 3.6) infants; median daily
Day 56: 1.7 (1.1, 2.8) zinc retention was
Day 85: 1.3 (0.8, 1.4) positive in all cases.
Day 117: 1.2 (0.6, 1.9) Infant growth not
reported
Greece, 2005 n = 180 mothers aged 25 ± Healthy; zinc status not spe- Mean ± SD: 10–20 mL collected in the Infant growth and zinc
[49] Leotsinidis 4 (SD) yr with full-term cifically measured. Anemia Day 3: 4.91 ± 1.73 morning by manual expres- status not reported
infants, mean birthweight prevalence was 41.9% Day 14: 2.99 ± 0.92 sion, 2 h after the previous
3,396 ± 480 g breastfeeding
Honduras and n = 263 mother-infant pairs Maternal zinc status not Mean ± SD at 9 mo: 10-40 mL collected by manual Infant growth and zinc
Sweden, 2004 (total for Honduras + reported. Prevalence of low Honduras: 0.7 ± 0.18 (n = 108) expression at 9 mo postpar- status not reported
[40] Dömellof Sweden); mothers > 16 yr of ferritin (< 12 μg/L) was 32% Sweden: 0.46 ± 0.26 (n = 86) tum in the morning > 1 h
age; exclusively breastfeed- in Honduras and 12% in after feeding
ing at 4 mo. Infants were Sweden
full term, with birthweight >
2,500 g, no illness
K. H. Brown et al.
India, 1997 Mothers of 68 preterm infants Maternal health and zinc Mean ± SD, term infants only: Full breast expression from Mean ± SEM plasma
[16] Hemalatha (gestational age at birth not status not reported Colostrum: 2.27 ± 0.191 the left breast between 10 zinc (μg/dL):
(colostrum and given), 92 term breastfed Transitional: 2.19 ± 0.297 and 11am Cord: 118.3 ± 5.07
transitional milk infants, and 26 term for- Mo 1–3: 1.90 ± 0.20 Mo 3: 76.9 ± 8.85
data not included mula-fed infants. All weaned Mo 4–6: 1.50 ± 0.225 Mo 6: 74.2 ± 5.12
because day of with cereal at 6 mo Mo 7–12: 1.15 ± 0.15 Mo 9: 83.8 ± 7.84
collection not Yr 1: 114.6 ± 13.02
reported) Infant growth not
reported
Italy, 1999 n = 22 mothers of appropriate Healthy, nonsmokers, non- Mean ± SD, placebo group 10 mL of milk was obtained No difference in gains
[26] Chierici weight for gestational age vegetarians; uncomplicated only (n = 11): from both breasts with a in infant weight,
term infants. Half were ran- pregnancy, labor, and Day 3: 8.16 ± 2.96 breast pump before the baby length, or head cir-
domized to receive vitamin delivery Day 30: 3.9 ± 1.01 nursed cumference between
and mineral supplements Day 90: 2.87 ± 1.23 the 2 maternal study
groups
Japan, 1982 n = 65 mothers 21–37 yr of Mothers judged to have good Mean ± SD: 10 mL of milk obtained by Infant growth and zinc
[23] Higashi (colos- age (mean, 27.3 yr) with nutritional status based on Colostrum: 10.39 ± 4.43 (n manual expression in the status not reported
trum data excluded healthy, full-term infants clinical examination. Mean = 65) morning before the baby was
because day of with birthweight > 2,500 g serum zinc ranged from 55 Wk 1: 4.56 ± 3.01 (n = 65) due to be fed
collection not to 115 μg/dL Mo 1: 2.66 ± 1.03 (n = 65)
specified) Mo 3: 1.14 ± 0.67 (n = 45)
Mo 5: 1.05 ± 0.46 (n = 35)
Japan, 2005 n = 1,197 mothers under 40 Zinc status not specifically Mean ± SD: ~50 mL of milk obtained at Mean birthweight,
[61] Yamawaki yr of age (mean, 29.12 ± measured Days 1–5: 4.75 ± 2.48 (n = 20) an intermediate time during 3,135 ± 346 (SD) g.
3.99 [SD] yr) who did not Days 6–10: 3.84 ± 1.39 (n = suckling Infant zinc status not
smoke or use vitamin sup- 38) specifically reported
plements, with infants with Days 11–20: 3.37 ± 0.89 (n
no symptoms of atopy and = 40)
birthweight > 2,500 g (mean, Days 21–89: 1.77 ± 1.08 (n =
3,135 ± 346 [SD] g) 551)
Days 90–180: 0.67 ± 0.8 (n =
476)
Days 181–365: 0.65 ± 0.43 (n
= 39)
Nigeria, 1982 n = 20 low-SES mothers with Maternal health and zinc Mean ± SD: 10 mL of milk collected by Infant growth and zinc
[36] Atinmo term infants status not reported Days 2–4: 5.98 ± 1.12 manual expression at noon status not reported
Days 8–14: 5.49 ± 0.73 just before infant feeding
Mo 1–2: 3.93 ± 0.78
continued
S151
Nigeria, 1984 n = 240 low-SES mothers Mothers judged healthy by Mean ± SEM: 8–12 mL of milk collected Infant growth and zinc
[50] Mbofung 17–40 yr of age who had clinical examination and Days 1–7: 5.83 ± 0.13 (n = 96) between 9 am and noon status not reported
given birth to a normal, full- medical history; zinc not Wk 4: 3.2 ± 0.10 (n = 54) from the breast due for the
term infant specifically measured Wk 8: 3.0 ± 0.10 (n = 34) next feeding by the mother
Wk 12: 2.3 ± 0.11 (n = 84) using manual expression
Wk 16: 1.88 ± 0.13 (n = 40)
Wk 20: 1.08 ± 0.10 (n = 31)
Wk 24: 0.98 ± 0.08 (n = 21)
Wk 32: 0.82 ± 0.09 (n = 12)
Wk 36: 0.68 ± 0.10 (n = 8)
Spain, 1997 n = 57 lactating women aged Comparison of healthy moth- Mean ± SD, mothers with low 5-mL samples collected from Normal growth
[52] Ortega 18–35 yr with healthy infants ers with low zinc intake (< zinc intake: each breast pre- and post- No difference in infant
with normal birthweight and 50% RI; mean serum zinc, Days 13–14: 3.05 ± 0.48 feeding between 10 and 11 growth between
a mean gestational age of 78.5 ± 9.8 μg/dL) and high Day 40: 1.88 ± 0.41 am groups
39.4 wk zinc intake (≥50% RI; mean
serum zinc, 87.0 ± 17 μg/dL) Mean ± SD, mothers with high
(p< .05) zinc intake:
Days 13–14: 3.34 ± 0.60
Day 40: 2.16 ± 0.52
8% of mothers with low zinc
intake and none with high
zinc intake had < 0.75 mg/L
breastmilk zinc
Spain, 2001 n = 22 women, well-nourished Well nourished; zinc not Mean ± SD: Foremilk collected from both Healthy; infant growth
[56] Silvestre and healthy with term measured specifically Days 2–4: 7.99 ± 3.23 breasts prior to feeding; col- and zinc status not
infants Day 14: 3.31 ± 1.06 lection between 11 am and reported
Day 30: 2.41 ± 0.90 4 pm
Day 60: 1.40 ± 0.65
Day 90: 1.05 ± 0.71
Turkey, 2005 n = 20 mothers (mean age, 23 Healthy mothers; weight Mean ± SD, mothers of term Milk collected by manual Infant growth and zinc
[13] Ustundag ± 1 [SD] yr) of term infants ranged from 50 to 65 infants only: expression within 2 h of the status not reported
and n = 20 mothers (mean kg. Zinc not measured Days 0–7: 2.41 ± 0.28 first feeding in the morning
age, 21 ± 2 [SD] yr) of pre- specifically Days 7–14: 2.28 ± 0.19 between 8 and 11 am
term infants (< 37 wk gesta- Day 21: 2.39 ± 0.20
tional age); all mothers were Mo 2: 2.01 ± 0.18
nonsmokers
K. H. Brown et al.
UK, 1982 n = 10 mothers of term infants Healthy; zinc not specifically Mean ± SD: Complete expression from Infant growth and zinc
[44] Hibberd measured Day 2: 7.75 ± 1.33 both breasts at each feeding status not reported
Day 4: 4.51 ± 1.32 with the use of an electric
Day 6: 3.90 ± 1.23 breast pump; an aliquot was
Day 9: 3.32 ± 1.27 taken for analysis and sam-
Day 16: 2.81 ± 1.16 ples from each 24-h period
Day 23: 2.33 ± 1.31 were pooled
Day 30: 1.71 ± 1.15
Day 37: 1.71 ± 0.92
USA, 1979a n = 38 Caucasian women aged “Good or excellent” nutri- Mean ± SD: 150–200 mL of milk collected Infant growth and zinc
[59] Vaughan 19–42 yr, mothers of term tional status Mo 1–3: 1.60 ± 0.23 (n = 28) once per month at 4-wk status not reported
infants; 1–31 mo postpartum Maternal zinc intake (mg/ Mo 4–6: 1.05 ± 0.15 (n = 39) intervals. Mothers instructed
(average participation was 4 day): Mo 7–9: 0.75 ± 0.11 (n = 23) to collect milk over a 3–5-
mo); 22 primiparous and 16 Mo 4–6: 12.7 Mo 10–12: 0.63 ± 0.09 (n = 13) day period in the morning,
multiparous Mo 7–9: 17.3 Mo 13–18: 0.69 ± 0.18 (n = 28) afternoon, and evening at
Mo 10–12: 16.5 Mo 19–31: 0.60 ± 0.19 (n = 30) random intervals within
Mo 13–18: 16.1 Total: 15.0 ± 4.2 feeding
Mo 19–31: 13.1
Serum zinc concentration (μg/
dL):
Mo 4–6: 230
Mo 7–9: 220
Mo 10–12: 140
Mo 13–18: 150
Mo 19–31: 220
USA, 1983a n = 102 mothers aged 16 to n = 39 (38%) took self-pre- Mean ± SD: 30 mL (1/3 foremilk, 1/3 Infant birthweight
[7] Feeley 38 yr with healthy full-term scribed zinc supplements Days 4–7: 5.36 ± 0.02 (n = 91) halfway through feeding, ranged from 2.5 to
infants; no difficulties during at a dose of 15 mg/day; n = Days 10–14: 4.22 ± 0.01 (n = 1/3 hindmilk) milk sample 4.8 kg with a mean
delivery. Middle SES; 97 2 (2%) took self-prescribed 163) obtained early morning and of 3.5 ± 0.5 (SD) kg.
Caucasian, 3 African-Ameri- zinc supplements at a dose of Days 30–45: 2.99 ± 0.01 (n = late evening by the mother Infant zinc status not
can, 2 Asian-American 25 mg/day 158) with the use of manual reported
No association between use of Means for each stage of lacta- expression or pump
zinc supplements and milk tion are significantly differ-
zinc concentration ent (p< .05)
USA, 1983b Longitudinal study of n = 6 Healthy, nonsmoking, limited Mean ± SEM: Entire contents of a single Infant growth and zinc
[43] Garza mothers 26–35 yr of age coffee, tea, and alcohol, no Mo 6: 1.2 ± 0.2 (n = 5) breast obtained with an status not reported
with healthy term infants medications. Zinc not spe- Mo 6.5: 1.2 ± 0.4 (n = 6) Egnell pump 3 to 4 h after
with appropriate weight for cifically measured Mo 7: 1.2 ± 0.3 (n = 6) the previous feeding,
gestational age; and exclu- Mo 7.5: 0.8 ± 0.3 (n = 6) between 8 am and noon
sively breastfeeding until Mo 8: 0.7 ± 0.3 (n = 5)
beginning of study (5–7 mo Mo 8.5: 0.8 ± 0.3 (n = 6)
postpartum) Mo 9: 0.7 ± 0.1 (n = 4)
continued
S153
USA, 1983c n = 23 mothers of term infants Mean ± SEM maternal zinc Mean ± SEM: 30 mL collected at the first Infant growth and zinc
[51] Moser recruited from private intake (mg/day): Mo 1: 2.6 ± 0.2 (n = 21) morning feeding after 6 am status not reported
obstetrics clinic. Mean age Wk 37 gestation: 9.8 ± 0.6 Mo 3: 1.3 ± 0.1 (n = 20) containing ½ foremilk and ½
was 30 ± 2 (SEM) yr; mean (n = 23) Mo 6: 1.1 ± 0.1 (n = 18) hind milk
education was 15 ± 2 (SEM) Mo 1 postpartum: 9.4 ± 0.5
yr (n = 23)
Mo 3 postpartum: 12.8 ± 1.8
(n = 20)
(n = 19)
Mean ± SEM maternal plasma
zinc (μg/dL):
Wk 37 gestation: 63.8 ± 2.0
(n = 23)
(n = 23)
(n = 21)
(n = 19)
USA, 1984a Longitudinal study of n = 8 Mothers of term infants Mean ± SD, term infants only: Full breast expression with Mean birthweight,
[17] Butte preterm infants (gestational 20–35 yr of age who did Wk 2: 3.4 ± 0.8 breast pump between 8 and 2.99 ± 0.46 (SD) kg.
age, 30–36 wk; mean, 33.9 not consume > 2 servings/ Wk 4: 2.9 ± 0.9 12 am, at least 2 hours after Infant zinc status not
wk; birthweight, 1,920 ± day of coffee, tea, alcohol, or Wk 6: 2.1 ± 0.9 feeding reported
70 [SD] g) and n = 13 term carbonated beverages; parity Wk 8: 1.9 ± 0.6
infants (gestational age, ≤ 2, no routine medications. Wk 10: 1.8 ± 1.0
37–42 wk; mean, 39.2 wk; Maternal health was judged Wk 12: 1.4 ± 0.7
birthweight, 2,990 ± 46 [SD] to be good by health history
g)
USA, 1985a n = 11 mothers aged 26-39 yr. Apparently well-nourished, Mean ± SD: 5 mL collected from both Mean birthweight,
[38] Casey White, middle-class, well with uncomplicated preg- Day 1: 7.94 ± 5.65 (n = 7) breasts in the middle of the 3.33 ± 0.46 (SD) kg
nourished. Mean parity was nancy; zinc not measured Day 2: 11.5 ± 4.7 (n = 8) midmorning feeding (2 min (range, 2.70–4.05);
2.5 (range, 2–4) specifically Day 3: 9.34 ± 1.97 (n = 9) after letdown) mean weight at 1 mo,
Day 4: 6.40 ± 0.66 (n = 9) 3.95 ± 0.33 (SD) kg.
Day 5: 5.42 ± 1.11 (n = 8) Infant zinc status not
Day 8 ± 2: 4.74 ± 1.02 (n =10) reported
Day 14 ± 3: 3.88 ± 0.91 (n =10)
K. H. Brown et al.
Day 21 ± 3: 3.71 ± 1.09 (n = 9)

Day 28 ± 3: 2.98 ± 0.78 (n = 8)
USA, 1985b Middle-income, healthy lac- Healthy; some took zinc sup- Mean ± SD: 5 mL collected by manual Infant growth and zinc
[5, 45] Krebs (only tating women, longitudinal plements during pregnancy Placebo group: expression. Pooled samples status not reported
data from placebo trial, zinc supplement (ZS): n [47]. NZS: mean dietary zinc Mo 1: 2.65 ± 0.81 from 6 collections/day
group included) = 14, placebo (NZS): n = 39. intake 10.7 ± 4.1 mg/day Mo 9: 0.67 ± 0.40
Mean infant age was 3.4 mo Zinc supplement group:
Mo 1: 2.83 ± 1.05
Mo 9: 0.82 ± 0.54
(at mo 9, 8 subjects remained
in placebo and 4 in zinc sup-
plement group)
USA, 1987 n = 45 mothers of healthy, Healthy; zinc not measured Mean ± SD: 24-h alternate breast pooled Infant growth pro-
[37] Butte full-term, exclusively breast- specifically Mo 1: 2.37 ± 0.82 sample: full breastmilk gressed satisfactorily.
fed infants aged 0–4 mo Mo 2: 1.55 ± 0.62 samples (entire content of 1 Infant length posi-
Mo 3: 1.13 ± 0.52 breast), at each feeding over tively associated with
Mo 4: 1.03 ± 0.52 24-h period zinc intake at mo 1
Interindividual variability and 2
(CV), 0.33
Intraindividual variability
(CV), 0.42
USA, 1989 n = 13 well-educated moth- Healthy mothers, nonsmok- Mean breastmilk zinc peaked 2–5 mL obtained by manual Infant growth and zinc
[39] Casey ers (mean age, 31.9 ± 4.4 ers, all multiparous. Dietary at 2 days, then declined expression from both status not reported
[SD] yr), median income assessment: mean zinc throughout lactation; decline breasts at midfeeding in the
> US$35,000, with normal intake, 10.9 ± 0.5 (SD) mg/ was nonlinear and more morning
prepregnancy weight day; 5/13 took supplemental rapid in 1st 2 wk. Higher
and weight gain during zinc (additional 10–45 mg/ zinc concentrations at 3, 6, 9,
pregnancy. Uneventful day); no effect of supplement and 12 mo were not associ-
pregnancy reported ated with supplemental zinc
intake
Mean ± SD (n = 13):
Day 7: 4.70 ± 1.20
Mo 1: 2.90 ± 0.70
Mo 12: 0.46 ± 0.30
continued
S155
USA, 1995 Healthy, lactating women. Healthy (plasma zinc was Mean ± SD (n = 71): Milk collection: 1st 29 Healthy infants. Mean
[28] Krebs (pooled Prospective, randomized, 79.8 μg/dL in both the zinc- Mo 0.5: 3.88 ± 0.99 subjects, 5–10 mL hand- weight-for-age per-
data from maternal double-blind, controlled supplementation and the Mo 1: 3.06 ± 1.12 expressed midfeeding from centile was 62 at 2
supplementation trial; maternal zinc sup- non-zinc-supplementation Mo 2: 2.04 ± 0.83 both breasts during 3-day mo, 33 at 7 mo, and
trial) plementation group (15 mg/ groups, compared with 88.3 Mo 3: 1.48 ± 0.65 period. 2nd 42 subjects, 3 25 at 9 mo of age.
day) (n = 40), placebo group μg/dL in the nonlactating Mo 4: 1.44 ± 0.66 midfeeding samples/day for Mean length-for-age
(n = 31) control group) Mo 5: 1.16 ± 0.62 3 days, with > 4 h between percentile was 43 at
Mo 6: 1.09 ± 0.67 samples 2 wk, 28 at 7 mo, and
Randomized, controlled Mo 7: 0.85 ± 0.51 26 at 9 mo of age [48]
trial. Maternal zinc supple- Mo 8: 0.87 ± 0.58
mentation, 15 mg/day. No Mo 9: 0.78 ± 0.51
difference in milk zinc con-
centration between groups;
thus, data were pooled
USA, 1996 n = 9 mothers (individually Healthy; zinc not measured Mean ± SD for each infant: 5–10 mL expressed at each Mean weight change
[46] Krebs identified as: A–I) of normal, specifically Mo 2.0: A (1.37 ± 0.2), B (1.96 feeding over 3 days during 1-wk study
term, exclusively breastfed ± 0.26), period was 20 ± 7.9
infants aged 2–5 mo Mo 2.3: C (1.01 ± 0.28), D g/day. Mean total
(1.96 ± 0.25) zinc absorbed was
Mo 4.0: E (2.76 ± 0.3), F (1.69 0.61 ± 0.23 mg/day
± 0.32),
Mo 4.5: G (1.40 ± 0.21)
Mo 4.8: H (1.33 ± 0.15)
Mo 5.0: I (1.08 ± 0.14)
RI, recommended intake; SES, socioeconomic status
K. H. Brown et al.
TABLE 2. Publications excluded from the data analysis
Zinc concentration in
Country, year Health and zinc status of Description of milk milk according to stage of Growth or zinc status Rationale for
[reference] author Study group mothers sample lactation (mg/L) of infant exclusion
Bangladesh, 1990 n = 34 lactating mothers in No information on All milk extracted from Mean ± SD: No information on No information on
[82] Simmer rural area maternal health or zinc both breasts over 24-h Mo 1: 1.96 ± 0.99 infant growth or preterm births
status period using mechan- Mo 2–3: 1.25 ± 0.54 zinc status
ical pump; 24-h infant Mo 6: 0.93 ± 0.41
test-weighings per- Mo 9: 0.73 ± 0.61
formed on 2 days Mo 12: 0.54 ± 0.27
Bangladesh, 1996 n = 34 low-SES mothers in 2/3 of mothers had BMI 5 mL collected from Mean ± SD: No information on Data not presented by
[9] Hussain Dhaka in wk 6–36 of lac- < 20 alternate breasts by 5–7 am: 2.10 ± 0.83 infant growth or stage of lactation
tation. Mean age, 22 yr manual expression 2–3 pm: 1.74 ± 0.53 zinc status
10 pm: 1.84 ± 0.69
Brazil, 1985 n = 8 mothers living in a General decrease in ~5 mL collected at Breastmilk zinc concen- Infants grew normally Numerical results not
[66, 67] Dorea low-income community. WHZ over 0–6 mo beginning and end of tration reported as a during the study reported
Mean age, 22.5 yr lactation but remained 1 nursing period at figure [67] period.
> 90% WHZ reference the same time of day Breastmilk zinc was a
population significant predictor
of weight gain (p =
.0006) and linear
growth (p = .02)
Brazil, 1989b n = 25 Amazonian mothers Generally inadequate Foremilk of the fullest Mean ± SD: No information on No information on
[71, 72] Lehti nutritional status breast collected by Mo 0–1: 2.2 ± 0.9 (n = 37) infant growth or preterm births
manual expression Mo 1–2: 1.6 ± 0.9 (n = 59) zinc status
between 8:30 and Mo 2–3: 1.4 ± 0.9 (n = 41)
11:30 am
Brazil, 2007 n = 117 mothers (mean Mothers were apparently Manual expression on 1.22 ± 0.78 mg% [sic] No information on No information on
[73] Melnikov age, 22.1 yr) and their healthy. day 2 postpartum infant growth or preterm births
infants. None of the Breastmilk zinc was not zinc status
infants showed malfor- related to maternal age,
mations or clinically parity, or history of
detectable impairment miscarriage
China, 2002 n = 18 mothers (mean age, Mean maternal zinc 5 mL collected by At ~2 mo postpartum, No information on No information on
[81] Sian 26 ± 2 [SD] yr) of infants intake estimated by manual expression of mean breastmilk zinc infant growth or preterm births
~2 mo of age (mean age, 3-day diet records, 7.8 midfeeding sample concentration was 2.34 zinc status
48 ± 11 [SD] days) ± 1.7 mg/day (n = 18) ± 0.83 mg/L. Milk zinc
output was 2.01 ± 0.97
mg/day
continued
S157
TABLE 2. Publications excluded from the data analysis (continued) S158
Côte d’Ivoire, n = 33 mothers in a rural General health was good Aliquot taken from Mean ± SD: Weight-for-age No information on
[70] Lauber area; 8 were primiparous. and nutritional status full breast expression Mo 1: 3.5 ± 0.9 (n = 4) declined from mo 5 preterm births
19 infants were aged 1–3 was fairly satisfactory by electric pump in Mo 6: 2.3 ± 0.9 (n = 7) onward and leveled
mo and 10 infants were early morning, from Mo 12: 1.6 ± 0.9 (n = 8) off at mo 10
aged 4–15 mo the breast not used Mo 18: 1.5 ± 1.3 (n = 6) Height-for-age
to feed the infant the remained at 94%–
previous night 96% of Harvard
standards; practi-
cally no change in
weight-for-height
Croatia, 1996 n = 29 mothers of wide SES No information on 80 mL collected Mean (range): Mean cord blood zinc No information on
[24] Frković range; mean age, 28.9 yr maternal health or zinc Days 2–12: 4.98 was 118 ± 21 μg/dL preterm births
(range, 17–45 yr); mean status (1.69–11.60)
parity, 1.8
Ethiopia, 2003 Rural Ethiopian mothers Mothers of stunted and ~15-mL sample col- Age-adjusted breastmilk Infants of mothers No information on
[83] Umeta with infants aged 5–11 nonstunted children lected from right zinc (5–11 mo post- with low breastmilk preterm births and
mo; n = 253 (83%) agreed breast, ~60 min after partum) of mothers of zinc were more wide infant age
to give breastmilk sample last feeding stunted infants (mean ± stunted ranges
SEM): 0.60 ± 0.03 mg/L
(n = 92) vs. 0.68 ± 0.02
mg/L (n = 161) in moth-
ers of nonstunted infants
(p = .02)
Finland, 1980 n = 28 mothers aged 24–35 Lowest individual zinc 8-mL aliquots at begin- Mean ± SD: No information on No information on
[84] Vuori yr (mean age, 28 yr). All intake was 9.3 mg/day. ning and end of each 1st survey week: 1.89 ± infant growth or preterm births
mothers belonged to 2 No correlation between feeding during a 24-h 0.74 zinc status
highest (out of 3) SES zinc levels in diet and period and pooled to 2nd survey week: 0.72 ±
groups. Mean height, 164 milk. Mean ± SD zinc one sample 0.44 (p < .001)
cm; mean weight, 54.4 intake (range): 13.7 ±
kg prepregnancy, 59.1 kg 2.7 (10.0–19.4) mg/day
after delivery, 57.6 kg at at 1st survey week, 12.8
1st survey week (6–8 wk ± 2.8 (9.3–18.6) mg/day
postpartum), 55.5 kg at at 2nd survey week
2nd survey week (17–22
wk postpartum)
K. H. Brown et al.
Finland, 1994 n = 200 exclusively breast- Healthy, nonsmoking Milk was manually Median (range) for pla- No significant differ- All women received
[25] Salmenperä feeding mother–infant mothers with uncom- expressed; 10 mL of cebo group (received ence in serum zinc supplemental
pairs (n = 116 exclusively plicated pregnancy milk was pooled from iron): between exclusively micronutrients
breastfed to 6 mo; n = 36 and delivery; full-term the beginning and Days 4–5: 4.75 (3.27–6.9) breastfed infants
exclusively breastfed to healthy infants of end of each feeding (n = 75) and infants fed com-
7–9 mo). Mothers sup- appropriate weight for Mo 2: 1.41 (1.1–2.19) (n plementary foods
plemented with 20 mg/ gestational age = 77) at 7.5–9 mo. Milk
day or 40 mg/day zinc or Mo 4: 0.9 (0.58–1.38) (n zinc concentration
placebo = 67) greater in mothers
Mo 6: 0.67 (0.4–1.13) given 40 mg/day

(n = 56) zinc than in those
Mo 7.5: 0.61 (0.39–0.97) given 20 mg/day
(n = 31) zinc at 6 and 7.5 mo
Mo 9: 0.6 (0.38–0.95) (p = .02)
(n = 14)
Mo 10: 0.61 (0.42–0.87)
(n = 8)
Mo 11: 0.43 (0.33–0.57)
(n = 6)
Mo 12: 0.43 (0.33–0.56)
(n = 5)
Guatemala, 2005 n = 56 women with healthy, Zinc not specifically Full expression with Mean ± SD: No information on Women with para-
[65] Dhonukshe- exclusively breastfed measured; 17/42 had hand pump plus Mo 1: 3.85 ± 0.76 (n = 4) infant growth or sites had higher
Rutten infants aged 1–6 mo in intestinal parasites manual expression > Mo 2: 2.38 ± 0.53 (n = 14) zinc status breastmilk zinc than
periurban Guatemala 1 h after last feeding Mo 3–6: 1.53 ± 0.64 (n non-infected coun-
City and low-income from breast not used = 29) terparts: 2.30 ± 1.10
rural area in last feeding (SD) vs. 1.80 ± 0.82
mg/L
India, 1980 Cross-sectional study: n = High-income women had 10 mL collected by Breastmilk zinc values No information on No information on
[6] Rajalakshmi 412 low- income urban greater height, weight, manual expression according to stage of infant growth or preterm births
mothers, n = 100 high- MUAC, and triceps just prior to infant lactation were available zinc status
income urban mothers, n skinfold thickness than feeding. Day-to-day from cross-sectional and
= 208 low-income rural low-income women variation assessed by semilongitudinal stud-
mothers Mean hemoglobin, 13.0 daily collection for ies, as well as data on
Semilongitudinal study: ± 0.15 (SEM) g/dL in 5–7 days. Diurnal diurnal and day-to-day
n = 24 low-income low-income group and variation assessed by variation in breastmilk
mothers 12.9 ± 0.26 (SEM) g/dL collections at 10 am zinc
in high-income group and 6 pm in 6 women
continued
S159
India, 1999 Mothers of 155 infants of Maternal serum zinc Milk samples (colos- Mean ± SD: Breastmilk zinc lower Time of milk collec-
[19] Sharda gestational age 26–41 concentration ranged trum) obtained by Gestational age 37–41 wk: in preterm than in tion postpartum not
wk and birthweight from 57.5 to 76.5 μg/dL mechanical expres- 5.03 ± 0.078 (n = 5) term infants and reported
550–3,800 g Serum zinc concentra- sion from both breasts Term, appropriate weight in low-birthweight
Appropriate for gestational tion in nonpregnant for gestational age: 5.04 than in normal-
age and small for gesta- women was 99.4 ± 14.4 ± 0.072 (n = 3) birthweight infants
tional age not defined; (SD) μg/dL Term, small for gestational No difference in milk
time of collection of milk age: 4.98 ± 0.098 (n = 2) zinc concentration
samples not reported Gestational age 26–30 wk: between mothers of
4.75 ± 0.078 (n = 3) term infants with
Gestational age 31–33 wk: different birth-
4.71 ± 0.098 (n = 11) weights and mothers
Gestational age 34–36 wk: of preterm infants
4.75 ± 0.10 (n = 11) with different
Preterm, appropriate birthweights
weight for gestational Mean ± SD milk zinc:
age: 4.72 ± 0.12 (n = 17) Birthweight 0.5–1.5
Preterm, small for gesta- kg: 4.75 ± 0.085
tional age: mg/L (n = 12)
4.75 ± 0.072 (n = 8) 1.5–2.5 kg: 4.77 ± 1.64
mg/L (n = 14)
2.5–4.0 kg: 5.00 ± 0.14
mg/L (n = 4)
Indonesia, 1998 n = 92 middle-income Mean BMI, 22.0; 40% Entire milk content Median: 2.7 Mean HAZ, WHZ, Data from wide infant
[68] Gross mothers in a poor urban anemic (hemoglobin < from the breast last 10th, 90th percentiles: WAZ near 0 and age range presented;
area (East Jakarta), 25.4 ± 120 g/L); mean plasma suckled was collected 1.3, 4.7 distribution similar values for milk
5.2 (SD) yr of age zinc, 85.5 ± 24.2 (SD) with a manual pump Range: 0.5–12.8 (n = 91) to NCHS reference zinc concentration
Mean infant age, 2.4 ± 1.4 μg/dL (increased with during 5 consecutive population according to infant
(SD) mo (range, 0.1–5.2 time postpartum); 29% days between 9 and age not presented
mo) had low plasma zinc 11 am numerically
concentrations
Italy, 1993 n = 26 mothers aged 17–37 Apparent good health Full breast manual Median ± SD mg zinc/ Healthy; zinc status Data presented as
[22, 78] Perrone yr with term infants and expression at the 2nd kg dry weight of breast- not specifically pooled wide range
6 mothers with preterm nursing, 9–11 am milk, term infants: measured of infant ages; zinc
infants. No supplementa- Wk 1: 36.4 ± 2.8 (n = 46) measured as mg per
tion during pregnancy or Wk 2: 24.2 ± 1.6 (n = 15) dry weight
lactation Wk 3: 28.6 ± 6.8 (n = 19)
K. H. Brown et al.
Wk 4: 21.7 ± 1.4 (n = 59)

Japan, 2003 n = 68 Japanese mothers No information on Milk samples col- Mean ± SD, 5–8 days No information on No information on
[69] Honda aged 19–38 yr maternal health or zinc lected at 5–8 days postpartum: infant growth or preterm births
status postpartum Mother > 35 yr: 5.41 ± zinc status
Details of milk collec- 1.44
tion not specified Mother < 35 yr: 5.90 ±
1.83
Nullipara: 6.27 ± 1.92
Multipara: 5.35 ± 1.5
Cesarean delivery: 5.8 ±
1.58
No Cesarean delivery:
5.83 ± 1.84
Kenya, 2002 Mothers of 250 tod- Not supplemented; Details of milk collec- Breastmilk zinc content No information on Breastmilk composi-
[77] Onyango dlers (mean age, 13.9 zinc not specifically tion not specified not directly measured infant growth or tion taken from
mo; range, 12–26 mo) measured zinc status another study in The
from rural, low-income Gambia; no infor-
families mation on preterm
births
Kuwait, 2000 n = 34 mid-upper-class, Healthy; none had Manual expression Mean ± SD: No information on Pooled data presented
[62] Al-Awadi lactating mothers of term plasma zinc below ref- between 6 and 8 am Mo 0–6: infant growth or from wide range of
infants living in Kuwait. erence values before the infant’s 1st Kuwaiti: 3.2 ± 0.12 zinc status infant ages
Maternal age range was feeding Non-Kuwaiti: 2.4 ± 0.06
25–40 yr Mo 6-12:
Kuwaiti: 2.4 ± 0.14
Non-Kuwaiti: 1.9 ± 0.05
Mo 12-18:
Kuwaiti: 2.0 ± 015
Non-Kuwaiti: 1.7 ± 0.09
Nepal and USA, n = 26 Nepalese mothers Mean ± SD plasma zinc Early morning, 5–10- Mean ± SD: Mean ± SD infant Wide age ranges and
1988 2–6 mo postpartum, and (µg/dL): mL prefeeding sample Nepal: 1.10 ± 0.098 plasma zinc (µg/dL): results not disaggre-
[74] Moser n = 23 US mothers 3–6 USA: 1.20 ± 0.098 Nepal: 68.0 ± 5.9 gated by age
mo postpartum Nepal: 66.7 ± 2.0 USA: not determined
USA: 88.3 ± 2.0 (p< .05)
No significant difference
in dietary zinc intake
Nigeria, 2000 n = 15 mothers 26.2 ± 3.3 Healthy mothers; mean Breastmilk collected at Mean at ~mo 6: 1.52 Mean serum zinc con- No information on
[76] Okolo (SD) yr of age and infants serum zinc, 117 μg/dL 6 mo of lactation; no centration was 164 preterm births
6.12 ± 0.26 (SD) mo of information on col- μg/dL. Infant growth
age lection method not reported
continued
S161
The Gambia and n = 56 mothers (mean age, Maternal health and zinc 1–5 mL expressed Mean: No information on Mothers received
UK, 1990 29.1 yr) in Gambia and status not reported between feedings The Gambia infant growth or micronutrient sup-
[63] Bates n = 57 mothers (mean Mo 1: 4.22 zinc status plements; no infor-
age, 31.8 yr) in UK; stage Mo 2: 3.50 mation on preterm
of lactation varied from Mo 3: 2.78 births
0 to 99 wk postpartum Mo 4: 1.71
in Gambia and from 0 Mo 5: 1.69
to 112 wk postpartum Mo 6: 2.15
in UK UK
Mo 2: 1.34
Mo 3: 2.06
Mo 4: 0.87
Mo 5: 0.73
Mo 6: 0.73
Turkey, 2006 n = 21 mothers, 27.3 ± 5.4 Healthy; no chronic Manual full breast Mean ± SD: Increases in the body No information on
[79] Rakicioğlu (SD) yr, nonsmokers, diseases expression after the 1.5 ± 0.4 before Ramadan, length and weight of preterm births
not taking supplements; first nursing period of 1.8 ± 0.3 after Ramadan infants were within
infants aged 2–5 mo the day between 9 and (p = .001) the normal NCHS
11 am range
UK, 1993 Mothers of 39 normal, No information on No information on milk Infant zinc intake from No information on Zinc concentration of
[80] Richmond healthy children 17–61 maternal zinc status sample collection milk (geometric mean ± infant growth milk not presented.
wk of age SD): 1.06 ± 2.10 mg/day No information on
preterm births
USA, 1976 Mothers in 6th to 12th wk 37 of the 50 women were 1 40-mL sample/day at Mean ± SD: Healthy; zinc status Data not presented by
[10] Picciano of lactation (n = 50) with taking supplemental early morning feed- Daily period: 1.68 ± 0.78 not specifically stage of lactation;
healthy, full-term infants. vitamins. “A few sup- ing for 5 consecutive Weekly period: 1.59 ± 0.84 measured wide range of infant
42 mothers were 20–30 yr plements contained days (daily period). Within-day period: 1.58 ages
old and 8 were > 30 yr. 16 less than 10% RDA for 2 additional samples ± 0.81
were primiparous and 34 zinc” at weekly intervals
were multiparous (weekly period) or Mean:
within 1 day (within- Time of day: morn-
day period) ing 1.71, midday 1.54,
evening 1.49 (morning
value was significantly
K. H. Brown et al.
and evening values)
Age: 20–30 yr, 1.55; > 30
yr, 2.00 (p < .001)
Parity: 0, 0.39; ≥1, 1.75
(p < .001)
Lactation history: 0, 1.46;
≥1, 1.75 (p < .001)
USA, 1979b n = 52 Caucasian, middle- Healthy; zinc intake 33 obstetrical patients: Mean breastmilk zinc No information on Labels missing from
[8] Kirksey SES mothers aged 18–31 exceeded 2/3 of RDA 5–10 mL collected concentration decreased infant growth or data table
yr. 33 were patients at (includes supplements); after milk letdown from 4.72 mg/L on zinc status
an obstetrical clinic at zinc intake from diet at 1st morning feed- day 3 to 3.20 mg/L on
enrollment; 19 were was less than 2/3 RDA ing on days 3 and day 14. Correlation
members of La Leche for 30 women 14 postpartum and between zinc and iron
League (stages of lacta- prior to multivitamin in colostrum: r = 0.730.
tion in these women supplementation Levels of zinc in milk
ranged from 1 to 30 mo) 19 La Leche League were similar for supple-
patients: similar to mented subjects (mean
obstetrical patients, zinc intake, 28 mg) and
except milk collected unsupplemented sub-
on 3 consecutive days jects (mean zinc intake,
11 mg)
USA, 1982 n = 34 mothers recruited Healthy mothers 5–15-mL samples col- Mean: No information on Sample included 7 pre-
[86] Zimmerman at hospital (7 preterm lected after letdown Wk 1: 3.50 infant growth or term infants
infants, 27 full-term and before feeding, Wk 15: 0.850 zinc status
infants) at 1, 15, and 30 wk Wk 30: 0.550
(single samples) No difference between
breastmilk zinc of moth-
ers of preterm and full-
term infants
USA, 1984b n = 46 mothers during mo Apparently healthy; zinc Full breast expression Mean ± SD: Weight-for-age was No information on
[64] Dewey 7–20 of lactation aged status not specifically in the morning at the Mo 4–6: 0.79 ± 0.47 at or above the 25th preterm births
21–37 yr measured 2nd feeding of the (n = 38) percentile of NCHS
day, either manual or Mo 7–11: 0.42 ± 0.22 standards for 6 of 8
with an Egnell pump (n = 27) infants at 7 mo and 5
Mo 12–20: 0.33 ± 0.26 of 6 infants at 11 mo.
(n = 4) 4 of 6 infants had
decreased gradually
in weight for age
since mo 7
continued
S163
USA, 1985c n = 62 mothers with a Half of the mothers were Manual expression of 1 Data for milk zinc con- No information on No information on
[85] Finley mean age of 29 yr vegetarian and half breast in the morning centration according to infant growth or preterm births
were non-vegetarian. at the 2nd feeding of month postpartum are zinc status
Zinc status not specifi- the day presented as a figure
cally measured.
USA, 1988 n = 49 middle-income Apparently healthy and Milk samples collected Data for milk zinc con- No information on All women received
[11] Karra mothers; mean age, 28.8 adequately nourished; at every infant feeding centration according to infant growth or micronutrient sup-
yr; mean parity, 1.7 zinc not specifically over a 24-h period at month postpartum are zinc status plements; numerical
measured 1, 2, 3, 4, 5, and 6 mo presented as a figure results not reported
postpartum. 10 mL
of milk was expressed
manually or by breast
pump
USA, 1990 n = 20 lactating women Healthy; mean plasma Manual expression of Mean ± SEM, placebo No information on All mothers received
[75] Moser-Veillon aged 20–36 yr. Double- zinc values ranged from half-foremilk, half- group: infant growth or a multivitamin
blind, randomized, 77.2 to 96.1 µg/dL hindmilk, in the Wk 1: 4.62 ± 0.48 zinc status supplement
controlled trial. Multivi- morning Wk 2: 3.63 ± 0.33
tamins (n = 10) or mul- Wk 4: 2.73 ± 0.18
tivitamins + 25 mg zinc Wk 12: 1.72 ± 0.12
(n = 10). Supplementa- Wk 24: 1.37 ± 0.22
tion from 1 day to 36 wk Wk 36: 0.60 ± 0.22
postpartum
BMI, body mass index; HAZ, height-for-age z-score; MUAC, mid-upper-arm circumference; NCHS, US National Center for Health Statistics; RDA, recommended daily allow-
ance; SES, socioeconomic status; WAZ, weight-for-age z-score; WHZ, weight-for-height z-score
K. H. Brown et al.
TABLE 3. Breastmilk zinc concentration according to child’s

on zinc absorption from breastmilk. Moreover, because
agea
the FAZ varies according to the amount of zinc con-
Mean ± SD milk sumed in a particular meal [89, 90], and the amount
No. of zinc concentra- of milk (and hence zinc) consumed at a single feeding
Age group (mo) observations tion (mg/L) may be extremely variable, the available information
<1 74 4.11 ± 1.50 on mean FAZ from breastmilk should be considered a
1 to 2 42 1.91 ± 0.53 rough approximation, at best, of the true value for an
3 to 5 24 0.98 ± 0.35 individual child.
6 to 11 24 0.77 ± 0.22 Two studies have been published on infant FAZ from
12 to 21 2 0.64 ± 0.05 breastmilk. One study, which reported results from
a. Data were extracted from studies presented in table 1. Figures for
nine exclusively breastfed infants 2 to 5 months of age
milk zinc concentration represent weighted means of results from who received zinc stable isotope tracers along with six
all studies that provided information for respective age ranges. feedings per day, found a mean FAZ of 0.54 (range,
0.40 to 0.68) [46]. Another study of 10 non-exclusively
by 1 month. Breastmilk zinc transfer declines further, breastfed infants 5 to 7 months of age reported a mean
albeit more slowly, to ~0.7 mg/day by 6 months. The FAZ of 0.50 (range, 0.27 to 0.78) when they were fed
prediction interval displayed in the figure indicates a total of 250 mL of expressed breastmilk, along with
the ranges of values for daily milk zinc intakes that are other foods, over the course of three feedings during
likely to occur for children at a particular age. Because an 8- to 12-hour study period [91]. Breastmilk zinc
the available data do not permit adjustment for day-to- was highly bioavailable under the circumstances of
day (intraindividual) variability in milk zinc transfer, these studies, although the actual study conditions
the portrayed range overestimates the distribution of were not necessarily consistent with usual breastfeeding
children’s usual zinc intakes. Thus, the curves can be practices, so the ultimate implications of these results
used to describe the ranges of daily zinc intakes, but remain uncertain. In developing countries, the mean
they cannot be used reliably to assess the adequacy of frequency of breastfeeding is often greater than 12
usual zinc intake from breastmilk. times per day [1, 30, 32], so the amount consumed per
feeding may be less than that provided in the foregoing
Adequacy of breastmilk zinc transfer in relation to the zinc studies. In that case, the zinc intakes per feeding would
requirements of exclusively breastfed term infants less than be less, and the FAZ might be even greater than was
6 months of age observed. Because of the broad range of FAZ reported
There are two possible theoretical approaches to assess from these studies and the uncertainty regarding
the adequacy of breastmilk as a source of zinc for exclu- the true FAZ from different amounts of breastmilk
sively breastfed term infants: to compare usual zinc fed during real-life feeding conditions, we applied
intake and assumed absorption from breastmilk with assumptions for FAZ values ranging from 0.40 to 0.60
the physiological requirements for absorbed zinc, or to explore the range of zinc that might be absorbed
to test for any functional response after providing sup- from breastmilk by infants of different ages, according
plemental zinc to such infants. The former approach is to these assumptions. The results of these exploratory
particularly challenging, for several reasons. First, not calculations are shown in table 5.
only is information needed on the usual total amount of Several expert groups have published estimates of the
breastmilk zinc consumed at each age, but knowledge physiological requirements for absorbed zinc. A WHO
is also required regarding the absorption of zinc from committee calculated the amount of zinc that must be
this food source. Furthermore, the true physiological absorbed each day to replace endogenous losses, based
requirements for absorbed zinc at each age must be on data extrapolated from adults, and to allow for
known with some degree of certainty. Although infor- infant growth, based on estimates of zinc accretion in
mation can be compiled on zinc consumption from newly synthesized tissues of growing infants [92]. The
breastmilk at different ages (as described above), there WHO committee estimated that the requirements for
is less information on individuals’ usual zinc intakes or absorbed zinc derived by this method are 1.3 mg/day
TABLE 4. Amount of breastmilk transferred from mother to child (g/day) according to child’s age and breast-
feeding patterna
Age group (mo)
Breastfeeding pattern 0–2 3–5 6–8 9–11 12–13
Exclusive breastfeeding 714 ± 131 784 ± 128 — — —
Partial breastfeeding 644 ± 159 706 ± 146 674 ± 151 616 ± 172 549 ± 187
a. Data from Brown et al. [87]. Values represent mean ± SD for children from developing countries.
during the first 3 months of life and 0.7 mg/day for the supplemented boys had significantly greater weight and
period from 3 to 5 months. The Steering Committee length gains after 3 months of supplementation than
of the International Zinc Nutrition Consultative Group boys in the placebo group, but there were no significant
(IZiNCG) accepted the WHO recommendations but effects of supplementation among girls [96]. However,
further noted that these requirements may be partially these latter findings are difficult to interpret, because
offset by hepatic zinc reserves accumulated during differential dropout rates by treatment group may have
gestation [93]. confounded the results. Moreover, the infants in this
According to the comparisons in table 5, breastmilk trial were receiving nonquantified amounts of other
alone would provide the WHO/IZiNCG estimated foods, so it is uncertain whether the groups differed
mean requirements [92, 93] during the first month of only with regard to zinc supplementation. In summary,
life (without considering the possibility of drawing on the results of these three studies carried out in industri-
hepatic zinc reserves) if the FAZ is greater than 0.50 but alized countries are inconsistent, and none clearly dem-
would provide less than the mean requirement there- onstrates a uniformly positive impact of additional zinc
after. Nevertheless, IZiNCG concluded that breastmilk on the growth of infants under 6 months of age. Thus,
alone is an adequate source of zinc for approximately it seems that breastmilk alone is probably an adequate
6 months, based on the assumption that some reserves source of zinc for approximately the first 6 months of
present at birth could be used to meet physiological life for normal-birthweight, term infants in industrial-
demands and the fact that supplementation trials have ized settings. We could not locate any relevant infor-
not demonstrated consistent benefits of additional zinc mation from developing countries that was obtained
supplementation for exclusively breastfed term infants, specifically from exclusively breastfed infants.
as discussed below.
Zinc supplementation trials among breastfed infants less Section 2
than 6 months of age
Three intervention trials in which breastfed infants less How much zinc is transferred in breastmilk to infants and
than 6 months of age received either zinc supplements young children less than 24 months of age who are con-
or placebo have been reported from industrialized suming breastmilk in addition to complementary foods?
countries. However, in two of these trials, the period
of supplementation lasted until the infants attained 10 Conclusions
to 11 months of age, so the results do not specifically
reflect the effect of providing additional zinc to infants The amount of zinc transferred in breastmilk to par-
less than 6 months of age. In the one trial completed tially breastfed infants less than 6 months of age is
among term infants less than 6 months of age, zinc- approximately 15% less than that described above for
supplemented US girls gained 3 g/day more weight exclusively breastfed infants, because of the smaller vol-
than their nonsupplemented counterparts from 4 to umes of milk consumed by partially breastfed infants.
7 months of age, but there were no significant effects The age-related pattern of change in milk zinc intakes
on the linear growth of girls or on the weight or length is similar for both groups of infants. After 6 months of
gains of boys [94]. In another study of US term infants age, breastmilk continues to be an important source
4 to 10 months of age, there was no effect of zinc sup- of zinc, providing ~0.5 mg/day from 6 to 8 months
plementation on growth or other functional outcomes and ~0.4 to 0.3 mg/day thereafter, a time when the
among either girls or boys [95]. Finally, in a study estimated physiological requirements for absorbed zinc
conducted in France among infants 3 to 11 months range from 0.84 to 0.5 mg/day, depending on age and
old, most of whom were African immigrants, the zinc- the source of these estimates.
TABLE 5. Estimated daily zinc intake among exclusively breastfed infants less than 6 months of age and estimated total
absorbed zinc at three assumed levels of fractional zinc absorption (FAZ), and relation to estimated physiological require-
ments for zinc, according to infant’s age
Amount of absorbed zinc (mg/day), Estimated aver-
according to assumed FAZ age physiological
Milk zinc transfer requirements for
Age group (mo) (mg/day) FAZ = 0.40 FAZ = 0.50 FAZ = 0.60 zinc (mg/day)a
<1 2.52 1.01 1.26 1.51 1.3
1–2 1.37 0.55 0.68 0.82 1.3
3–5 0.86 0.34 0.43 0.52 0.7
a. Requirements from WHO and IZiNCG [92, 93]
Detailed review of evidence Assessment of the adequacy of breastmilk zinc trans-

fer in relation to theoretical requirements is even more
Studies of breastmilk zinc transfer to partially breastfed problematic for partially breastfed children than for
infants less than 24 months of age those who are exclusively breastfed. Because partially
Identification of studies and milk zinc concentration by breastfed children are consuming other foods in addi-
child age. The same process that was described above tion to breastmilk, these other foods may contribute
was used to identify and analyze studies of breastmilk additional zinc and they may affect zinc absorption
zinc transfer to partially breastfed infants and young from breastmilk [98]. Thus, we have not attempted to
children, and the same summary values for milk zinc assess the adequacy of breastmilk zinc transfer at dif-
concentration were used for the specific age intervals, ferent ages, and we simply present the mean amount
as shown in figure 2. of milk zinc that is transferred to demonstrate the
potential contribution that breastmilk might provide
Amount of milk consumed according to child age. As
in relation to these requirements. Clearly, breastmilk
with the preceding analysis, we extracted information
provides a potentially important proportion of the
on breastmilk volumes from the WHO/UNICEF publi-
daily zinc requirement, even among partially breastfed
cation on complementary feeding [87]. For the present
children in the second year of life.
analysis, we selected the age-specific information for
partially breastfed infants and young children from
developing countries, as summarized in table 4.
Section 3
Total zinc transfer in breastmilk to partially breastfed
children less than 24 months, and relation to require- What are the programmatic implications of the answers
ments. The amount of zinc transferred in breastmilk to the foregoing questions, and what are the remaining
to partially breastfed children was estimated using the research needs?
same simulation procedure described above. The analy- The foregoing analyses indicate that breastmilk
sis had to be limited to children less than 18 months of seems to provide an adequate amount of zinc to exclu-
age because there was only one data point available for sively breastfed term infants less than 6 months of age.
milk zinc concentration from children older than 18 Breastmilk continues to be an important source of zinc
months, and the available information on breastmilk for partially breastfed infants well into the second year
volumes provided a single figure for the entire age of life. Thus, breastfeeding should be promoted and
range from 12 to 23 months. Because the volume of supported as an intervention that contributes to the
milk consumed by partially breastfed infants less than 6 prevention of zinc deficiency. WHO currently recom-
months is approximately 15% less than the volume con- mends that “infants should be exclusively breastfed
sumed by exclusively breastfed infants, the total milk
zinc transfer to the former infants is correspondingly
less, although the age-related pattern of change is simi- TABLE 6. Estimated daily zinc intake among partially
lar, as summarized in figure 4 and table 6. Even after breastfed children less than 18 months of age and relation
9 months of age, breastmilk continues to provide an to estimated average physiological requirements for zinc,
average of ~0.4 mg/day to partially breastfed infants. according to child’s age
Estimated average physiological
requirement for zinc (mg/day)a
Age Milk zinc
Zinc transferred in breastmilk (mg/day)
8 group transfer WHO and

(mo) (mg/day) IZiNCG IOM, 2000
7
<1 2.18 ± 1.06 1.3 —
6
1–2 1.18 ± 0.56 1.3 —
5
3–5 0.73 ± 0.39 0.7 —
4
6–8 0.51 ± 0.33 0.8 0.84
3
9–11 0.39 ± 0.31 0.8 0.84
2
1
12–17 0.29 ± 0.31 0.8 / 0.5 0.74
a. WHO and IZiNCG provided similar estimates for children < 12
0
0 3 6 9 12 15 18 months of age, although IZiNCG assumed that breastmilk was
adequate for exclusively breastfed infants < 6 months of age, so
Age (mo)
the estimates in the table refer to non-exclusively breastfed infants
[92, 93]. At 12 to 17 months of age, the two estimates provided by
FIG. 4. Simulated mean and 95% prediction interval of daily WHO and IZiNCG differ. The US Institute of Medicine (IOM)
zinc transfer in breastmilk to partially breastfed children does not provide estimates of physiological requirements for zinc
according to age for children < 7 months of age [97].
during the first 6 months of life. Thereafter they should organizations to provide information, counseling, per-
receive nutritionally adequate and safe complementary sonal support, and referral when necessary [101].
foods while breastfeeding continues up to 2 years of Additional research is needed on the zinc require-
age and beyond” [99]. Thus, these infant and young ments of infants and young children and the bio-
child feeding recommendations are consistent with availability of zinc from breastmilk when fed in usual
the current conclusion that breastfeeding should be amounts, either alone or with other foods. This infor-
promoted as an important dietary source of highly mation is needed to help determine both the period
bioavailable zinc. for which breastmilk alone is an adequate source of
Information is available on how best to promote and zinc and how much additional zinc is needed from
support appropriate breastfeeding behaviors [100], and complementary foods. Information is also needed on
a detailed discussion of these issues is beyond the scope the period of adequacy of breastmilk for providing zinc
of the present review. Briefly, attention must be devoted to low-birthweight infants. Finally, different approaches
to advocacy, policy formulation, health curriculum for promoting optimal breastfeeding behaviors require
reform and in-service training of health service person- further evaluation in the context of large-scale pro-
nel, and the development of suitable community-based grammatic interventions.
References
1. Brown KH, Black RE, López de Romaña G, de Kanashiro and between-breast variation in selected components
HC. Infant-feeding practices and their relationship with of human milk. Am J Clin Nutr 1984;40:635–46.
diarrheal and other diseases in Huascar (Lima) Peru. 13. Ustundag B, Yilmaz E, Dogan Y, Akarsu S, Canatan H,
Pediatrics 1989;83:31–40. Halifeoglu I, Cikim G, Aygun AD. Levels of cytokines
2. Popkin BM, Adair L, Akin JS, Black R, Briscoe J, Flieger (IL-1beta, IL-2, IL-6, IL-8, TNF-alpha) and trace ele-
W. Breast-feeding and diarrheal morbidity. Pediatrics ments (Zn,Cu) in breast milk from mothers of preterm
1990;86:874–82. and term infants. Mediators Inflamm 2005;6:331–6.
3. Bahl R, Frost C, Kirkwood BR, Edmond K, Martines J, 14. Ronayne de Ferrer PA, Weisstaub A, López N, Ceriani
Bhandari N, Arthur P. Infant feeding patterns and risks Cernadas JM. Zinc levels in term and preterm milk.
of death and hospitalization in the first half of infancy: Arch Latinoam Nutr 2001;51:33–6.
Multicentre cohort study. Bull World Health Organ 15. Trugo NM, Donangelo CM, Koury JC, Silva MI, Freitas
2005;83:418–26. LA. Concentration and distribution pattern of selected
4. Castillo-Duran C, Vial P, Uauy R. Trace mineral micronutrients in preterm and term milk from urban
balance during acute diarrhea in infants. J Pediatr Brazilian mothers during early lactation. Eur J Clin Nutr
1988;113:452–7. 1988;42:497–507.
5. Krebs NF, Hambidge KM, Jacobs MA, Mylet S. Zinc in 16. Hemalatha P, Bhaskaram P, Kumar PA, Khan MM, Islam
human milk: Diurnal and within-feed patterns. J Pediatr MA. Zinc status of breastfed and formula-fed infants of
Gastroenterol Nutr 1985;4:227–9. different gestational ages. J Trop Pediatr 1997;43:52–4.
6. Rajalakshmi K, Srikantia SG. Copper, zinc, and magne- 17. Butte NF, Garza C, Johnson CA, Smith EO, Nichols BL.
sium content of breast milk of Indian women. Am J Clin Longitudinal changes in milk composition of mothers
Nutr 1980;33:664–9. delivering preterm and term infants. Early Hum Dev
7. Feeley RM, Eitenmiller RR, Jones JB Jr, Barnhart H. 1984;9:153–62.
Copper, iron, and zinc contents of human milk at early 18. Mendelson RA, Anderson GH, Bryan MH. Zinc, copper
stages of lactation. Am J Clin Nutr 1983;37:443–8. and iron content of milk from mothers of preterm and
8. Kirksey A, Ernst JA, Roepke JL, Tsai TL. Influence of full-term infants. Early Hum Dev 1982;6:145–51.
mineral intake and use of oral contraceptives before 19. Sharda B, Adhikari R, Ajmera M, Gambhir R, Singh PP.
pregnancy on the mineral content of human colos- Zinc and copper in preterm neonates: Relationship with
trum and of more mature milk. Am J Clin Nutr 1979; breast milk. Indian J Pediatr 1999;66:685–95.
32:30–9. 20. Friel JK, Andrews WL, Jackson SE, Longerich HP, Mercer
9. Hussain M, Sikder ZU, Wahed MA, Haque AT, Jahan C, McDonald A, Dawson B, Sutradhar B. Elemental
F. Zinc concentration of breast milk and its diurnal composition of human milk from mothers of premature
variation in Bangladeshi mothers. Bangladesh Med Res and full-term infants during the first 3 months of lacta-
Counc Bull 1996;22:70–3. tion. Biol Trace Elem Res 1999;67:70–7.
10. Picciano MF, Guthrie HA. Copper, iron, and zinc 21. da Costa RS, do Carmo MG, Saunders C, de Jesus EF,
contents of mature human milk. Am J Clin Nutr Lopes RT, Simabuco SM. Characterization of iron,
1976;29:242–54. copper and zinc levels in the colostrum of mothers of
11. Karra MV, Kirksey A. Variation in zinc, calcium, and term and pre-term infants before and after pasteuriza-
magnesium concentrations of human milk within a tion. Int J Food Sci Nutr 2003;54:111–7.
24-hour period from 1 to 6 months of lactation. J Pediatr 22. Perrone L, Di Palma L, Di Toro R, Gialanella G, Moro R.
Gastroenterol Nutr 1988;7:100–6. Trace element content of human milk during lactation.
12. Neville MC, Keller RP, Seacat J, Casey CE, Allen JC, J Trace Elem Electrolytes Health Dis 1993;7:245–7.
Archer P. Studies on human lactation. I. Within-feed 23. Higashi A, Ikeda T, Uehara I, Matsuda I. Zinc and
copper contents in breast milk of Japanese women. human milk. Am J Clin Nutr 1989;49:773–85.
Tohoku J Exp Med 1982;137:41–7. 40. Dömellof M, Lönnerdal B, Dewey KG, Cohen RJ, Her-
24. Frković A, Medugorac B, Alebić-Juretić A. Zinc levels in nell O. Iron, zinc, and copper concentrations in breast
human milk and umbilical cord blood. Sci Total Environ milk are independent of maternal mineral status. Am J
1996;192:207–12. Clin Nutr 2004;79:111–5.
25. Salmenperä L, Perheentupa J, Nanto V, Siimes MA. Low 41. Donangelo CM, Trugo NM, Koury JC, Barreto Silva
zinc intake during exclusive breast-feeding does not impair MI, Freitas LA, Feldheim W, Barth C. Iron, zinc, folate
growth. J Pediatr Gastroenterol Nutr 1994;18:361–70. and vitamin B12 nutritional status and milk composi-
26. Chierici R, Saccomandi D, Vigi V. Dietary supplements tion of low-income Brazilian mothers. Eur J Clin Nutr
for the lactating mother: Influence on the trace element 1989;43:253–66.
content of milk. Acta Paediatr Suppl 1999;88:7–13. 42. Friel JK, Andrews WL, Jackson SE, Longerich HP, Mercer
27. Hess SY, King JC. Effects of maternal zinc supplementa- C, McDonald A, Dawson B, Sutradhar B. Elemental
tion on pregnancy and lactation outcomes. Food Nutr composition of human milk from mothers of premature
Bull 2009;30:S60–78. and full-term infants during the first 3 months of lacta-
28. Krebs NF, Reidinger CJ, Hartley S, Robertson AD, tion. Biol Trace Elem Res 1999;67:225–47.
Hambidge KM. Zinc supplementation during lactation: 43. Garza C, Johnson CA, Smith EO, Nichols BL. Changes in
Effects on maternal status and milk zinc concentrations. the nutrient composition of human milk during gradual
Am J Clin Nutr 1995;61:1030–6. weaning. Am J Clin Nutr 1983;37:61–5.
29. Islam MM, Khatun M, Peerson JM, Ahmed T, Mollah 44. Hibberd CM, Brooke OG, Carter ND, Haug M, Harzer
MA, Dewey KG, Brown KH. Effects of energy density G. Variation in the composition of breast milk during
and feeding frequency of complementary foods on total the first 5 weeks of lactation: Implications for the feeding
daily energy intakes and consumption of breast milk by of preterm infants. Arch Dis Child 1982;57:658–62.
healthy breastfed Bangladeshi children. Am J Clin Nutr 45. Krebs NF, Hambidge KM, Jacobs MA, Rasbach JO. The
2008;88:84–94. effects of a dietary zinc supplement during lactation on
30. Brown KH, Black RE, Becker S, Nahar S, Sawyer J. longitudinal changes in maternal zinc status and milk
Consumption of foods and nutrients by weanlings in zinc concentrations. Am J Clin Nutr 1985;41:560–70.
Bangladesh. Am J Clin Nutr 1982;36:878–89. 46. Krebs NF, Reidinger CJ, Miller LV, Hambidge KM.
31. Brown KH, Stallings RY, de Kanashiro HC, López de Zinc homeostasis in breast-fed infants. Pediatr Res
Romaña G, Black RE. Effects of common illnesses on 1996;39:661–5.
infants’ energy intakes from breast milk and other foods 47. Hambidge KM, Krebs NF, Jacobs MA, Favier A, Guyette
during longitudinal community-based studies in Huas- L, Ikle DN. Zinc nutritional status during pregnancy: A
car (Lima), Peru. Am J Clin Nutr 1990;52:1005–13. longitudinal study. Am J Clin Nutr 1983;37:429–42.
32. Kimmons JE, Dewey KG, Haque E, Chakraborty J, 48. Krebs NF, Reidinger CJ, Robertson AD, Hambidge KM.
Osendarp SJ, Brown KH. Low nutrient intakes among Growth and intakes of energy and zinc in infants fed
infants in rural Bangladesh are attributable to low intake human milk. J Pediatr 1994;124:32–9.
and micronutrient density of complementary foods. J 49. Leotsinidis M, Alexopoulos A, Kostopoulou-Farri E.
Nutr 2005;135:444–51. Toxic and essential trace elements in human milk from
33. Brown KH, López de Romaña D, Arsenault JE, Peer- Greek lactating women: Association with dietary habits
son JM, Penny ME. Comparison of the effects of zinc and other factors. Chemosphere 2005;61:238–47.
delivered in a fortified food or a liquid supplement 50. Mbofung CM, Atinmo T, Omololu A. Mineral content
on the growth, morbidity, and plasma zinc concen- of colostrum and mature milk of lactating Nigerian
trations of young Peruvian children. Am J Clin Nutr women as influenced by stage of lactation. Nutr Rep Int
2007;85:538–47. 1984;30:1137–46.
34. Bégin F, Santizo MC, Peerson JM, Torún B, Brown KH. 51. Moser PB, Reynolds RD. Dietary zinc intake and zinc
Effects of bovine serum concentrate, with or without concentrations of plasma, erythrocytes, and breast milk
supplemental micronutrients, on the growth, morbid- in antepartum and postpartum lactating and nonlac-
ity, and micronutrient status of young children in a tating women: A longitudinal study. Am J Clin Nutr
low-income, peri-urban Guatemalan community. Eur J 1983;38:101–8.
Clin Nutr 2007;62:39–50. 52. Ortega RM, Andrés P, Martinez RM, López-Sobaler AM,
35. US Institute of Medicine. Nutrition during lactation. Quintas ME. Zinc levels in maternal milk: The influence
Washington, DC: National Academy Press, 1991. of nutritional status with respect to zinc during the third
36. Atinmo T, Omololu A. Trace element content of breast- trimester of pregnancy. Eur J Clin Nutr 1997;51:253–8.
milk from mothers of preterm infants in Nigeria. Early 53. Robberecht H, Benemariya H, Deelstra H. Daily dietary
Hum Dev 1982;6:309–13. intake of copper, zinc, and selenium of exclusively
37. Butte NF, Garza C, Smith EO, Wills C, Nichols BL. breast-fed infants of middle-class women in Burundi,
Macro- and trace-mineral intakes of exclusively breast- Africa. Biol Trace Elem Res 1995;49:151–9.
fed infants. Am J Clin Nutr 1987;45:42–8. 54. Benemariya H, Robberecht H, Deelstra H. Daily
38. Casey CE, Hambidge KM, Neville MC. Studies in human dietary intake of copper, zinc and selenium by different
lactation: Zinc, copper, manganese and chromium in population groups in Burundi, Africa. Sci Total Environ
human milk in the first month of lactation. Am J Clin 1993;136:49–76.
Nutr 1985;41:1193–200. 55. Sievers E, Oldigs HD, Dörner K, Schaub J. Longitudinal
39. Casey CE, Neville MC, Hambidge KM. Studies in human zinc balances in breast-fed and formula-fed infants. Acta
lactation: Secretion of zinc, copper, and manganese in Paediatr 1992;81:1–6.
56. Silvestre D, Martìnez-Costa C, Lagarda MJ, Brines J, 74. Moser PB, Reynolds RD, Acharya S, Howard MP, Andon
Farré R, Clemente G. Copper, iron, and zinc contents in MB, Lewis SA. Copper, iron, zinc, and selenium dietary
human milk during the first three months of lactation: A intake and status of Nepalese lactating women and their
longitudinal study. Biol Trace Elem Res 2001;80:1–11. breast-fed infants. Am J Clin Nutr 1988;47:729–34.
57. Vuori E. Intake of copper, iron, manganese and zinc by 75. Moser-Veillon PB, Reynolds RD. A longitudinal study
healthy, exclusively-breast-fed infants during the first 3 of pyridoxine and zinc supplementation of lactating
months of life. Br J Nutr 1979;42:407–11. women. Am J Clin Nutr 1990;52:135–41.
58. Vuori E, Kuitunen P. The concentrations of copper and 76. Okolo SN, Onwuanaku C, Okonji M, VanderJagt DJ,
zinc in human milk. A longitudinal study. Acta Paediatr Millson M, Churchwell C, Glew RH. Concentration of
Scand 1979;68:33–7. eight trace minerals in milk and sera of mother-infant
59. Vaughan LA, Weber CW, Kemberling SR. Longitudinal pairs in northern Nigeria. J Trop Pediatr 2000;46:160–2.
changes in the mineral content of human milk. Am J 77. Onyango AW, Receveur O, Esrey SA. The contribution
Clin Nutr 1979;32:2301–6. of breast milk to toddler diets in western Kenya. Bull
60. Xiang M, Harbige LS, Zetterström R. Breast milk World Health Organ 2002;80:292–9.
levels of zinc and omega-6 polyunsaturated fatty acids 78. Perrone L, Di Palma L, Di Toro R, Gialanella G, Moro R.
and growth of healthy Chinese infants. Acta Paediatr Interaction of trace elements in a longitudinal study of
2007;96:387–90. human milk from full-term and preterm mothers. Biol
61. Yamawaki N, Yamada M, Kan-no T, Kojima T, Kaneko Trace Elem Res 1994;41:321–30.
T, Yonekubo A. Macronutrient, mineral and trace ele- 79. Rakicioğlu N, Samur G, Topçu A, Topçu A. The effect
ment composition of breast milk from Japanese women. of Ramadan on maternal nutrition and composition of
J Trace Elem Med Biol 2005;19:171–81. breast milk. Pediatr Int 2006;48:278–83.
62. Al-Awadi FM, Srikumar TS. Trace-element status in 80. Richmond J, Strehlow CD, Chalkley SR. Dietary intake
milk and plasma of Kuwaiti and non-Kuwaiti lactating of Al, Ca, Cu, Fe, Pb, and Zn in infants. Br J Biomed Sci
mothers. Nutrition 2000;16:1069–73. 1993;50:178–86.
63. Bates CJ, Tsuchiya H. Zinc in breast milk during pro- 81. Sian L, Krebs NF, Westcott JE, Fengliang L, Tong L,
longed lactation: Comparison between the UK and the Miller LV, Sonko B, Hambidge M. Zinc homeostasis
Gambia. Eur J Clin Nutr 1990;44:61–9. during lactation in a population with a low zinc intake.
64. Dewey KG, Finley DA, Lönnerdal B. Breast milk volume Am J Clin Nutr 2002;75:99–103.
and composition during late lactation (7–20 months). J 82. Simmer K, Ahmed S, Carlsson L, Thompson RP. Breast
Pediatr Gastroenterol Nutr 1984;3:713–20. milk zinc and copper concentrations in Bangladesh. Br
65. Dhonukshe-Rutten RA, Vossenaar M, West CE, Schu- J Nutr 1990;63:91–6.
mann K, Bulux J, Solomons NW. Day-to-day variations 83. Umeta M, West CE, Verhoef H, Haidar J, Hautvast JG.
in iron, zinc and copper in breast milk of Guatemalan Factors associated with stunting in infants aged 5–11
mothers. J Pediatr Gastroenterol Nutr 2005;40:128–34; months in the Dodota-Sire District, rural Ethiopia. J
discussion 0-1. Nutr 2003;133:1064–9.
66. Dorea JG, Horner MR, Bezerra VLVA, Campanate ML. 84. Vuori E, Mäkinen SM, Kara R, Kuitunen P. The effects
Longitudinal study of major milk constituents from two of the dietary intakes of copper, iron, manganese, and
different socioeconomic groups of mothers in Brazil. zinc on the trace element content of human milk. Am J
Nutr Rep Intern 1984;29:699–710. Clin Nutr 1980;33:227–31.
67. Dorea JG, Horner MR, Campanate ML. Lacteal zinc 85. Finley DA, Lonnerdal B, Dewey KG, Grivetti LE.
and copper in relation to volume, total ash and energy Inorganic constituents of breast milk from vegetar-
during the first three months of lactation of Brazilian ian and nonvegetarian women: Relationships with
women. Acta Paediatr Scand 1985;74:891–6. each other and with organic constituents. J Nutr
68. Gross R, Hänsel H, Schultink W, Shrimpton R, Matulessi 1985;115:772–81.
P, Gross G, Tagliaferri E, Sastroamdijojo S. Moderate 86. Zimmerman AW, Hambidge KM, Loepow ML, Green-
zinc and vitamin A deficiency in breast milk of mothers berg RD, Stover ML, Casey CE. Acrodermatitis in
from East-Jakarta. Eur J Clin Nutr 1998;52:884–90. breast-fed premature infants: Evidence for a defect of
69. Honda R, Tawara K, Nishijo M, Nakagawa H, Tanebe mammary zinc secretion. Pediatrics 1982;69:176–83.
K, Saito S. Cadmium exposure and trace elements in 87. Brown KH, Dewey KG, Allen LH. Complementary feed-
human breast milk. Toxicology 2003;186:255–9. ing of young children in developing countries: A review
70. Lauber E, Reinhardt M. Studies on the quality of breast of current scientific knowledge. Geneva: World Health
milk during 23 months of lactation in a rural community Organization, 1998.
of the Ivory Coast. Am J Clin Nutr 1979;32:1159–73. 88. De Groot MH. Probability and statistics, Reading, Mass,
71. Lehti KK. Iron, folic acid and zinc intakes and status of USA: Addison-Wesley, 1975.
low socio-economic pregnant and lactating Amazonian 89. Sandström B, Cederblad Å, Lönnerdal B. Zinc absorp-
women. Eur J Clin Nutr 1989;43:505–13. tion from human milk, cow’s milk, and infant formulas.
72. Lehti KK. Breast milk folic acid and zinc concentrations Am J Dis Child 1983;137:726–9.
of lactating, low socioeconomic, Amazonian women and 90. López de Romaña D, Salazar M, Hambidge KM, Penny
the effect of age and parity on the same two nutrients. ME, Peerson JM, Krebs NF, Brown KH. Longitudinal
Eur J Clin Nutr 1990;44:675–80. measurements of zinc absorption in Peruvian children
73. Melnikov P, da Cruz Montes Moura AJ, Batista Palhares consuming wheat products fortified with iron only
DB, Martimbianco de Figueiredo CS. Zinc and copper or iron and 1 of 2 amounts of zinc. Am J Clin Nutr
in colostrum. Indian Pediatr 2007;44:355–7. 2005;81:637–47.
91. Abrams SA, Wen J, Stuff JE. Absorption of calcium, zinc, 96. Walravens PA, Chakar A, Mokni R, Denise J, Lemon-
and iron from breast milk by five- to seven-month-old nier D. Zinc supplements in breastfed infants. Lancet
infants. Pediatr Res 1997;41:384–90. 1992;340:683–5.
92. World Health Organization/Food and Agriculture 97. US Institute of Medicine. Zinc. In: Dietary reference
Organization/International Atomic Energy Agency. intakes for vitamin A, vitamin K, boron, chromium,
Trace elements in human health and nutrition. Geneva: copper, iodine, iron, manganese, molybdenum, nickel,
WHO, 1996. silicon, vanadium, and zinc. Washington, DC: National
93. International Zinc Nutrition Consultative Group Academy Press, 2000:442–501.
(IZiNCG), Brown KH, Rivera JA, Bhutta Z, Gibson 98. Jalla S, Westcott J, Steirn M, Miller LV, Bell M, Krebs
RS, King JC, Lönnerdal B, Ruel MT, Sandtröm B, NF. Zinc absorption and exchangeable zinc pool sizes in
Wasantwisut E, Hotz C. Assessment of the risk of zinc breast-fed infants fed meat or cereal as first complemen-
deficiency in populations and options for its control. tary food. J Pediatr Gastroenterol Nutr 2002;34:35–41.
Food Nutr Bull 2004;25(1 suppl 2):S99–203. 99. UNICEF/World Health Organization. Global strategy
94. Krebs NF, Westcott JE, Butler-Simon N, Hambidge KM. for infant and young child feeding, Geneva: WHO, 2003.
Effects of a zinc supplement on growth of normal breast- 100. World Health Organization. Community-based
fed infants. FASEB J 1996;10:1324. strategies for breastfeeding promotion and support in
95. Heinig MJ, Brown KH, Lönnerdal B, Dewey KG. Zinc developing countries. Geneva: WHO, 2003.
supplementation does not affect growth, morbidity, or 101. Wellstart International. Community-based breast-
motor development of US term breastfed infants at 4–10 feeding support: a planning manual. Washington, DC:
mo of age. Am J Clin Nutr 2006;84:594–601. 1996.
The potential to improve zinc status through
biofortification of staple food crops with zinc
Christine Hotz
Abstract Introduction
Biofortification is an agricultural strategy that aims to Biofortification is an intervention strategy under

increase the content of select micronutrients, including development with the goal of increasing the content
zinc, in staple food crops such as rice, wheat, maize, pearl of select micronutrients, including zinc, in the edible
millet, and others. When consumed among zinc-deficient portion of staple food crops by agricultural, agronomic,
populations, zinc-biofortified staple foods should improve or genetic means. When consumed, biofortified staple
the adequacy of zinc intakes and hence reduce the risk foods would lead to improved adequacy of zinc intakes
of dietary zinc deficiency. Several conditioning factors and hence a reduced risk of dietary zinc deficiency,
will contribute to the potential for this strategy to meet among those who currently have high rates of inad-
its goal, including the additional amount of zinc that can equate intakes. At present, the potential for biofortifica-
be bred into the staple crop food, the amount of zinc that tion with zinc, iron, and provitamin A carotenoids is
remains in the staple crop food following usual processing primarily being evaluated or explored in some of the
methods, and the bioavailability of zinc from the staple world’s most important staple food crops for the poor,
crop food in the context of the usual diet. Reduction of including rice, wheat, maize, cassava, beans, and sweet
the phytate content of cereals with the use of agricultural potatoes [1].
techniques is a potential complementary strategy for Biofortification of staple food crops can be achieved
improving the bioavailability of zinc. The feasibility of through the following processes: conventional breed-
biofortification to result in a meaningful increase in the ing, by selecting for genotypes with the highest micro-
adequacy of population zinc intakes and to reduce the nutrient content observed for that crop; use of genetic
consequences of zinc deficiencies still needs to be deter- modifications, such as gene insertions or induced
mined through efficacy trials. At the program level, the mutations; and use of agronomic practices, such as
ability to widely disseminate biofortified crop varieties application of zinc-containing fertilizer.
and the willingness of farmers to adopt them will also The focus of this strategy is to increase the adequacy
affect the magnitude of the impact of this strategy. of micronutrient intakes among rural, agriculturally
based populations that already produce and consume
the staple food vehicle. As is the case for universal
Key words: Agriculture, biofortification, diet, phytate, staple food fortification, biofortification is intended
staple foods, zinc to contribute to the prevention of micronutrient defi-
ciencies, reaching all household members. Unlike tra-
ditional fortification, biofortification does not require
that the food vehicle be centrally processed. Therefore,
this strategy has the potential to fill the gap in coverage
by fortification, as it can be more accessible to those
The author is affiliated with HarvestPlus and the Interna- who consume staple foods from local or self produc-
tional Food Policy Research Institute, Washington, DC.
Please direct queries to the author: Christine Hotz, Har- tion. On the other hand, the time required to develop
vestPlus and International Food Policy Research Institute, c/o the biofortified crops and the available natural variation
the Micronutrient Initiative, 180 Elgin St., Suite 1000, Ottawa, of certain micronutrients in different crops may not
ON K2P 2K3, Canada; email: c.hotz@cgiar.org. permit as great an increase in micronutrient content
Improving zinc status through biofortification S173
as fortification may be able to achieve, and greater trials but may be expected to range from 30% to 60%
effort would be required to add MMN through biofor- in Asia (e.g., rice and wheat) and from 20% to 40% in
tification than is the case for fortification. Transgenic Africa (e.g., maize).
approaches for increasing micronutrient contents in
staple crops may enable greater enhancements in the Detailed review of the evidence
contents of several micronutrients, but there are many
regulatory and safety issues that need to be addressed To assess the feasibility of biofortification to result in
before that option can become widely available. Also, to a meaningful increase in the adequacy of population
reach a sufficient number of farmers, dissemination of zinc intakes, there are three main issues that need to be
planting materials for biofortified varieties will depend considered: What are the levels of additional absorbed
on the effectiveness of seed systems and agricultural zinc intake that need to be achieved to contribute
extension activities, which in some countries are still meaningfully to human zinc requirements? Can those
rather limited. levels be achieved through the possible biofortification
Research and development activities for zinc-biofor- processes? If minimum target levels for zinc contents
tified crops are ongoing, and conventionally bred crops of staple foods are achieved, will they reach the target
may be available for testing and release as approved population at risk?
varieties within the next 5 years.
This paper is divided into four sections, which What are the levels of additional absorbed zinc intake
address the following sets of questions in relation to that need to be achieved to contribute meaningfully to
biofortification of staple food crops with zinc: zinc requirements?
The amount of additional zinc in various food crops
Section 1: Can the achievable level of increased zinc
that will result in a biologically meaningful improve-
content contribute to meeting zinc requirements,
ment in population zinc status needs to be established
taking into account bioavailability, in a large proportion
through research. The theoretical optimal amount
of the vulnerable population?
of additional zinc to be added through biofortifica-
Section 2: What are the major factors potentially tion would be calculated to cover the deficit between
modifying the impact of zinc-biofortified staple food actual zinc intakes and the intake amount that would
crops? minimize the proportion of the population with inad-
equate zinc intakes. This is the approach that has been
Section 3: Can biofortified staple foods improve popu-
recommended for the design of universal food forti-
lation zinc status and associated health outcomes?
fication programs [2]. However, because biofortified
Section 4: What are the risks of consuming zinc- staple foods are still being developed and zinc content
biofortified foods? is being increased incrementally over time, there is
presently less flexibility in catering to specific needs of
populations than there may be with commercial food
Section 1 fortification. For this reason, minimum target levels
have been set as preliminary goals for the first phase of
Can the achievable level of increased zinc content development of biofortified staple foods. The proposed
contribute to meeting zinc requirements, taking into minimum target level for zinc has been set to provide
account bioavailability, in a large proportion of the an additional amount of bioavailable zinc in the food
vulnerable population? supply that is equivalent to ~40% of the physiological
requirement for absorbed zinc for nonpregnant women
Conclusions and nonbreastfed children 4 to 6 years of age [3, 4]
among populations with high intakes of the staple food
Achievements in conventional breeding for zinc-rich of interest.
staple food crops suggest that there is adequate breed- The latter target increment makes assumptions about
ing potential for several crops, including rice, maize, the usual amount of staple food intake per day; the loss
wheat, and pearl millet. If assumptions are met for daily of zinc from the seed, root, or tuber during processing
intake of the staple food, bioavailability of zinc from (e.g., milling) and cooking; and the bioavailability of
the diet, and amount of zinc lost during processing zinc from the staple food in the context of the usual
and cooking, the additional amount of zinc achieved diet [3]. The specific assumptions used in setting these
through biofortification in these crops would contribute minimum, generic breeding targets for zinc contents in
approximately 40% of the absorbed zinc requirement various crops are summarized in table 1.
for nonpregnant women and for children 4 to 6 years For the appropriate development of biofortified
of age. Adoption rates or “coverage” of zinc-biofortified staple foods, each of these assumptions will need to
crops and their consumption by different age groups be verified and adjusted accordingly in populations
will need to be assessed in large-scale implementation targeted for biofortification. Dietary intake data from
S174 C. Hotz
a range of rural, low-income populations at risk for content and bioavailability are discussed in Section 2.
dietary zinc deficiency are lacking, so quantification Another approach to assessing the adequacy of dif-
of usual, individual intakes of zinc and of the targeted ferent increments in the zinc contents of staple foods is
staple foods is required. The losses of zinc with cooking through modeling with dietary intake data. One such
of staple foods is typically very low (e.g., 0% to 10%; analysis was conducted with the use of dietary intake
[5]), but losses with milling of grains at different extrac- data from the National Nutrition Survey of Mexico
tion rates can be high (table 2) and must be quantified. of 1999 [7] and proposed zinc requirements [4]. The
Losses of phytate should be measured together with analysis examined the effect of increasing the content
zinc losses, since both the total zinc content and the of zinc in maize from 18 to 33 µg/g dry weight on the
phytate:zinc molar ratio determine the amount of adequacy of bioavailable zinc intakes. Bioavailability
absorbable zinc [4]. The assumed 25% bioavailability was estimated with the use of a prediction equation
of zinc is only an average figure, and since the target [4], and the zinc content and phytate:zinc molar ratio
levels, and presumably the biological impact of dietary of the diets and estimated absorbable zinc were com-
zinc interventions, will depend on the net amount of pared with the age-specific physiological requirement
zinc absorbed, zinc absorption from biofortified crops for absorbed zinc to determine adequacy. At the level
should be measured directly. Zinc bioavailability is of maize consumption in the population, the model
particularly important to confirm for children, since predicted a decrease in the prevalence of inadequate
the absorption of zinc from different staple food-based zinc intakes from 28% to 15% among rural preschool
diets has not frequently been quantified in children of children and from 43% to 19% among rural women.
different ages and it cannot be directly predicted from With the use of this approach, appropriate target
the equations derived from studies in adults [4, 6]. The levels for a specific population could be determined
potential modifying effects of milling in relation to zinc by assessing the simulated impact of different levels of
TABLE 1. Assumptions used to set minimum target levels for increased zinc content of biofortified staple food cropsa
Bioavail-
Estimated Additional able zinc (as
Average Staple food zinc content percentage of
Requirement intake Average zinc content to be achieved physiological
Staple foodb Group for zincc (g/day) of crop (mg/kg) (mg/kg) requirementd)
Whole maize Nonbreastfed chil- 2 100 Whole maize: ~30 +8 34
Whole wheat dren 1–3 yr Whole wheat: ~30
Polished rice Children 4–6 yr 4 200 Polished rice: ~16 43
Cassava Nonpregnant 7 400 Cassava: ~10 39
women ≥ 19 yr
Sweet potato Nonbreastfed chil- 2 50 Sweet potato: ~9 + 17 36
Common dren 1–3 yr Common beans: ~32
beans Children 4–6 yr 4 100 46
Nonpregnant 7 200 41
women ≥ 19 yr
a. For all staple crops, zinc losses during processing were assumed to be 10% and zinc bioavailability was assumed to be 25%.
b. For maize, rice, wheat, and beans, intakes are expressed on a dry weight basis; for cassava and sweet potato, intakes are expressed on a fresh
weight basis.
c. The Estimated Average Requirements for unrefined cereal-based diets or refined vegetarian or mixed diets recommended by the Interna-
tional Zinc Nutrition Consultative Group (IZiNCG) [4] are given.
d. The physiological requirements for absorbed zinc presented by IZiNCG [4] were used.
TABLE 2. Zinc contents of whole cereal grains and milled products

Zinc content Zinc content Zinc content
Cereal (µg/g dry weight) Cereal (µg/g dry weight) Cereal (µg/g dry weight)
Maize Wheat Rice
Whole 21.0 Whole 37 Brown 28
High extraction 17.3 85% extraction 19 Polished 17
Low extraction 8.0 70% extraction 12
Degermed 4.4
Sources: Welch and Graham [32], Ferguson et al. [33], US Department of Agriculture [34], and International Minilist/WorldFood Dietary
Assessment System, 2.0 (University of California, Berkeley, CA, USA).
additional zinc on the prevalence of inadequate intakes and barley were found to increase by 39% to 77% with
of bioavailable zinc. soil zinc application in either rain-fed or irrigated
fields [17], and the concentration of zinc in rice pad-
Can those levels be achieved through the possible dies increased up to 2.3-fold with soil zinc application
biofortification processes? in Pakistan [18]. In Bangladesh, foliar fertilization
At present, it is estimated that the minimum target level with zinc resulted in grain zinc content that was 2.6
for zinc in adapted varieties of rice, maize, wheat, and times greater in rice and 2 times greater in wheat than
pearl millet is achievable through conventional breed- in controls [19]. Further, these increments were esti-
ing within the next 2 to 5 years [8]. For other crops, mated to result in a doubling of the total zinc intake of
such as yams and cassava, sufficiently high levels of Bangladeshi children. Assuming a 43% increase in the
zinc content may not be achievable through conven- zinc content of milled rice due to soil zinc application,
tional breeding in a measurable time frame because of Gibson and colleagues recently estimated that the total
limited natural genetic variation in zinc content. For zinc intake of Thai schoolchildren at high risk for zinc
some crops, such as beans and potatoes, adequately deficiency would be nearly doubled from 4.8 to 8.6 mg/
increased zinc content may be achievable, but the time day with treatment, on the basis of their usual intakes
frame for doing so cannot yet be estimated. of rice and zinc [20]. However, variation in grain zinc
Although proof-of-concept has been achieved for concentration in response to zinc fertilizer applica-
genetic modification to synthesize β-carotene and tion will occur with different crops, genotypes of the
increase iron content through synthesis of ferritin same crop, and environmental conditions [16], so the
protein, the use of biotechnology to increase the zinc benefits of these techniques for human zinc nutrition
content of staple foods is still at an exploratory stage cannot be extrapolated to all situations.
[9]. Research is still required to identify genes involved
with the plants’ uptake and translocation of zinc and its If minimum target levels for zinc contents of staple
deposition into the edible portion of the grain, root, or foods are achieved, will they reach the target population
tuber. Information is accumulating on zinc transport- at risk?
ers and how they function in different plant tissues To reach large segments of populations vulnerable to
throughout their development [10, 11]. zinc deficiency, biofortified staple food crops must be
Some progress has been made with the use of bio- adopted by farmers for regular production and by con-
technology that may affect zinc nutrition indirectly by sumers for regular consumption. This requires several
improving its bioavailability. Staple food crops, such as conditioning factors. First, the strategy must tap into
maize, have been modified by mutagenesis to achieve a available mechanisms for disseminating new seed or
reduced grain content of phytic acid [12]. Progress has planting materials to farmers, such as through public
also been made with the use of transgenic approaches extension systems, private seed companies, or non-
to express heat-stable phytase from microbial sources profit or nongovernmental organizations involved in
in the endosperm of wheat [13]. Although plenty of rural extension. Adoption of the seed will require that
evidence exists to indicate that reduced phytate intakes the biofortified varieties have acceptable traits, such
relative to zinc intakes result in increased efficiency of as adequate (if not improved) yield, pest and disease
absorption of zinc [4, 14], Mazariegos and colleagues resistance, and preferred physical and organoleptic
[15] recently concluded that zinc absorption was not properties. If these latter conditions are met, adoption
significantly greater among children consuming a diet of zinc-biofortified staple foods should not require a
based on a low-phytate maize mutant (~60% reduction) strong behavior change component. In general, it is
than among children consuming an isogenic wild- reasonable to expect that these latter conditions are or
type maize, possibly because of greater than expected will be met for zinc-biofortified crops [21].
variation in total phytate intakes, inadequate statistical Adoption rates for new crop varieties can vary widely
power, and differences in the total zinc content of the by region; for example, conservative estimates of adop-
diet, which independently affects the fractional absorption rates of 30% to 60% have been assumed for Asia,
tion of dietary zinc. There is as yet no clear evidence where seed systems are better developed, and 20% to
available from controlled trials for the benefits to zinc 40% for Africa, where seed systems are less well devel-
status from long-term intake of diets in which only the oped [22]. Where adoption rates are typically lower,
phytate content is reduced. additional investment in interventions through nutri-
Zinc-containing fertilizers for wheat production tion education, farmer education, and seed systems
in areas with zinc-deficient soils, such as Turkey, are support may be required to assure adequate coverage
already in use as a means of improving crop yields in a reasonable length of time. Coverage for zinc-bio-
[16]. The application of zinc in soil fertilizers or foliar fortified staple food crops will need to be determined
sprays can, in some situations, also result in increased through the evaluation of large-scale interventions to
zinc content of the grain. For example, the grain zinc determine the percentage of the population that pro-
concentrations of bread wheat, durum wheat, triticale, duces or purchases biofortified foods, the percentage
S176 C. Hotz
of production represented by the biofortified variety, portions of the grain lost in milling, the benefits of the
and, ultimately, the amount of biofortified varieties strategy may be more limited for those populations
consumed by the target population that is at risk for that consume milled grains. Efforts of breeders and
zinc deficiency. A pilot intervention to introduce plant scientists need to focus on increasing micro-
β-carotene–rich sweet potato in Mozambique as a nutrient content of the starchy portion of the grain
biofortified staple crop included an assessment cap- (endosperm) to have maximum benefit for the widest
turing many of these elements [23]. Similar research range of populations.
on a larger scale will need to be carried out for zinc- On the other hand, milled grain products will
biofortified crops in a variety of settings; the potential have a lower content of phytate, the primary plant
coverage and impact may only be known in the next component that inhibits zinc absorption in humans.
5 to 10 years. Therefore, although a lower amount of additional zinc
may be achievable in the starchy endosperm, what is
there will be more bioavailable. For example, using a
Section 2 recently published trivariate model for estimating zinc
absorption from diets with varying levels of zinc and
What are the major factors potentially modifying the phytate [14], Hambidge and colleagues [6] predicted
impact of zinc-biofortified staple food crops? that the same amount of additional zinc would be
absorbed from biofortified wheat flour whether it was
Conclusions whole or of 85% extraction, based on biofortified and
control wheat produced in Mexico by the Center for
The effects of milling degree on the potential useful- the Improvement of Maize and Wheat (CIMMyT).
ness of zinc biofortification in grains require further A human study of zinc absorption using these wheat
study. Quantification of the zinc and phytate contents products is in progress to verify the predictions.
of whole and refined biofortified grain products and Strategies for reducing the phytate content of zinc-
the absorption of zinc from them should be measured biofortified cereal grains may help to increase their
in human studies. It is possible that loss of zinc with impact on zinc intake adequacy. Induced mutations
milling may be offset by the parallel loss of phytate in several crops, including maize, wheat, and barley,
and a commensurate increase in zinc bioavailability. have led to varying degrees of decrease in grain phytate
Nonetheless, maximum benefits of biofortification may content [24–26]. Although an increase in zinc absorp-
be observed if more of the additional grain zinc can be tion with decreased phytate content of maize from such
deposited into the starchy endosperm. Another pos- mutants was well demonstrated among US adults [27],
sible strategy for positively affecting the impact of zinc the increase in the amount of absorbed zinc following
biofortification is a concomitant decrease in phytate the longer-term consumption of the reduced-phytate
content. Some experience with low-phytate mutant maize mutants consumed with typical diets was not
cereal crops suggests that it is possible to decrease significantly greater than that for diets with the wild-
grain phytate content substantially, but the agronomic type control maize or a local maize variety with a
viability of these mutants and the impact on zinc status naturally higher zinc content [15]. The long-term
after long-term consumption still require assessment. effects of reduced phytate intake on zinc absorption
The potential impact of host or environmental factors and excretion should be reassessed in further studies.
affecting intestinal health and either zinc absorption Some of the low-phytate mutant crops were shown
or intestinal losses of endogenous zinc needs to be to have reduced yields [24], and this would need to
quantified. If these conditions are highly prevalent be overcome in order for these mutants to be further
in populations that could benefit from biofortifica- tested and released as varieties.
tion, the additional amount of zinc achieved through Other factors that may alter the efficacy of zinc-
biofortification may need to be higher, or the impact biofortified wheat include any of those that limit the
of zinc-biofortified foods may be conditional on their ability of individuals to maintain zinc homeostasis [4].
treatment or prevention. For example, potentially common conditions, such
as frequent diarrhea, intestinal parasitic infections,
Detailed review of the evidence or tropical enteropathy characterized by increased
intestinal permeability [28, 29], could lead to increased
In cereal grains, rather large concentrations of zinc typ- intestinal losses of endogenous zinc [30] and hence
ically occur in the outer layers (e.g., aleurone) and the increased requirements for dietary zinc. Such factors
germ. As a result, large proportions of the total grain could therefore limit the benefit of any food-based
zinc content are lost during milling; some examples of intervention, including zinc fortification. It is uncertain
the content of zinc in grain and their milling products to what extent poor intestinal health caused by any of
are shown in table 2. If the process of biofortification those conditions may affect zinc absorption and zinc
results primarily in increased amounts of zinc in the homeostasis [4]; research in this area is needed.
Section 3 Detailed review of the evidence
The increased content of endogenous zinc in staple

Can biofortified staple foods improve population zinc foods is not expected to cause any changes to the color,
status and associated health outcomes? texture, or organoleptic qualities of these foods, and
therefore the additional zinc may be considered as
Conclusions “invisible” to the population [21]. This is in contrast
to biofortification with provitamin A, which causes
There is no direct evidence yet available to demon- distinct changes in color and other organoleptic prop-
strate whether or not zinc-biofortified staple foods can erties. Nonetheless, high-zinc lines in many cases will
have an impact on zinc status and associated health need to be backcrossed into local or improved varieties
outcomes. in order to maintain all of the locally preferred agro-
nomic and organoleptic qualities of the crop [31]. Ulti-
Detailed review of the evidence mately, the preferences of farmers and consumers need
to be addressed when new varieties are to be released.
The availability of zinc-biofortified staple crops is still As described above in relation to assessing the appro-
limited, and thus randomized, controlled efficacy trials priateness of different target increments in the zinc
of their impact on zinc status and associated health content of staple foods, simulations based on existing
outcomes have not yet been conducted [3]. The first dietary intake data can be used to estimate the impact
of such trials are expected to be conducted within the of biofortification on the prevalence of intakes that
next 3 to 5 years. might exceed the upper limit for zinc. The simulation
model using data from the National Nutrition Survey
of Mexico of 1999 [7] predicted that only 3% and 0%
Section 4 of preschool children and women, respectively, would
have total zinc intakes exceeding the proposed upper
What are the risks of consuming zinc-biofortified foods? limit for zinc [4] after the biofortification of maize.
Nonetheless, to fully assess the prevalence of high zinc
Conclusions intakes in a population, it also would be necessary to
conduct similar modeling for other subgroups, such as
At the modest, physiological levels of increased zinc adolescent boys and adult men, because their intakes
content presently achievable through biofortification, of staple foods are usually higher than those of women
it is unlikely that there is any substantial risk of toxic and preschool children, and hence they are more likely
intakes of zinc as a result of biofortification. It is also to achieve levels of zinc intake that could exceed the
unlikely that an increased zinc content of staple foods upper limits. Ultimately, direct studies of the potential
would result in changes to their organoleptic qualities toxicity of zinc biofortification, such as through assess-
that could lead to rejection of biofortified varieties that ment of the effect of excess zinc on copper status, would
might be introduced. be necessary to confirm the true risk of toxicity; the
upper limits for zinc intakes were derived from stud-
ies of zinc intakes from supplements and may not be
relevant to intakes of zinc from food sources [4].
References
1. HarvestPlus. Research and implementation. Available B, Wasantwisut E, Hotz C. Assessment of the risk of zinc
at: http://www.harvestplus.org/pdfs/hpRIbrochure.pdf. deficiency in populations and options for its control.
Accessed 10 December 2008. Food Nutr Bull 2004;25(1 suppl 2):S99–203.
2. Allen L, de Benoist B, Dary O, Hurrell R, eds. Guide- 5. US Department of Agriculture. Table of nutrient reten-
lines on food fortification with micronutrients. Geneva: tion factors. Release 6. Washington, DC: US Department
World Health Organization, 2006. of Agriculture, 2007.
3. Hotz C, McClafferty B. From harvest to health: Chal- 6. Hambidge K, Krebs N, Miller L, Westcott J, Rosado
lenges for developing biofortified staple foods and J, Garcia O, Gonzalez K, Ortiz-Monasterio I, Pfeiffer
determining their impact on micronutrient status. Food W, Hotz C. Zinc (Zn) in biofortified crops: Efficacy
Nutr Bull 2007;28:S271–9. in human nutrition. Zinc crops 2007: Improving
4. International Zinc Nutrition Consultative Group crop production and human health. 24–26 May 2007,
(IZiNCG), Brown KH, Rivera JA, Bhutta Z, Istanbul, Turkey. Conference proceedings. Avail-
Gibson RS, King JC, Lönnerdal B, Ruel MT, Sandtröm able at: http://www.zinc-crops.org/ZnCrops2007/
S178 C. Hotz
PDF/2007_zinccrops2007_hambidge.pdf. Accessed 5 using zinc fertilizers to improve the adequacy of zinc

February 2009. intakes of children consuming rice-based diets in NE
7. De Nova E. Modelaje del impacto de un programa de Thailand. Zinc crops 2007: Improving crop production
biofortificación en la adecuación de la ingesta de zinc and human health. 24–26 May 2007, Istanbul, Turkey.
en mujeres y preescolares en México. Tesis publicada, Conference proceedings. http://www.zinc-crops.org/
Cuernavaca, Mexico: Instituto Nacional de Salud Pub- ZnCrops2007/PDF/2007_zinccrops2007_gibson.pdf.
lica/Escuela de Salud Publica de Mexico, 2006. Accessed 5 February 2009.
8. HarvestPlus. Medium term plan: 2008–2010. Washing- 21. Bouis HE. Micronutrient fortification of plants through
ton, DC: International Food Policy Research Institute / plant breeding: Can it improve nutrition in man at low
Centro Internacional de Agricultura Tropical, 2007. cost? Proc Nutr Soc 2003;62:403–11.
9. Bouis HE. The potential of genetically modified food 22. Meenakshi J, Johnson N, Manyong V, De Groote H,
crops to improve human nutrition in developing coun- Javelosa J, Yanggen D, Naher F, Gonzalez C, Garcia J,
tries. J Dev Studies 2007;43:79–96. Meng E. How cost-effective is biofortification in com-
10. Guerinot ML. The ZIP family of metal transporters. bating micronutrient malnutrition? An ex-ante assess-
Biochim Biophys Acta 2000;1465:190–8. ment. HarvestPlus Working Paper 2. Washington DC:
11. Grotz N, Guerinot ML. Molecular aspects of Cu, Fe HarvestPlus, 2007.
and Zn homeostasis in plants. Biochim Biophys Acta 23. Low JW, Arimond M, Osman N, Cunguara B, Zano F,
2006;1763:595–608. Tschirley D. A food-based approach introducing orange-
12. Raboy V, Gerbasi P, Young K, Stoneberg S, Pickett S, fleshed sweet potatoes increased vitamin A intake and
Bauman A, Murthy P, Sheridan W, Ertl D. Origin and serum retinol concentrations in young children in rural
seed phenotype of maize low phytic acid 1-1 and low Mozambique. J Nutr 2007;137:1320–7.
phytic acid 2-1. Plant Physiol 2000;124:355–68. 24. Bregitzer P, Raboy V. Effects of four independent low
13. Holm PB, Kristiansen KN, Pedersen HB. Transgenic phytate mutations on barley agronomic performance.
approaches in commonly consumed cereals to improve Crop Sci 2006;46:1318–22.
iron and zinc content and bioavailability. J Nutr 25. Guttieri M, Bowen D, Dorsch JA, Raboy V, Souza E.
2002;132:514S–6S. Identification and characterization of a low phytic acid
14. Miller LV, Krebs NF, Hambidge KM. A mathematical wheat. Crop Sci 2004;44:418–24.
model of zinc absorption in humans as a function of 26. Pilu R, Panzeri D, Gavazzi G, Rasmussen SK, Consonni
dietary zinc and phytate. J Nutr 2007;137:135–41. G, Nielsen E. Phenotypic, genetic and molecular char-
15. Mazariegos M, Hambidge KM, Krebs NF, Westcott JE, acterization of a maize low phytic acid mutant (lpa241).
Lei S, Grunwald GK, Campos R, Barahona B, Raboy V, Theor Appl Genet 2003;107:980–7.
Solomons NW. Zinc absorption in Guatemalan school- 27. Hambidge KM, Huffer JW, Raboy V, Grunwald GK,
children fed normal or low-phytate maize. Am J Clin Westcott JL, Sian L, Miller LV, Dorsch JA, Krebs NF. Zinc
Nutr 2006;83:59–64. absorption from low-phytate hybrids of maize and their
16. Cakmak I. Plant nutrition research: Priorities to meet wild-type isohybrids. Am J Clin Nutr 2004;79:1053–9.
human needs for food in sustainable ways. Plant Soil 28. Menzies IS, Zuckerman MJ, Nukajam WS, Somasunda-
2002;247:3–24. ram SG, Murphy B, Jenkins AP, Crane RS, Gregory GG.
17. Bagci A, Erdal I, Gultekin F, Yilmaz A, Ekiz H, Sade B, Geography of intestinal permeability and absorption.
Torun MB, Cakmak I. Effect of zinc fertilization and Gut 1999;44:483–9.
irrigation on grain yield, zinc concentration and quality 29. Goto K, Chew F, Torun B, Peerson JM, Brown KH.
of cereal species. Zinc crops 2007: Improving crop pro- Epidemiology of altered intestinal permeability to lac-
duction and human health. 24–26 May 2007, Istanbul, tulose and mannitol in Guatemalan infants. J Pediatr
Turkey. Conference proceedings. Available at: http:// Gastroenterol Nutr 1999;28:282–90.
www.zinc-crops.org/ZnCrops2007/PDF/2007_zinc- 30. Castillo-Duran C, Vial P, Uauy R. Trace mineral bal-
crops2007_bagci.pdf. Accessed 5 February 2009. ance during acute diarrhea in infants. J Pediatr 1988;
18. Umar Khan M, Qasim M, Jamil M. Effect of different 113:452–7.
levels of zinc on the extractable zinc content of soil 31. Pfeiffer W, McClafferty B. Biofortification: Breeding
and chemical composition of rice. Asian J Plant Sci micronutrient-dense crops In: Kang MPP, Priyadarshan
2002;1:20–1. PM, eds. Breeding major food staples. Ames, Ia, USA:
19. Duxbury J, Bodruzzaman M, Jonson S, Mayer A, Lauren Blackwell Publishing, 2007. pp 61–91.
J, Meisner C, Welch R. Impacts of increased mineral 32. Welch R, Graham R. A new paradigm for world agri-
micronutrient content of rice and wheat seed/grain on culture: Meeting human needs. Productive, sustainable,
crop productivity and human nutrition in Bangladesh. nutritious. Field Crops Res 1999;60:1–10.
In: Li C, Zhang F, Dobermann A, Hinsinger P, Lambers 33. Ferguson E, Gibson R, Thompson L, Ounpuu S, Berry
H, Li XL, Marschner P, Maene L, McGrath S, Oenema M. Phytate, zinc, and calcium contents of 30 East Afri-
O, Peng SB, Rengel Z, Shen QR, Welch R, Von Wiren can foods and their calculated phytate:Zn, Ca:phytate
N, Yan XL, Zhu YG,ed. Plant nutrition for food security, and [Ca][Phytate]/[Zn] molar ratios. J Food Comp Anal
human health and environmental protection. Beijing: 1988;1:316–25.
Tsinghua University Press, 2005:30–1. 34. US Department of Agriculture. Nutrient database for
20. Gibson R, Winichagoon P, Pongcharoen T, Thurlow R, standard reference. Release 14. Washington DC: US
Krittaphol W, Bailey K, Anderson V. The feasibility of Department of Agriculture, 2001.
Galvanizing action: Conclusions and next steps for
mainstreaming zinc interventions in public health
programs
Kenneth H. Brown, Shawn K. Baker, and the IZiNCG Steering Committee*
Abstract absorbed zinc, and recent studies are beginning to con-

firm a positive impact of zinc fortification on indicators
This paper summarizes the results of the foregoing reviews of population zinc status. To assist with the development
of the impact of different intervention strategies designed of zinc intervention programs, more information is
to enhance zinc nutrition, including supplementation, needed on the prevalence of zinc deficiency in different
fortification, and dietary diversification or modification. countries, and rigorous evaluations of the effectiveness
Current evidence indicates a beneficial impact of such of large-scale zinc intervention programs should be
interventions on zinc status and zinc-related functional planned. Recommended steps for scaling up zinc inter-
outcomes. Preventive zinc supplementation reduces the vention programs, with or without other micronutrients,
incidence of diarrhea and acute lower respiratory tract are described. In summary, there is now clear evidence
infection among young children, decreases mortality of of the benefit of selected interventions to reduce the risk
children over 12 months of age, and increases growth of zinc deficiency, and a global commitment is urgently
velocity. Therapeutic zinc supplementation during epi- needed to conduct systematic assessments of population
sodes of diarrhea reduces the duration and severity of zinc status and to develop interventions to control zinc
illness. Zinc fortification increases zinc intake and total deficiency in the context of existing public health and
nutrition programs.
Kenneth H. Brown is affiliated with the Department of

Nutrition and the Program in International and Community Key words: Dietary diversification, dietary modifica-
Nutrition, University of California, Davis, California, USA;
Kenneth H. Brown and Shawn K. Baker are affiliated with tion, research needs, zinc deficiency, zinc fortification,
Helen Keller International, Dakar, Senegal. zinc intervention, zinc supplementation
H. Brown, Department of Nutrition, University of California,
One Shields Ave., Davis, CA 95616, USA; e-mail: khbrown@
ucdavis.edu.
Background and summary of the evidence
* The other members of the Steering Committee of the This second technical document prepared by the Inter-
International Zinc Nutrition Consultative Group (IZiNCG)
are Zulfiqar A. Bhutta (Department of Paediatrics and Child national Zinc Nutrition Consultative Group (IZiNCG)
Health, the Aga Khan University, Karachi, Pakistan), Omar provides information on the range of intervention
Dary (A2Z/The USAID Micronutrient and Child Blindness strategies that can be considered for control of zinc
Project, Washington, DC, USA), Rosalind S. Gibson (Depart- deficiency and systematic reviews of current knowl-
ment of Human Nutrition, University of Otago, Dunedin,
New Zealand), Christine Hotz (HarvestPlus and International edge regarding their efficacy and effectiveness. The
Food Policy Research Institute, Washington, DC, USA), Janet results of these analyses confirm a beneficial impact
C. King (Children’s Hospital Oakland Research Institute, Oak- of preventive zinc supplementation for reducing the
land, California, USA), Bo Lönnerdal (Department of Nutri- incidence of selected childhood infections and increas-
tion and Program in International and Community Nutrition, ing children’s physical growth [1], and of therapeutic
University of California, Davis, California, USA), Juan A.
Rivera (Nutrition and Health, National Institute of Public zinc supplementation for reducing the duration and
Health, Cuernavaca, Mexico), Marie T. Ruel (Food Consump- severity of diarrhea [2]. In particular, preventive zinc
tion and Nutrition, International Food Policy Research Insti- supplementation reduces the incidence of diarrhea
tute, Washington, DC, USA), Emorn Wasantwisut (Institute of by approximately 27% among young children over 12
Nutrition, Mahidol University, Salaya, Thailand), and Sonja Y.
Hess (IZiNCG Executive Officer; Department of Nutrition, months of age and decreases the incidence of acute
and the Program in International and Community Nutrition, lower respiratory tract infections by approximately
University of California, Davis, California, USA). 15%. Preventive zinc supplementation may also reduce
the incidence of malaria, but the number of available on such outcomes.

studies is still relatively small, so more research is Less information is available on the impact of zinc
needed to confirm this outcome. Overall, zinc sup- fortification programs and of dietary intervention
plementation reduces child mortality by approximately strategies, although the available evidence suggests that
6%. This impact of preventive zinc supplementation is both of these approaches should enhance zinc status.
restricted to children over 12 months of age (in whom For example, studies indicate that zinc fortification
the mortality reduction is approximately 18%) and can increase zinc intake and total zinc absorption [9].
possibly to small-for-gestational-age infants. Preventive However, data are still lacking to demonstrate a similar
zinc supplementation also increases linear growth and positive impact of zinc fortification on young children’s
weight gain of young children, thereby contributing zinc status, and additional information is needed on
to reduced rates of stunting and underweight. Impor- the effectiveness of large-scale fortification programs,
tantly, available studies show that preventive zinc sup- among both children and adults. Information is also
plements provided in recommended amounts do not needed on the impact of point-of-use multiple micro-
have adverse effects on the status of other micronutri- nutrient (MMN) fortificants on zinc status and other
ents or cause any detectable functional abnormalities. zinc-related health outcomes.
Thus, preventive zinc supplementation is both safe and A number of dietary intervention strategies have
efficacious, contributing positive benefits for achieving the potential for improving zinc status. Breastmilk is
the United Nations Millennium Development Goals an important potential source of zinc for infants and
number 1 (reduction of poverty and hunger, as meas- young children, and current international guidelines
ured by rates of underweight and stunting) and number for the promotion of breastfeeding should be viewed
4 (reduction of child mortality). as an appropriate component of programs to support
The present analysis confirms that provision of ther- adequate zinc nutrition of young children [10]. Other
apeutic zinc supplementation as adjunctive treatment interventions to increase the availability, accessibility,
reduces the duration of acute diarrhea by 0.5 days and and consumption of animal-source foods or to increase
that of persistent diarrhea by 0.7 days [2]. The results the zinc content of plant-source foods or to increase
therefore support the recommendation by the World zinc absorption from these foods should all enhance
Health Organization (WHO) and the United Nations the consumers’ zinc status [11]. However, rigorous
Children’s Fund (UNICEF) to include therapeutic evaluations of large-scale dietary approaches are still
zinc supplementation in diarrhea control programs lacking. In the future, biofortification holds promise as
[3]. Insufficient information is available to determine a sustainable approach to improve the zinc content and/
whether zinc supplementation might also enhance the or bioavailability of staple food crops [12].
treatment of acute and chronic respiratory infections, To summarize, there is good evidence supporting
and available studies found no impact of therapeutic the beneficial impact of zinc interventions, especially
zinc on the outcome of malaria treatment [2]. of zinc supplementation and, to a lesser extent, zinc
Preventive zinc supplementation has also been exam- fortification, on zinc status and zinc-related functional
ined among pregnant and lactating women [4]. An outcomes. Moreover, recent reviews of major strategies
earlier meta-analysis concluded that zinc supplemen- to reduce childhood morbidity and mortality have
tation during pregnancy reduced the rate of preterm highlighted the importance of controlling zinc defi-
births by 14% [5]. However, there were no consistent ciency. For example, the Lancet series on maternal and
effects on complications of labor and delivery or on child undernutrition concluded that zinc deficiency
infant birthweight. Reports from studies in Peru and is responsible for approximately 4% of the worldwide
Bangladesh noted reduced rates of postnatal infant morbidity and mortality burdens of young children
infections [6, 7], and the one study that examined [13]. Similarly, the recently revised Copenhagen Con-
infant growth throughout the first year of life found sensus on the best ways of advancing global welfare
increased growth beginning at 4 months and continu- cited zinc supplementation as one of the top-ranked
ing through 12 months of age among infants whose interventions for reducing malnutrition and improving
mothers received supplemental zinc during pregnancy welfare overall [14].
[8]. However, there is still too little information to allow On the basis of these sets of information, it is reason-
definitive conclusions to be drawn on these postnatal able to ask why national governments and international
infant outcomes. Available information suggests that agencies are not investing more in programs to reduce
maternal zinc supplementation during lactation does zinc deficiency. Although there are multiple possible
not affect breastmilk zinc concentration [4], although answers to this question, the most likely explanations
most of these studies were conducted among presum- seem to be the lack of comprehensive information on
ably well-nourished women in industrialized countries, the prevalence of zinc deficiency, the limited program-
so more information is needed from undernourished matic experience in the delivery of zinc interventions,
women. Studies among lactating women have not the lack of guidelines from international agencies on
reported on maternal health, so information is needed the need to control zinc deficiency, and possibly the
Conclusions and mainstreaming zinc interventions S181
inherent inertia of under-resourced health systems and survival issues. These opportunities should be seized
donor agencies to implement new programs, especially to sensitize policy makers and program managers to
when total funding for nutrition programs is relatively the importance of mainstreaming zinc into child health
stagnant. and nutrition interventions. IZiNCG technical docu-
To address these formidable challenges, efforts are ments and a one-page summary of major advocacy
urgently needed to include assessment of zinc status points concerning zinc nutrition are available on the
in national nutritional status assessment surveys, to internet (www.IZiNCG.org) to assist with these com-
promote increased understanding of the importance munication tasks.
of zinc deficiency and to motivate greater interest in To motivate further discussions and actions at the
controlling this problem, and to conduct operational national level, it would be helpful to enlist existing
research to determine how best to deliver zinc-international nutrition stakeholders’ groups or to help
vention programs in the context of ongoing nutrition catalyze formation of such groups where they do not
and health activities. The following paragraphs summa- exist. Examples of important partners to include in
rize some of the major issues that need to be addressed such national stakeholders’ groups are representatives
to permit scaling up and mainstreaming of zinc inter- of the public health community, consumer groups and
ventions. Because information on the implementation other civil society organizations, industry, and donor
and impact of these programs is still rudimentary, it agencies, as well as individual politicians, academicians,
will be important to incorporate into the design and and journalists.
implementation of these programs rigorous monitor-
ing and evaluation systems so that they can be critically
assessed and modified as necessary. Implementing and scaling up preventive
zinc supplementation
Collecting information on population zinc Given the growing evidence and heightened awareness
status concerning the positive impacts of preventive zinc
supplementation in children, there is an urgent need
Objective information on population zinc status is to address specific operational issues that are currently
urgently needed for countries that have been identified impeding programmatic progress. The present review
as having an elevated risk of zinc deficiency. Recom- identified several key operational issues [1], including
mendations are now available on the best approaches the optimal dosage range for preventive zinc sup-
for assessing population zinc status [15]. To minimize plementation, with or without other micronutrients,
the cost of collecting information on population zinc for particular target groups within the population; the
status, opportunities should be sought to include appropriate frequency and duration of supplementa-
zinc-related information in already scheduled health tion; and the types of products that can be used to
or nutritional status surveys to be conducted in repre- deliver zinc, such as dispersible tablets, multimicro-
sentative samples of national populations. This infor- nutrient powders (e.g., products like Sprinkles), and
mation can be used to assess the need for specific zinc lipid-based nutrient supplements. The review also
interventions and for targeting these activities, and as a identified a number of existing delivery platforms,
baseline to assess the future impact of such programs. within both the public and the private sectors, that
could be exploited for delivering preventive zinc sup-
plements to children, because the currently ongoing
Mobilizing interest in zinc nutrition activities provide reasonably frequent and reliable con-
tacts with the target groups and high levels of coverage.
There has been a recent resurgence in global atten- Examples of potential delivery platforms that should
tion to the problem of malnutrition in general and be assessed for the feasibility of adding preventive zinc
of micronutrient deficiencies in particular, as is evi- supplementation are listed below. The choice of which
denced, for example, by the creation of the REACH platform(s) is(are) most appropriate in a particular set-
initiative (Ending child hunger and undernutrition ting needs to be made at the country level.
initiative [http://endingchildhunger.blogspot.com]) » Twice-yearly vitamin A supplementation, which
and the development of the new multipartner 10-year increasingly is being integrated into semiannual
strategy for reduction of vitamin and mineral defi- events for child survival, and expanded with multiple
ciencies [16]. Moreover, the aforementioned Lancet additional product-delivery components;
series on maternal and child undernutrition [13, 17] » Growth monitoring and promotion programs, if
and the Copenhagen Consensus publication [14] have these are sufficiently well organized and utilized to
captured global interest and can be used as powerful support and justify additional interventions;
advocacy tools. There are multiple global, regional, » Community-based or community-directed distri-
and national forums addressing nutrition and child bution programs, such as are being adopted, for
example, to distribute preventive and therapeutic motivating potential beneficiaries, and a strategy for
medicines for certain parasitic diseases; procuring and distributing zinc supplements. These
» Social marketing through private-sector distribution activities can be used to revitalize national diarrheal
channels disease control programs at the same time that zinc
Further, in view of the apparent impact of preventive treatment is introduced into the national diarrhea
zinc supplementation in reducing mortality in small- treatment protocol. Increased efforts are particularly
for-gestational-age infants and the multiple special important in Africa and South-East Asia, where deaths
needs of these infants, there are compelling reasons due to diarrhea remain very high [21].
for designing programs to identify and provide special
support to low-birthweight infants, including preven-
tive zinc supplementation. Implementing and scaling up zinc
As summarized above, there is less consistent evi- fortification
dence for a beneficial impact on pregnancy outcomes
when preventive zinc supplementation is given to Implementation of fortification programs requires a
pregnant women. However, in view of the possible number of steps, which have been described in detail
benefits of zinc supplementation for reducing the risk elsewhere [22]. Multiple vehicles for mass and targeted
of premature delivery, the possible positive impact of zinc fortification are available, and guidelines have
zinc supplementation on infant birthweight among recently been published on recommended levels of zinc
undernourished women, and the possible benefits for fortification of cereal flour [23]. The recommended
infant postnatal health, as well as the lack of reported level of zinc fortification of cereal flour depends on the
adverse effects, zinc should be included in maternal population’s usual intake of the cereal product(s), the
supplements given during pregnancy in populations degree of milling (hence, the intrinsic zinc and phytate
at risk for zinc deficiency. contents of the flour), and the zinc and phytate contents
of the remainder of the diet. For cereal flour fortifica-
tion programs that are already operating in countries at
Implementing and scaling up therapeutic risk for zinc deficiency, it is important to review current
zinc supplementation norms and to add or adjust zinc levels in line with the
new guidelines. For cereal flour fortification programs
As indicated in the present series of papers [2], recom- currently being planned in countries with an elevated
mendations already exist for including zinc supplemen- risk of zinc deficiency, zinc should be included in the
tation as adjunctive therapy during episodes of acute fortification premix at appropriate levels.
and persistent diarrhea [3]. These revised guidelines
take into account two new approaches: the new for-
mulation for oral rehydration solutions containing Implementing and scaling up dietary
lower concentrations of glucose and salt, and providing interventions
children with 20 mg of supplemental zinc per day for 10
to 14 days (10 mg per day for infants less than 6 months Ultimately, the most desirable approach to eliminate
old). Despite the global consensus on these recom- zinc deficiency will be to ensure universal access to
mendations, progress in mainstreaming these revised diets with adequate zinc content and bioavailability.
guidelines has been slow. To help countries revise their There is considerable experience in designing and
national diarrhea programs, useful implementation implementing programs to promote increased house-
guidelines have been developed in several languages hold food production through home gardens and
by the Zinc Task Force* [18–20]. As a first step in small-animal production, including poultry, fish, and
promoting inclusion of zinc in diarrhea treatment, small ruminants, but there is limited experience in
existing diarrhea program guidelines, professional taking such programs to scale [24]. These approaches
practices, and domiciliary treatment tendencies should have the added benefit of addressing multiple nutrient
be assessed. Information is also needed on care-seeking shortfalls, not just zinc deficiency, and of providing
behaviors and household procurement of medicines other economic advantages to the household [11].
and oral rehydration therapy. This information can For greatest success in terms of nutritional outcomes,
be used to develop training guidelines for health pro- household food production interventions should
fessionals, a communication plan for informing and be linked to effective behavior change communica-
tion interventions and focus on women, so that the
* The Zinc Task Force was formed as a collaborative products will be used by the household members to
working group among UNICEF, WHO, the Johns Hopkins address specific nutritional problems of the most vul-
Bloomberg School of Public Health, and the US Agency for
International Development (USAID), with support from the nerable members of the family. Although improved
Bill and Melinda Gates Foundation, to accelerate the adoption household production of zinc-rich foods, particularly
of zinc for diarrhea management. those foods derived from animal sources, is expected
Conclusions and mainstreaming zinc interventions S183
to enhance the consumers’ zinc status, there have been needed from efficacy trials for particular diseases, such
few attempts to conduct rigorous evaluations. Thus, as acute lower respiratory infection and tuberculosis,
it is still not possible to state which specific dietary and for selected age groups.
approaches are most likely to mitigate the problem of A great deal of research is still needed regarding the
zinc deficiency. impact of maternal zinc supplementation during preg-
nancy on both maternal and infant health outcomes.
In particular, studies should focus on the optimal time
Research needs of initiating supplementation (pre- or periconceptual
vs. later in pregnancy), maternal factors, such as body
There are a number of remaining knowledge gaps con- weight and zinc status, that may modify the response
cerning zinc nutrition, assessment of zinc status, and to zinc supplementation, and possible adverse effects
the efficacy, effectiveness, and cost of zinc intervention of supplementation. Longer-term outcomes for child
programs, which need to be addressed to permit more health also should be explored. The effect of zinc
rapid and extensive implementation of zinc interven- supplementation of undernourished women on their
tion programs. The most important research needs breastmilk zinc concentrations should be studied
were described in the individual sections of this docu- further.
ment, so they will be reviewed only in general terms The efficacy and effectiveness of zinc fortification
in this concluding paper. Readers are also referred to a of different food vehicles should be determined for
recent publication that applied a new research priority- different age and sex groups. Recent guidelines on
setting methodology to review zinc-related research appropriate levels of zinc fortification of cereal flour
needs for public health programs [25]. [23] should be assessed in the context of ongoing pro-
A recurring theme in many recent documents on grams to confirm their adequacy. Greater attention also
zinc nutrition is the need for novel biomarkers of zinc should be devoted to the design and evaluation of the
status that are reliable, low-cost, and feasible for field nutrition and health impacts of dietary interventions
implementation. Nevertheless, consensus has been designed to enhance zinc status. Such interventions
reached on possible approaches to assess population may include increased production and consumption of
zinc status [15], using currently available indicators, so animal-source foods, food processing to reduce phytate
the most pressing research need for public health pro- content, and strategies to increase the zinc content of
grams is to apply the existing techniques so as to gener- staple foods, for example, through biofortification or
ate more information on the prevalence and major risk better agronomic practices.
factors of zinc deficiency in different countries. When-
ever nutrition status surveys are conducted in countries
that have been identified as having an elevated risk of Conclusions
zinc deficiency, the possibility of including indicators
of zinc status should be considered. In summary, there is now clear evidence of the benefit
On the topic of preventive zinc supplementation, of selected interventions to eliminate or reduce the risk
the major research questions deal with optimal dosage of zinc deficiency. Such interventions, if implemented
regimens (amount, frequency, and duration of sup- at scale and attaining high coverage, would have a
plementation), appropriate combinations of micro- direct impact on achieving several Millennium Devel-
nutrients, and the effectiveness of available delivery opment Goals, including reducing child morbidity,
platforms that might be exploited for distributing the mortality, and restricted growth. A global commitment
supplements. Information is also needed on the efficacy is urgently needed to conduct systematic assessments of
of preventive supplementation for reducing mortality population zinc status and to develop and mainstream
among low-birthweight infants. Many of the same interventions to control zinc deficiency in the context
dosage and delivery issues pertain to therapeutic zinc of existing health and nutrition programs.
supplementation, although more information is also
References
1. Brown KH, Peerson JM, Baker S, Hess SY. Preventive zinc ment of acute diarrhoea. Geneva: WHO/UNICEF, 2004.
supplementation among infants, preschoolers, and older 4. Hess SY, King JC. Effects of maternal zinc supplementa-
prepubertal children. Food Nutr Bull;2009;30:S12–40. tion on pregnancy and lactation outcomes. Food Nutr
2. Haider BA, Bhutta ZA. The effect of therapeutic zinc Bull 2009;30:S60–78.
supplementation among young children with selected 5. Mahomed K, Bhutta Z, Middleton P. Zinc supplemen-
infections: A review of the evidence. Food Nutr Bull tation for improving pregnancy and infant outcome.
30:S41–59. Cochrane Database Syst Rev 2007;CD000230.
3. World Health Organization/UNICEF. Clinical manage- 6. Osendarp SJ, van Raaij JM, Darmstadt GL, Baqui AH,
Hautvast JG, Fuchs GJ. Zinc supplementation during 17. Bhutta ZA, Ahmed T, Black RE, Cousens S, Dewey K,
pregnancy and effects on growth and morbidity in low Giugliani E, Haider BA, Kirkwood B, Morris SS, Sach-
birthweight infants: A randomised placebo controlled dev HP, Shekar M. What works? Interventions for
trial. Lancet 2001;357:1080–5. maternal and child undernutrition and survival. Lancet
7. Osendarp SJ, West CE, Black RE. The need for maternal 2008;371:417–40.
zinc supplementation in developing countries: An unre- 18. World Health Organization/UNICEF. Implementing the
solved issue. J Nutr 2003;133:817S–27S. new recommendations on the clinical management of
8. Iannotti LL, Zavaleta N, Leon Z, Shankar AH, Caulfield diarrhoea. Guidelines for policy makers and programme
LE. Maternal zinc supplementation and growth in Peru- managers. Geneva: WHO, 2006.
vian infants. Am J Clin Nutr 2008;88:154–60. 19. World Health Organization/UNICEF/US Agency for
9. Hess SY, Brown KH. Impact of zinc fortification on zinc International Development/US Pharmacopeia/Drug
nutrition. Food Nutr Bull 2009;30:S79–107 Quality and Information Program/Johns Hopkins
10. Brown KH, Engle-Stone R, Krebs NF, Peerson JM. Bloomberg School of Public Health. Production of zinc
Dietary intervention strategies to enhance zinc nutri- tablets and oral rehydration solution. Guidelines for
tion: Promotion and support of breastfeeding for infants programme managers and pharmaceutical manufactur-
and young children. Food Nutr Bull 2009;30:S144–71. ers. Geneva: WHO, 2007.
11. Gibson RS, Anderson VP. A review of interventions 20. World Health Organization. The treatment of diarrhoea:
based on dietary diversification or modification strate- Manual for physicians and other senior health workers.
gies with the potential to enhance intakes of total and Geneva: WHO, 2005.
absorbable zinc. Food Nutr Bull 2009;30:S108–43. 21. Boschi-Pinto C, Velebit L, Shibuya K. Estimating child
12. Hotz C. The potential to improve zinc status through mortality due to diarrhoea in developing countries. Bull
biofortification of staple food crops with zinc. Food Nutr World Health Organ 2008;86:710–7.
Bull 2009;30:S172–8. 22. Allen L, de Benoist B, Dary O, Hurrell R, eds. Guide-
13. Black RE, Allen LH, Bhutta ZA, Caulfield LE, de Onis lines on food fortification with micronutrients. Geneva:
M, Ezzati M, Mathers C, Rivera J. Maternal and child World Health Organization, 2006.
undernutrition: Global and regional exposures and 23. Brown KH, Hambidge KM, Ranum P, Tyler V, and the
health consequences. Lancet 2008;371:243–60. Zinc Fortification Working Group. Zinc fortification of
14. Horton S, Alderman H, Rivera JA. Copenhagen consen- cereal flours: Current recommendations and research
sus challenge paper: Hunger and malnutrition. Copen- needs. Food Nutr Bull 2009; in press.
hagen, Denmark: Copenhagen, Consensus Center, 24. World Bank/International Food Policy Research Insti-
2008. tute. From agriculture to nutrition: Pathways, synergies
15. de Benoist B, Darnton-Hill I, Davidsson L, Fontaine O, and outcomes. Washington, DC: World Bank, 2007.
Hotz C. Conclusions of the joint WHO/UNICEF/IAEA/ 25. Brown KH, Hess SY, Boy E, Gibson RS, Horton S, Osen-
IZiNCG interagency meeting on zinc status indicators. darp SJ, Sempertegui F, Shrimpton R, Rudan I. Setting
Food Nutr Bull 2007;28:S480–S79. priorities for zinc-related health research to reduce
16. Sanghvi T, van Ameringen M, Baker J, Fiedler J. Vitamin children’s disease burden worldwide: An application
and mineral deficiencies technical situation analysis: of the Child Health and Nutrition Research Initiative’s
A report for the ten year strategy for the reduction research priority-setting method. Public Health Nutr
of vitamin and mineral deficiencies Food Nutr Bull 2009;12:389–96.
2007;28:S157–S219.
List of Contributors
IZiNCG Steering Committee Christine Hotz, Ph.D. (author, IZiNCG internal

reviewer)
Kenneth H. Brown, M.D. (guest editor, author, Nutrition Coordinator
IZiNCG internal reviewer) HarvestPlus, and International Food Policy Research
Professor Institute
Department of Nutrition, and Program in Interna- Washington, DC, USA
tional and Community Nutrition
University of California Janet C. King, Ph.D. (author, IZiNCG internal
Davis, California, USA reviewer)
and Africa Regional Advisor in Nutrition and Child Senior Scientist
Survival Children’s Hospital Oakland Research Institute
Helen Keller International Oakland, California, USA
Dakar, Senegal Professor Emeritus, University of California at Berke-
ley and Davis
Sonja Y. Hess, Ph.D. (guest editor, author, IZiNCG
internal reviewer) Bo Lönnerdal, Ph.D. (author, IZiNCG internal
Research Nutritionist reviewer)
Department of Nutrition, and Program in Interna- Professor
tional and Community Nutrition Department of Nutrition, and Program in Interna-
University of California tional and Community Nutrition
Davis, California, USA University of California
Davis, California, USA
Shawn K. Baker, M.P.H (author, IZiNCG internal
reviewer) Juan A. Rivera, Ph.D. (author, IZiNCG internal
Regional Director for Africa reviewer)
Helen Keller International Director
Dakar, Senegal Nutrition and Health
National Institute of Public Health
Zulfiqar A. Bhutta, M.D., Ph.D. (author, IZiNCG Cuernavaca, Mexico
internal reviewer)
Professor and Chair Omar Dary, Ph.D. (IZiNCG internal reviewer)
Department of Paediatrics and Child Health Food Fortification Specialist
The Aga Khan University A2Z/The USAID Micronutrient and Child Blindness
Karachi, Pakistan Project
Washington, DC, USA
Rosalind S. Gibson, Ph.D. (author, IZiNCG internal
reviewer) Marie T. Ruel, Ph.D. (IZiNCG internal reviewer)
Research Professor Director
Department of Human Nutrition Food Consumption and Nutrition Division
University of Otago International Food Policy Research Institute
Dunedin, New Zealand Washington, DC, USA
S186 List of contributors
Emorn Wasantwisut, Ph.D. (IZiNCG internal K. Ryan Wessells (research assistant)

reviewer) Research Assistant
Senior Advisor Department of Nutrition, and
Institute of Nutrition Program in International and Community Nutrition
Mahidol University University of California
Salaya, Nakhon Pathom, Thailand Davis, California, USA
Ricardo Uauy, M.D., Ph.D. (Coordinator of external

Other contributors review)
President of the International Union of Nutritional
Victoria P. Anderson, M.Sc. (author) Sciences (IUNS)
Research Assistant Professor of Public Health Nutrition
Department of Human Nutrition London School of Hygiene and Tropical Medicine
University of Otago London, UK
Dunedin, New Zealand and Institute of Nutrition University of Chile INTA
Santiago, Chile
Reina Engle-Stone (author)
Research Assistant Nita Dalmiya, M.P.H. (external reviewer)
Department of Nutrition, and Nutrition Specialist
Program in International and Community Nutrition Nutrition Section
University of California UNICEF
Davis, California, USA New York, USA
Batool A. Haider, M.D. (author) Ian Darnton-Hill, M.B.B.S., Ph.D., M.P.H. (external
Epidemiologist and Senior Instructor (Research) reviewer)
Department of Paediatrics and Child Health UNICEF Special Adviser to the Executive Director on
The Aga Khan University Ending Child Hunger and Undernutrition
Karachi, Pakistan UNICEF, New York, USA
and REACH: Ending Child Hunger and Undernutri-
Nancy F. Krebs, M.D. (author) tion Inter-Agency Taskforce
Professor World Food Programme
Department of Pediatrics Rome, Italy
University of Colorado School of Medicine
Denver, Colorado, USA Olivier Fontaine, M.D. (external reviewer)
Medical Officer
Janet M. Peerson, MS (author) Department of Child and Adolescent Health and
Statistician Development
Department of Nutrition, and World Health Organization
Program in International and Community Nutrition Geneva, Switzerland
University of California
Davis, California, USA Dominic Schofield, M.B.A. (external reviewer)
Nutrition Specialist (Food Fortification)
Josh Jorgensen (research assistant) Nutrition Section
Research Assistant UNICEF
Department of Nutrition, and New York, USA
Program in International and Community Nutrition
University of California
Davis, California, USA

FNB v30n1 Supplement Izinc PDF

Uploaded by

Copyright:

Available Formats

FNB v30n1 Supplement Izinc PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

FNB v30n1 Supplement Izinc PDF

Uploaded by

Copyright:

Available Formats

FOOD AND

Preventive zinc supplementation in children

Therapeutic zinc supplementation in children

Zinc supplementation during pregnancy and lactation

Dietary diversification or modification to enhance zinc intakes

Zinc intake through breastmilk

Improving zinc status through biofortification

Conclusions and mainstreaming zinc interventions

International Nutrition Foundation

International Zinc Nutrition Consultative Group Technical Document No. 2

Editor: Dr. Irwin H. Rosenberg, Friedman School of Nutrition Science

Food and Nutrition Bulletin, vol. 30, no. 1, Supplement

Abstract Key words: Assessment, zinc, zinc deficiency, zinc

Biol Chem 2002;277:26389–95. insulin-like growth factor-I, and the glucose/insulin

Kenneth H. Brown, Janet M. Peerson, Shawn K. Baker, and Sonja Y. Hess

Abstract There were no significant effects on children’s behavioral

Section 1 in the meta-analyses, the analytic methods that were

1,620 articles identified in PubMed; 1,538 articles rejected

87 articles reporting on randomized,

55 individual randomized trials

Placebo vs. zinc (n = 35)

FIG. 1. Number of articles and individual studies included in the meta-analysis

[30] Hong risk pregnancy B vitamins + zinc

[45, 47] Bhandari schoolchildren Vitamin A + zinc

[10, 93] Friis

tion on the incidence of diarrhea Bangladesh 2003b Fe [95] (n=327)

found that there was no relation Peru 2007 [58] (n=198)

a total of 1,692 children. The ln (relative risk)

10 Malaria morbidity. The effects of zinc supplementa-

in clinic visits for malaria, and one of which found

Bangladesh 2002 [22] (n=299)

Subgroup (n= 6,950)

Bangladesh 2005 [98] (n=1,474)

–1.0 –0.5 0 0.5 1.0 1.5 2.0

lasted at least 6 months or to those studies that enrolled subjects.

–1.0 –0.5 0 0.5 1.0 1.5 2.0

All (n= 6,221)

–1.0 –0.5 0 0.5 1.0 1.5 2.0

FIG. 8. Effect of zinc supplementation on change in weight-for-height z-score

Bangladesh 2001 [21] (n=168)

All (n= 4,571)

–1.0 –0.5 0 0.5 1.0 1.5 2.0 2.5 3.0

–1.0 –0.5 0 0.5 1.0 1.5 2.0

Bangladesh 2002 [23] (n=198)

–1.0 –0.5 0 0.5 1.0 1.5 2.0

Bangladesh 2003b [97] (n=124)

–1.0 –0.5 0 0.5 1.0 1.5 2.0

FIG. 12. Effect of zinc supplementation on change in hemoglobin concentration

–1.0 –0.5 0 0.5 1.0 1.5 2.0

FIG. 13. Effect of zinc supplementation on change in serum or plasma ferritin

Bangladesh 2003b [97] (n=80)

–1.0 –0.5 0 0.5 1.0 1.5 2.0

FIG. 14. Effect of zinc supplementation on change in serum copper concentra-

Section 3 Twice-yearly vitamin A supplementation (VAS). Glo-

Batool A. Haider and Zulfiqar A. Bhutta

Abstract Key words: Diarrhea treatment, infections, malaria,

supplementation as an adjunct in the treatment of supplementation in the treatment of acute diarrhea in

Methods Detailed review of the evidence

in the supplemented group