Journal of Community Positive Practices, XXII(2) 2022, 79-98

ISSN Print: 1582-8344; Electronic: 2247-6571

A LOGISTIC REGRESSION ANALYSIS

OF DETERMINANTS OF CHILD
MALNUTRITION IN UTTAR PRADESH,
INDIA

Reena KUMARI11
Rekha GUPTA12

DOI: https://doi.org/10.35782/JCPP.2022.2.06

Abstract: This paper examines the effects of socio-economic determinants on child malnutrition in
Uttar Pradesh- the most populous and malnourished state of India. Using data from NFHS-4,
this study highlights district-wide variations in nutritional status of children less than five years in
the state. The percentage of stunted, wasted, and underweight have been taken as dependent
variables, while main explanatory variables of child malnutrition were female education, maternal
malnutrition (BMI), breastfeeding practices, children's place of residence, father’s education, wealth
index of household, family planning, sex of the child, sex of the household and age cohort of mother.
Logistic regression models have been applied in order to observe the association between explanatory
variables and malnutrition. The findings revealed a high incidence of less stunted, wasted, and
underweight infants in case of an educated parents, particularly educational level of mother. Wealth
index representing the possession of assets by the household found to be an important indicator in
the sense that children born to households with richer and richest quintiles showing better-
nourished children aged under-5. The modern method of using contraceptive, breastfeeding practices,
maternal malnutrition were also important indicators influencing the rate of malnutrition in
children. The study suggested that imparting education to females, breastfeeding practices, economic
status, and contraceptive use can play an important role in reducing nutritional status of children.
Keywords: malnutrition, stunted, wealth index, education, contraceptive use.

11 Assistant Professor, Department of Economics, C.M. College, L.N. Mithila University, Darbhanga,
Bihar-846004, E-mail: raireena86@gmail.com, ORCID ID- 0000-0003-2338-2059
12 Assistant Professor, Department of Economics, University of Allahabad, Prayagraj-211002, E-mail:

rekhaeco@gmail.com.
80 A logistic regression analysis of determinants of child malnutrition in Uttar Pradesh, India

1. Introduction
Malnutrition, defined as deficiency or imbalances in a person’s nutrients intake, is
considered a sensitive indicator and most common nutritional disorder for any
developing nation. The high prevalence of child malnutrition reflects not only a
country’s past health status but also its future health trails (Subramanyam et al., 2010).
Addressing the problem of malnutrition, the sustainable development goals (SDG)
adopted by United Nations (2019) indicates ‘By 2030, end all forms of malnutrition, including
achieving, by 2025, the internationally agreed targets on stunting and wasting in children under 5 years
of age, and address the nutritional needs of adolescent girls, pregnant and lactating women, and older
persons’ under Goal 2.
In this context, monitoring child malnutrition is crucial for planners and policy makers
in a country like India, a country which is facing the high prevalence of child
malnutrition, accounting for one third of the world’s wasted (height-for-age). UNICEF
contends that in India, the percentage of underweight and stunted children less than
five years were 43 percent and 48 percent respectively because of enduring
undernutrition. It is estimated that worldwide, India executes poorly in child
malnutrition status and ranked 114 out of 132 countries, just prior to Afghanistan and
Pakistan (IFPRI, 2016) and in the early 2000s, it was around one-third of all
malnourished children in the world (Svedberg, 2008).
Several studies indicated that child malnutrition affected child morbidity, health-care
expenditure and economic progress of any country (Measham & Chatterjee,1999;
Mishra, Lahiri & Luther, 1999) which further increase the risk of child ailment,
retardation in mental and physical growth, resulting into low ability to work, loss of
output and efficiency, high burden of health care expenditures and lastly can pledge the
poverty trap (World Bank, 2006; Grantham et al., 2007; Tarozzi & Mahajan, 2007;
Walker et al., 2007). Also. It is considered that severe or chronic malnutrition increase
the educational attainment and outcome of the children (Ahmed et al., 2012).
International Food Policy Research Institute (IFPRI) reports that in developing
countries, one out of every three children under-five, is malnourished. It is clearly noted
that the high prevalence of malnutrition not only cause severe cognitive and physical
loss of children but also it is a collapse or exploitation of a child’s human right (Das &
Sahoo, 2011).
The population of Uttar Pradesh is 199.5 million (India’s most populated state) and 39
percent of the total child population of India resides in the state. According to the
report published by UP State Nutrition Mission, in this state, half of all children age-5
are stunted, and 10 percent were wasted in 2014. In Uttar Pradesh, infant mortality rate
(IMR) and malnutrition rate are the highest in the country. National Family Health
Survey-4 (NFHS) data revealed that in Uttar Pradesh, the percentage of malnourished
children has declined significantly in the last ten years. In 2005-06, 56.8 percent of
children under-5 years of age were stunted, and 42.4 percent were underweighted; these
figures fell to 46.3 percent and 39.5 percent respectively, in 2015-16 (figure 1).
As per the report published by National Sample Survey Office (NSSO), the percentage
of poor in the state is 30.4 percent higher than all India (25.7%) poor. IMR is 48 per
 Reena KUMARI, Rekha GUPTA  81

thousand live births, the under-5 mortality rate is 73 per thousand live births, and life
expectancy is 62.9 years for male and 65.4 years for female. However, the mortality rate
among the children under-5 years of age fell by 18 percent from 2005-06 to 2015-16,
and IMR fell by 9 percent during the same period. Improvement in child nutrition
during the inter-survey period is the result of interventions such as broader coverage of
midday meals to children at school and rigorous operation of Integrated Child
Development Scheme (ICDS) at a more disaggregated level, underpinning of health
outreach facilities to rural and poor ménages for on-time vaccination of children and
strengthening of the public distribution system (PDS) of providing food grains at
promoted prices to the weaker section.
However, undernutrition of children remains a major health problem at the
disaggregated level in India. There has been a remarkable difference in IMR across the
districts; for example, highest in Shrawasti (96) while lowest in Kanpur Dehat (37). In
Uttar Pradesh, a considerable variation has been observed between rural and urban in
terms of child malnutrition. In rural areas of Uttar Pradesh, prevalence of stunted and
underweight among children was 48.5 percent and 41 percent respectively, as compared
with 37.9 percent and 33.7 percent among urban children. In this framework, a study of
child malnutrition and its important determinants is required for policy implications
and therefore, the present study examined the effects of various socio-economic factors
on child malnutrition children less than five years in Uttar Pradesh, employing the data
from NFHS-4 (2015-16).

Figure 1: Malnutrition trends in India and Uttar Pradesh

children among aged 0-5 years
82 A logistic regression analysis of determinants of child malnutrition in Uttar Pradesh, India

2. Previous research
There are several socio-economic and demographic factors (place of residence, age, and
nutritional status; and maternal education) responsible for poor nutritional status of
children (Bangladesh Demographic and Health Survey, 2007). The benefits of mother’s
education and its impact on child health status is well documented in literature which
lowers the rate of child malnutrition, in other words, the prevalence of malnutrition is
lower in case of educated mother (Frost, Forste & Hass, 2005; Basu & Stephenson,
2005; Miller & Rodgers, 2009; Cleland, 2010; Bbaale, 2014; Burroway, 2016). A study
by Mosle & Chen (1985) reported that effect of mother’s education on children’s
nutritional status works through a set of ‘proximate determinants’, such as, fertility
factors, environmental amenities, breastfeeding practices, and utilization of childcare
services. Similarly, a host of studies contend that more educated mothers may have
more nourished children because they have better knowledge about child rearing
practices and heath care and utilization (Thomas et al., 1991; Desai & Alva, 1998;
Glewwe 1999; Currie & Moretti, 2003).
A substantial range of literature in the domain of health and education have discussed
the casual relationship between mother’s education and nutritional status of children
and given some possible trails to find the link. The important studies include: (1) more
educated mothers can take better care of their children about nutrition and health care,
for example, washing of clothes, feeding practices, treatment during sickness and
availing timely vaccination services etc. (Caldwell 1979; Mondal et al., 2009) and better
sanitation such as flushing toilets and running water (Horton, 1998; Case, 2001,
Choudhury, 2015 ) (2) An educated mother can make healthier choices for themselves
during pregnancy that effectively prevent fatal childhood diseases (Govindasamy &
Ramesh, 1997) and directly affected the health of the child at birth (Miller & Rodgers,
2009); (3) Educated mother can give better level of living as previous research
documented that education provides women ample opportunities to work outside the
home and earn an income which empower them through greater authority and
bargaining capability in the family (Nussbaum, 2004; Sen, 1999).
However, a growing body of literature has drawn attention towards the impact of
father’s education in improving child nutritional status (Case et al., 2002). It is well-
established fact now that educational level of father plays a crucial role in family’s
financial status which in turn ensures better access to child health facilities and further
better child nutritional status (Choudhury, 2015). India’s NFHS data also revealed a
very high percentage of maternal malnutrition which shows the worse-off condition of
women. Studies claimed that mother’s nutritional status had a positive association with
child nutritional status (Sethuraman, 2006).
Poor breast-feeding practices in early childhood contribute to the burden of
malnutrition and infant and child mortality (Patel et al., 2010). It is observed that
working mothers have limited time for breastfeeding and to manage other nutritional
and health services to their children. Working mothers are dependent on a caretaker for
their child and caretaker’s poor attention, affection and involvement in child feeding are
the major causes for undernutrition of children. Moreover, a large body of empirical
research claimed that caretakers might not make the best use of offered resources due
 Reena KUMARI, Rekha GUPTA  83

to less information of optimal feeding behaviours and inappropriate cultural beliefs and
practice about nourishing (Allen and Gillespie, 2001; Moore, Akhter and Aboud, 2006).
Also, location or place of residence is one of the socioeconomic covariates often used
in the literature which determines child malnutrition across the region (Sastry, 1997). A
growing body of literature demonstrated that wealth status of a family has a significant
and positive impact on nutritional status of children (Smith & Haddad, 2000; Haddad
et al., 2003; Heltberg, 2009; Headey, 2013). A variety of indicators has been used to
measure economic status of the household, however, sorting of households according
to income level is more relevant proxy. While many studies claimed that malnutrition is
associated with poverty and disease (Dasgupta et. al., 2005) and argued that due to poor
access of food a large segment of children is suffering from severe health problems and
morbidity.
To improve child nutritional status, Government of India has introduced major
nutrition supplementation programs such as Integrated Child Development Services
(ICDS) and Mid-day Meal (MDM) Programme at the disaggregated level. However,
little is known about the empirical findings of the factors that drive child malnutrition,
especially in poor and bigger states like Uttar Pradesh. In this connection, in order to
formulate appropriate strategies related to child health outcomes, policy makers and
planners must prerequisite familiarity of the factors responsible for poor child outcome
such as stunting, wasting and underweight.

Therefore, the main objectives of this paper are to understand nutritional status based
on stunted, wasted and underweight of children less than the age of five and to examine
the factors which influence child malnutrition in respect of certain socio-economic
variables, for instance, father’s education, place of residence, wealth status, family
planning, sex of the child, sex of the household, religion; and maternal indicators like
mother’s malnutrition (BMI) and age cohort of mother. This study examined the
various socio-economic and household factors which are responsible for child
malnutrition in Uttar Pradesh.

3. Data and methods

In this study, data has been assessed from the NFHS-4 conducted by the International
Institute of Population Sciences (IIPS), Mumbai in 2015-16 and constrained to children
under the age of five. NFHS is a collaborative project of IIPS and many organizations
including foreign institutions. IIPS was assigned as the nodal agency by the Ministry of
Health and Family Welfare, Government of India and responsible for providing
coordination and technical assistance for the NFHS-413.

13 NFHS-4 is funded by the United States Agency for International Development (USAID), DFID,
the Bill and Melinda Gates Foundation (BMGF), UNICEF, UNFPA, the MacArthur Foundation
and the Ministry of Health and Family Welfare (MoHFW), Government of India.
84 A logistic regression analysis of determinants of child malnutrition in Uttar Pradesh, India

The survey used a uniform questionnaire, two-stage sampling design with stratification
by rural-urban background and field practice to make possible comparability of the data
and to achieve enhanced data quality. For anthropometric analysis, malnutrition of
children is enumerated in terms of anthropometrical measures- weight-for-age
(stunted), height-for-age (wasted) and weight-for-height (underweight) and Z-score ≤2
standard deviation, World Health Organisation/National Centre for Health Statistics
reference standards were applied to compute the standardized Z-score for all three
measures (WHO, 1978; Mazumdar, 2010).
In this survey, data on children’s height, weight and underweight have been collected
less than 5 years. Mother’s education, place of residence (rural-urban), sex of the child
(boys-girls), sex of the household, father’s education, wealth index (based on an
ownership-of-goods index), family planning (use of contraceptive), mother’s BMI,
breastfeeding practices and mother’s age cohort are the important indicators that are
assumed to effect nutritional status of a child. The logistic regression analysis has been
employed in the analysis for estimation of the odds of being malnourished in the
children of Uttar Pradesh.
Dependent variable denotes the number of children whose z-scores are below -2 are
coded 1 and those with z-scores of -2 or higher are coded 0. The details of explanatory
variables have been given in table 1 and thus results obtained are compared with
reference category. In order to compute variables, research design and sample weights,
statistical tool is carried out using SPSS (20.0 version) software for window and the
significance levels of p<0.01, 0.05 and 0.10 were taken.

Table 1: Symbolization and definition of variables used

in the logistic regression
Symbolization Independent variables Definition
Mother’s Educational level of the mother
education (dummy variable)
Mother_illitR Mother is illiterate =1, if mother is illiterate
=0, otherwise
Mother_elemedn Mother received elementary level of =1, if mother is having 1-8
education years of schooling
=0, otherwise
Mother_secedn Mother received secondary level of =1, if mother is having 9-10
education years of schooling
=0, otherwise
Mother_hiedn Mother received senior secondary or =1, if mother is having 11 or
above level of education more years of schooling
=0, otherwise
Sex_child Sex of child
(dummy variable)
Child_boyR Boy child =1, if the child is boy
=0, otherwise
Child_girl Girl child =1, if the child is girl
=0, otherwise
 Reena KUMARI, Rekha GUPTA  85

Symbolization Independent variables Definition

Father’s education Educational level of the father
(dummy variable)
Father_illiterateR Father is illiterate =1, if father is illiterate
=0, otherwise
Father_elemedn Father received elementary level of =1, if father is having 1-8
education years of schooling
=0, otherwise
Father_secedn Father received secondary level of =1, if father is having 9-10
education years of schooling
=0, otherwise
Father_hiedn Father received senior secondary or =1, if father is having 11 or
above level of education more years of schooling
=0, otherwise
Sex_hhn Sex of Household
(Dummy variable)
MaleR Male member =1, if the sex of household is
male
=0, otherwise
Female Female member 1, if the sex of household is
female
=0, otherwise
Place_residence Place of residence (dummy variable)
UrbanR Place of residence in urban areas =1, if the location of
household is urban
=0, otherwise
Rural Place of residence in rural areas =1, if the location of
household is rural
=0, otherwise
Contraceptive_use Use of contraceptive or family planning
by household (dummy variable)
Trad_methodR Household using traditional method =1, if the household is using
traditional method
=0, otherwise
Mode_method Household using modern method =1, if the household is using
modern method
=0, otherwise
Others_method Household using other method =1, if the household is using
other method excluding
traditional and modern
=0, otherwise
Wealth Index Wealth index reflecting the ownership
of some basic assets by the household
(dummy variable)
PoorestR Household is belonging to the poorest =1, if the household is
income category belonging to the poorest
category
=0, otherwise
Poor Household is poor 1, if the household is poor
=0, otherwise
86 A logistic regression analysis of determinants of child malnutrition in Uttar Pradesh, India

Symbolization Independent variables Definition

Middle Household is belonging to the middle- 1, if the household is
income category belonging to the middle-
income category
=0, otherwise
Richer Household is belonging to the richer 1, if the household is
income category belonging to the richer
income category
=0, otherwise
Richest Household is belonging to the richest 1, if the household belonging
income category to the richest income category
=0, otherwise
Body Mass Index BMI of the mother
(BMI) of mother (dummy variable)
BMI ≤18.5 Mother’s BMI is BMI ≤18.5 kg/m2 =1, if the mother’s BMI is
kg/m2R BMI ≤18.5 kg/m2
=0, otherwise
BMI ≥18.5 kg/m2 Mother’s BMI is BMI ≥18.5 kg/m2 =1, if the mother’s BMI is
BMI BMI ≥18.5 kg/m2
=0, otherwise
Breastfeeding Breastfeeding practices by mother
(dummy variable)
Within 1st hourR Child received breastmilk within 1st =1, if child received breast-
hour milk within 1st hour
=0, otherwise
More than 1st hour Child received breastmilk more than 1st =1, if child received breast-
hour milk more than 1st hour
=0, otherwise
Mother's Age Cohort Age of the mother (dummy variable)
Mother is 15-20 years old =1, if the mother is 15-20
years old
15-19R =0, otherwise
Mother is 20-24 years old =1, if the mother is 20-24
years old
20-24 =0, otherwise
25-29 Mother is 25-29 years old =1, if the mother is 25-29
years old
=0, otherwise
30-34 Mother is 30-34 years old =1, if the mother is 30-34
years old
=0, otherwise
35-39 Mother is 35-39 years old =1, if the mother is 35-39
years old
=0, otherwise
40-44 Mother is 40-44 years old =1, if the mother is 40-44
years old
=0, otherwise
45-49 Mother is 45-49 years old =1, if the mother is 45-49
years old
=0, otherwise
 Reena KUMARI, Rekha GUPTA  87

4. Results and discussion

Table 2 clearly depicted that the percentage of underweight and stunted children
increased up to one year for the boys and then decreased to about 4 years. The
percentage of wasted children increased up to one year and then decreased up to 4
years for both, boys and girls. The data demonstrates that the percentage of
underweight, stunted and wasted below the age of five was 39.7 percent, 46.5 percent
and 19.2 percent, respectively for boys and it was 39.8 percent, 46.2 percent and 16.7
percent for girls. The data also revealed that the percentage of underweight was less in
girls in comparison to boys, but it contrasts with the case of stunted and wasted.
District-wide variations in frequencies and ranks among the under-five children by sex
were given in appendix (Table 1A). It is clearly noted that inter-district variation is high
in terms of stunted, wasted and underweight in Uttar Pradesh. In terms of boys
stunted, the top five poor performer districts were Gautam Buddha Nagar, Jhansi,
Ballia, Meerut and Saharanpur. While all the bottom five performer districts where the
prevalence of stunting was high were; Bahraich, Shrawasti, Balrampur, Gonda and
Siddharth Nagar, located in the eastern region of Uttar Pradesh.
Despite being placed in the western region in the state, Gautam Buddha Nagar and
Saharanpur, both have a high percentage of stunted children. In terms of boys wasted,
Mahamaya Nagar, Etah, Farrukhabad, Gonda and Sant Kabir Nagar were the top five
performers, while Hamirpur, Lucknow, Lalitpur, Kaushambi and Jalaun were the
bottom five performers. The districts which fall under the top five performers in boys
underweight include Mathura, Mahamaya Nagar, Gorakhpur, Gautam Buddha Nagar
and Firozabad. In contrast, Budaun, Kaushambi, Chitrakoot, Jaunpur and Pilibhit were
the bottom five performers.
In terms of girls stunted, Ghaziabad, Baghpat, Lucknow, Gautam Buddha Nagar and
Rae Bareli were the top five performers while, Bahraich, Balrampur, Shrawasti,
Siddharth Nagar and Maharajganj were the poor performers. In girls wasted category,
Lalitpur, Chitrakoot, Rae Bareli, Lucknow and Hamirpur were the bottom five
performers while, Mainpuri, Farrukhabad, Kanshiram Nagar, Gonda and Deoria were
the top five performers. In terms of girls underweight, districts fall into the five top
categories were Firozabad, Mathura, Ghaziabad, Ballia and Gautam Buddha Nagar
while, Shahjahanpur, Jaunpur, Budaun, Jalaun and Sant Ravidas Nagar were the bottom
five performers in the state.
Districts with a high prevalence of stunted for both the boys and girls were observed in
Bahraich, Balrampur, Shrawasti, Siddharth Nagar while the low prevalence for both
sexes was observed in Gautam Buddha Nagar. Likewise, in terms of wasted, the high
prevalence for both sexes have been observed in some districts like Hamirpur, Lalitpur
and Lucknow. In contrast, the low prevalence for boys and girls has been observed in
Gonda. High prevalence of underweight for both boys and girls has been observed in
Budaun. In contrast, low prevalence of underweight for both has been seen in the
districts of Mathura, Forozabad and Gautam Buddha Nagar.
The percentage of stunted is higher in girls in comparison to boys in the districts like,
Rampur, Shahjahanpur, Unnao, Jalaun, Fatehpur, Basti, Ballia, Chandauli and Etah.
88 A logistic regression analysis of determinants of child malnutrition in Uttar Pradesh, India

Some districts in the state where the percentage of stunted was lower in girls in
comparison to their boy’s counterpart include Jyoptiba Phule Nagar, Firozabad,
Pilibhit, Hardoi, Lalitpur, Faizabad, Ambedkar Nagar, Jaunpur, Ghazipur and
Sonbhadra. No substantial differences between boys and girls were observed in case of
wasting. Although, considerable variation between boys and girls were found in some
districts such as Saharanpur, Mahamaya Nagar, Bareily, Hamirpur, Fatehpur,
Balrampur, Sant Kabir Nagar and Gorakhpur where the prevalence of underweight was
higher in boys than girls. On the other hand, there were some districts where the
percentage of underweight was higher in boys in comparison to girls include Mainpuri,
Pilibhit, Kanpur Nagar, Kanpur Dehat, Barabanki, Sultanpur, Bahraich, Maharajganj
and Jaunpur.

Table 2: Age and sex wise nutritional status of under- five children
in Uttar Pradesh
Sex of Stunted Wasted Underweight
Year
child N % N % N %
Boy 0 914 24.9 1113 30.3 1185 32.3
1 2032 53.2 826 21.6 1618 42.4
2 1982 52.9 629 16.8 1573 42.0
3 2046 52.5 588 15.1 1611 41.4
4 1860 48.1 499 12.9 1557 40.3
Total 8834 46.5 3655 19.2 7544 39.7
Girl 0 724 21.7 955 28.6 1032 30.9
1 1659 48.1 642 18.6 1322 38.3
2 1889 55.3 460 13.5 1467 43.0
3 1970 54.0 415 11.4 1562 42.8
4 1719 51.1 396 11.8 1468 43.6
Total 7961 46.2 2868 16.7 6851 39.8
Source: Calculated from NFHS-4 data, Government of India.

Table 3 reveals the parameter estimates for stunting. Coefficient of covariates and the
related standard error is shown in parentheses. Mother’s level of education is an
important factor in determining stunting of a child. It was found that mothers who
received elementary (OD=1.598, SE=.129), secondary (OD=1.762, SE=.137) and
higher (OD=1.442, SE=.117) level of education lower the rates of stunting. This
finding is unfailing and supported by many studies such as Roy (2000); Smith &
Hadded (2000); Headey (2013) and many others.
Likewise, father’s education also has a significant influence on stunting. The result
indicates that fathers who have completed elementary (OD=1.347, SE=.125) and
secondary (OD=1.451, SE=.128) level of education condense the rate of stunting and it
is statistically significant at 1 percent level of significance. Though, the result is not
found to be statistically significant if father has above secondary or higher level of
education.
However, household’s location of children no longer affected the rate of stunting. The
association between sex of household and stunting is negative and statistically
 Reena KUMARI, Rekha GUPTA  89

significant at 1 percent level of significance with low odd ratio. It confirms that if the of
household is headed by female (OD=.760, SE=.082) it lowers the rate of stunting.
Another important and determining factor of stunting is wealth index. It demonstrates
that children born to families with richer and richest wealth quintiles showing lower
stunting children. It means if the economic condition of a family is better it can upsurge
the level of living of the household, allow them to take indispensable care of the
children and further reduce stunting of children.
Use of modern method of contraceptive is significant indicator which also lowers the
rate of stunting. The results display that if family is using modern method (OD=1.291,
SE=.255) of contraceptive, they reduce stunting.
The association between mother’s BMI (OD=2.654, SE=.082) and stunting is robust
and highly significant with high odd ratio. The result is statistically significant at 1
percent level of significance which indicates if mother’s nutritional status is better, they
can lower the rate of stunting.
Our logistic regression results indicate a significant and negative relationship between
children belonging to the age up to 5 years who have never been breastfed within first
hour of birth (OD=.876, SE=.063) and stunting. There is no significance association
found between mother’s age cohort and stunting.

Table 3: Determinants of stunting using multiple logistic regression results

Variable Β Sig. Odd ratio 95% C.I.
Mother's Education
Mother_illitR
Mother_elemedn .469 (.129) .000 1.598 1.241-2.058
Mother_secedn .567 (.137) .000 1.762 1.347-2.306
Mother_hiedn .366 (.117) .002 1.442 1.147-1.814
Sex of Child
Child_boyR
Child_girl .093 (.056) .099 1.097 .983-1.225
Sex of Household
MaleR
Female -.275 (.082) .001 .760 .646-.893
Father's Education
Father_illiterateR
Father_elemedn .298 (.125) .017 1.347 1.054-1.720
Father_secedn .373 (.128) .004 1.451 1.130-1.864
Father_hiedn .101 (.104) .331 1.106 .903-1.356
Place of Residence
Rural .132 (.079) .095 1.142 .977-1.333
Use of Contraceptive
Trad_methodR .089 (.101) .380 1.093 .896-1.332
Mode_method .255 (.110) .021 1.291 1.040-1.602
Others_method .064 (.092) .489 1.066 .890-1.276
Wealth Index
90 A logistic regression analysis of determinants of child malnutrition in Uttar Pradesh, India

Variable Β Sig. Odd ratio 95% C.I.

PoorestR
Poor 1.121 (.130) .000 3.069 2.379-3.960
Middle .866 (.124) .000 2.377 1.865-3.030
Richer .593 (.122) .000 1.810 1.424-2.300
Richest .366 (.124) .003 1.442 1.131-1.840
Body Mass Index (BMI)
BMI ≤18.5 kg/m2R
BMI ≥18.5 kg/m2 .976 (.082) .000 2.654 2.260-3.116
Breastfeeding
Within 1st hourR
More than 1st hour -.132 (.063) .037 .876 .774-.992
Mother's Age Cohort
15-19R
20-24 -.524 (.460) .255 .592 .241-1.459
25-29 -.263 (.369) .476 .769 .373-1.585
30-34 -.048 (.367) .896 .953 .464-1.957
35-39 .008 (.369) .983 1.008 .489-2.075
40-44 -.070 (.375) .852 .933 .447-1.947
45-49 .011 (.403) .979 1.011 .458-2.229
N 41751
Cox & Snell R Square 0.085
Nagelkerke R Square 0.113
-2 Log-likelihood 7608.059
R, reference category
Figures in the parentheses are SE of estimates
Source: Author’s calculation

Table 4 represents that mother’s education has a robust and positive effect on wasting
relative to mothers with no education (model 1). The odds ratios for elementary
(OD=1.633, SE=.127), secondary (OD=1.781, SE=.135) and higher level of education
(OD=1.425, SE=.116) are more than one for all levels of education.
Father’s educational status also has a significant effect on wasting. The prevalence of
wasting is lower in the children whose father’s level of education is elementary
(OD=1.364, SE=.123) and secondary (OD=1.465, SE=.126). However, father’s
education above secondary level is not a necessary condition for wasting.
The prevalence of wasting is higher in children who are residing in rural areas. The
result demonstrates a significant effect on wasting with high odd ratio (OD=1.168,
SE=.078). As expected, due to poverty, rural people have low access to the health care
facility and better sanitation that result in a high percentage of wasting.
Also, sex of the household is another variable which determines wasting. The
prevalence of wasting is higher in the households which is headed by female
(OD=.759, SE=.081). The result is statistically significant at 1 percent level of
significance. Similar to this result, Burroway (2016) revealed an important finding that
females are more efficient than their male counterpart to use resources on nourishment
and other basic domestic needs.
 Reena KUMARI, Rekha GUPTA  91

There is a positive association between use of contraceptive and wasting. The

prevalence of wasting is lower in those families who are using modern method of
contraceptive (OD=1.307, SE=.109).
The forgoing analysis depicts that income is an important variable which determines
wasting. The results confirm that wealth index of household has a strong and positive
effect on reducing wasting in children. As expected, poor households do not provide
nutritional food and better health and medical facilities to their children which result in
a high rate of death and mortality. Due to low purchasing power, poor households do
not afford the cost of expenditure and improved sanitation.
Here, BMI is used as a proxy of nutritional status of mother which has a significant and
positive effect on wasting with high odd ratio (OD=2.597, SE=089). The coefficient is
statistically significant at 1 percent level of significance.
In this model, mothers age cohort and sex of the child do not have a significant effect
on wasting.

Table 4: Determinants of wasting using multiple logistic regression results

Variable β Sig. Odd ratio 95% C.I.
Mother's Education
Mother_illitR
Mother_elemedn .490 (.127) .000 1.633 1.273-2.094
Mother_secedn .577 (.135) .000 1.781 1.366-2.321
Mother_hiedn .354 (.116) .002 1.425 1.136-1.787
Sex of Child
Child_boyR
Child_girl .075 (.055) .172 1.078 .968-1.202
Sex of Household
MaleR
Female -.275 (.081) .001 .759 .648-.890
Father's Education
Father_illiterateR
Father_elemedn .310 (.123) .012 1.364 1.071-1.736
Father_secedn .382 (.126) .002 1.465 1.145-1.875
Father_hiedn .108 (.102) .292 1.114 .911-1.362
Place of Residence
Rural .155 (.078) .047 1.168 1.002-1.360
Use of Contraceptive
Trad_methodR .100 (.100) .317 1.105 .909-1.343
Mode_method .267 (.109) .014 1.307 1.055-1.617
Others_method .075 (.091) .410 1.078 .902-1.288
Wealth Index
PoorestR
Poor 1.103 (.128) .000 3.013 2.346-3.868
Middle .836 (.122) .000 2.307 1.818-2.928
Richer .561 (.120) .000 1.752 1.385-2.217
Richest .348 (.122) .004 1.416 1.116-1.798
92 A logistic regression analysis of determinants of child malnutrition in Uttar Pradesh, India

Variable β Sig. Odd ratio 95% C.I.

Body Mass Index (BMI)
BMI ≤18.5 kg/m2R
BMI ≥18.5 kg/m2 .954 (.080) .000 2.597 2.218-3.040
Breastfeeding
Within 1st hourR
More than 1st hour -.187 (.055) .001 .829 .744-.924
Mother's Age Cohort
15-19R
20-24 -.495 (.443) .264 .610 .256-1.452
25-29 -.222 (.354) .530 .801 .400-1.603
30-34 -.031 (.352) .929 .969 .486-1.932
35-39 .033 (.354) .927 1.033 .517-2.066
40-44 -.069 (.360) .849 .934 .461-1.893
45-49 -.023 (.387) .953 .978 .458-2.087
N 41751
Cox & Snell R Square 0.087
Nagelkerke R Square 0.116
-2 Log-likelihood 7595.278
R, reference category
Figures in the parentheses are SE of estimates
Source: Author’s calculation

The logit estimates of table 5 demonstrates that the effect of mother’s education on
child underweight is negative, with the high odds ratio more than one in each level of
education and coefficients are statistically significance at 1 percent levels of significance.
The rates of underweight were lower among children whose mothers have elementary
(OD=1.984, SE=.136), secondary (OD=1.764, SE=.146) and higher level of education
(OD=1.497, SE=.126) in comparison to mothers with no education.
This result is supported by many other studies in the domain of health and education
research which prove that educated and qualified mothers are contributing factors for
reducing malnutrition (Frosta et al., 2005; Basu & Stephenson, 2005; Miller & Rodgers,
2009; Cleland, 2010; Bbaale, 2014). Waihenya et al. (1996) described that child whose
mothers have received elementary education they lower malnutrition rate in
comparison to those whose mothers have completed higher education, especially in
case of stunting.
The effect of father’s education on children underweights at (OD=1.459, SE=.131),
secondary (OD=1.368, SE=.134) level is found to be statistically significant, with the
high odd ratio. However, our result is unfavourable in case of fathers who have earned
higher education. It suggests that there is essential to provide education to the fathers
up to the secondary level for dipping underweight in children.
Girl children have higher relative odds of 1.059 for underweight in comparison with
their boy counterparts and though, this association is not statistically significant. In our
society, it is hypothesized that girl children are supposed to be nutritionally ignored due
to their low status in society and hence are expected to be more underweight. No
significant association is found between sex of household and underweight. However, a
 Reena KUMARI, Rekha GUPTA  93

significant association is observed between location of household and underweight.

Children residing in rural areas (OD=1.154, SE=.082) are at higher risk of underweight
as that of children residing in urban areas, however, the coefficient is statistically
significant at 10 percent level of significance.
The estimated results indicate a significant and robust association between family using
the modern method of contraceptive and underweight. The rates of underweight in
children were lower whose family is using modern method of contraceptive
(OD=1.485, SE=.113) in reference to a family not using any contraceptive.
Similarly, a significant association is observed between the wealth status of the
households and underweight. Higher odds of underweight were observed among
children who belonged to poor wealth quartile (OD=2.562, SE=.131) than that of the
richest income quartile. Household’s level of living also determine the children’s
nutritional status as result shows that children from the poorest household are more
likely to malnourished than children from the richer and richest wealth quintile.
The findings of logit regression demonstrate that children whose mother’s BMI is
≥18.5 have statistically significant at 1 percent level of significance with a low odds
ratio (OD=.091, SE=.083). According to NFHS-4 report, the percentage of women
whose BMI was below normal (BMI<18.5 kg/m2) was 25.3 percent and children age 6-
59 months who were anemic, was 63.2 percent. It is appalling to note that in Uttar
Pradesh, more than half of pregnant women (51%) aged 15-49 years were anemic in
2016. However, in out result there is no significant association found between mother’s
age cohort and underweight in children.

Table 5: Determinants of underweight using multiple logistic regression results

Variable Β Sig. Odd ratio 95% C.I.
Mother's Education
Mother_illitR
Mother_elemedn .685 (.136) .000 1.984 1.518-2.592
Mother_secedn .567 (.146) .000 1.764 1.325-2.348
Mother_hiedn .404 (.126) .001 1.497 1.171-1.915
Sex of Child
Child_boyR
Child_girl .057 (.058) .323 1.059 .945-1.187
Sex of Household
MaleR
Female -.089 (.085) .296 .915 .774-1.081
Father's Education
Father_illiterateR
Father_elemedn .377 (.131) .004 1.459 1.128-1.886
Father_secedn .313 (.134) .019 1.368 1.052-1.780
Father_hiedn .160 (.111) .149 1.173 .944-1.457
Place of Residence
Rural .143 (.082) .081 1.154 .983-1.354
Use of Contraceptive
94 A logistic regression analysis of determinants of child malnutrition in Uttar Pradesh, India

Variable Β Sig. Odd ratio 95% C.I.

Trad_methodR .133 (.105) .204 1.142 .930-1.402
Mode_method .395 (.113) .000 1.485 1.189-1.854
Others_method .037 (.096) .699 1.038 .860-1.252
Wealth Index
PoorestR
Poor .941 (.136) .000 2.562 1.961-3.348
Middle .728 (.131) .000 2.072 1.604-2.677
Richer .542 (.129) .000 1.719 1.334-2.215
Richest .460 (.131) .000 1.584 1.224-2.049
Body Mass Index (BMI)
BMI ≤18.5 kg/m2R
BMI ≥18.5 kg/m2 -1.657 (.083) .000 .191 .162-.225
Breastfeeding
Within 1st hourR
More than 1st hour -.015 (.065) .821 .985 .867-1.120
Mother's Age Cohort
15-19R
20-24 -.466 (.469) .320 .628 .251-1.573
25-29 -.166 (.368) .652 .847 .411-1.743
30-34 .034 (.366) .927 1.034 .505-2.118
35-39 .023 (.367) .949 1.024 .498-2.102
40-44 .042 (.374) .910 1.043 .501-2.172
45-49 .251 (.402) .533 1.285 .584-2.826
N 41751
Cox & Snell R Square 0.118
Nagelkerke R Square 0.159
-2 Log-likelihood 6973.408
R, reference category
Figures in the parentheses are SE of estimates
Source: Author’s calculation

5. Conclusions
This paper examined the effect of socio-economic factors on child malnutrition status
in Uttar Pradesh using data from NFHS-4. This study accounts for maternal level
information such as the educational level of mother, malnutrition of mother (BMI as a
proxy) and age cohort of the mother that appears to cause malnutrition in children.
Among socio-economic determinants, mother’s education beyond elementary level
emerged as a significant factor that influenced stunting, wasting and underweight of
children belonging to the age of five. This result was buttressed by other empirical
studies and their results confirmed that mother’s educational level beyond secondary
level reduces the child nutrition outcomes (Bbaale, 2014).
However, the result was not supportive in case of father’s educational level more than
secondary. It indicated that father’s educational level up to secondary was necessary to
lower the rate of stunting, wasting and underweight in children. The analysis revealed
 Reena KUMARI, Rekha GUPTA  95

that sex of the household, sex of the child and mother’s age cohort no longer affected
the rate of stunting, wasting and underweight of children. However, place of residence
determines a significant and positive effect on nutritional status of the children. It
means if children reside in rural areas, they were more wasted in reference to those who
were residing in urban areas as given the fact that urban settings usually have improved
ailment prevention forces. Recently, Government of India has introduced two major
health programmes, for instance, Ayushman Bharat Pradhan Mantri Arogya Yojana
(PM-JAY) in September 2018, in order to provide primary, secondary and tertiary level
treatments to the poor and vulnerable and secondly, National Rural Health Mission in
April 2005 with the aim of enlightening the accessibility of and access to universal
health care for individuals living in rural areas. The study finds that wealth index is also
more likely to be an imperative factor which exhibits a lower level of child malnutrition.
It also represents if households belonging to the richest quintile of wealth index,
children’s live better quality of life and low mortality.
Based on our findings, the present study suggests that malnutrition in children is the
result of multiple factors. Low rate of female education is the principal factor for child
malnutrition, and this happens due to many socio-cultural reasons. To improve the
status of health and nutritional outcome, Government of India has launched many
programmes, however, more emphasis should also be given to improvement in female
education, employment opportunities to enhance economic status, access to rural
nutrition, awareness about use of contraceptive. Wealth status is another determinant
which indicates high incidence of poverty is a persistent problem affecting the ability to
access and attain an adequate diet.
This estimate shows the scarcity of material possession and unavailability of basic needs
to the poor which results in low demand for energy and protein intake. To improve
nutritional status of children, the state needs to eradicate all forms of poverty and to
provide employment opportunities to the poor.
The study recommends that efforts to improve female education need to be combined
with more specific schemes to progress and better child rearing practices. Also, it is
important to recommend that civic programmes to ensure education to female on how
to feed infants and safe them from infection, and adequate, accessible health services to
avoid and treat infections can collectively reduce malnutrition in children. Our results
show that family using modern of contraceptive is associated with positive child health
outcomes. Therefore, the present study suggests that government, donors, and private
players/NGO can perform an imperative role in improving women’s access to modern
healthcare services by ensuring affordability at the regional level.

References
Ahmed, T., Roy. S. & Alam. N. (2012). Determinants of undernutrition in children
under 2 years of age from rural Bangladesh”, Indian Pediatric, 49: 821–24.
Allen, L.H. & Gillespie, S.R. (2001). What works?: A review of the efficacy and effectiveness of
nutrition intervention. ACC/SCN Nutrition Policy Paper, 19, 27–8.
96 A logistic regression analysis of determinants of child malnutrition in Uttar Pradesh, India

Bangladesh Demographic & Health Survey (2007). Dhaka, Bangladesh and Calverton,
Maryland [USA]: National Institute of Population Research and Training, Mitra and
Associates, and Marco International.
Basu, A. M., & Stephenson, R. (2005). Low levels of maternal education and the
proximate determinants of childhood mortality: a little learning is not a dangerous
thing. Social Science & Medicine, 60(9), 2011-2023.
Bbaale, E. (2014). Maternal education and child nutritional status: evidence from
Uganda, African. Journal of Economic and Management Studies, 5(1), 52-74.
Burroway, Rebekah (2016). Empowering women, strengthening children: a multi-level
analysis of gender inequality and child malnutrition in developing countries. Gender
and Food: From Production to Consumption and After, 22, 117-142.
Caldwell, J. (1979). Education as a factor in mortality decline: an examination of
Nigerian Data, Population Studies, 33(3), 395-413.
Case, A. (2001). Does money protect health status? Evidence from South African
pensions. NBER working paper no. 8495.
Case, A., Lubotsky, D. & Paxson, C. (2002). Economic status and health in child: the
origins of the gradients. American Economic Review, 95(5), 1308-1334.
Choudhury, P.K. (2015). Explaining the role of parental education in the regional
variations in infant mortality in India. Asia & The Pacific Policy Studies, 2(3), 544-572.
Cleland, J. (2010). The benefits of educating women. The Lancet, 376(9745), 933-934.
Das, Sibabrata & Sahoo, Harihar (2011). An investigation into factors affecting child
undernutrition in Madhya Pradesh. Anthropologist, 13(3), 227-233.
Dasgupta, M., Lokshin, M., Gragnolati, M. & Ivaschenko, O. (2005). Improving Child
Nutrition Outcome in India-Can the Integrated Child Development Services Program be More
Effective? World Bank Policy Research Working Paper No. 3647. World Bank,
Washington, DC.
Desai, S., & Alva, S. (1998). Maternal education and child health: is there a strong casual
relationship? Demography, 35(1), 71-81.
Deshmukh, P.R., Sinha, N. & Dongre, A.R. (2013). Social determinants of stunting in
rural area of Wardha, Central India. Medical Journal Armed Forces India, 69, 213-217.
Frost, M. B., Forste, R., & Haas, D.W. (2005). Maternal education and child nutritional
status in Bolivia: finding the links. Social Science & Medicine, 60(2), 395-407.
Glewwe, P. (1999). Why does mother’s schooling raise child health in developing
countries? Evidence from Morocco. Journal of Human Resources, 34(1), 124-159.
Govindasamy, P. & Ramesh, B.M. (1997). Maternal education and utilisation of
maternal and child health services in India. NFHS subject reports, Number 5,
International Institute of Population Sciences, Mumbai.
Grantham-McGregor, S. Cheung, Y.B. Glewwe, P. Richter, L. & Strupp, B. (2007).
Developmental potential in the first 5 years for children in developing countries.
Lancet, 369 (9555): 60-70.
Haddad, L., Alderman, H., Appleton, S., Song, L., & Yohannes, Y. (2003). Reducing
child malnutrition: how far does income growth take Us? The World Bank
Economic Review 17 (1),107–131.
Headey, D., (2013). Developmental drivers of nutritional change: a cross-country
analysis. World Development, 42, 76–88.
 Reena KUMARI, Rekha GUPTA  97

Heltberg, R., (2009). Malnutrition, poverty, and economic growth. Health Economics, 18
(S1), S77–S88.
Horton, S. (1998). Birth order and child nutritional status: evidence from the
Philippines. Economic Development and Cultural Change, 36(2), 341-354.
Mazumdar, S. (2010). Determinants of inequality in child malnutrition in India. Asian
Population Studies, 6(3), 307-333.
Measham, A.R. & Chatterjee, M. (1999). Wasting away: The crisis of malnutrition in
India. Washington, DC: World Bank.
Miller, J. E., & Rodgers, Y. (2009). Mother’s education and Children’s nutritional status:
new evidence from Cambodia. Asian Development Review, 26(1),131-165.
Miller, J.E., & Rodgers, Y.V. (2009). Mother’s Education and Children’s Nutritional
Status: New Evidence from Cambodia. Asian Development Review, 26(1), 131-165.
Mishra, V.K., Lahiri, S., & Luther, N.Y. (1999). Child nutrition in India. National
Family Health Survey Subject Reports 14 (June 1999). Mumbai, India: International
Institute for Population Sciences, and Honolulu, Hawaii, USA: East-West Center.
Mondal, N.I., Hossain, K., & Ali, K. (2009). Factors influencing infant and child
mortality: a case study of Rajshahi district, Bangladesh. Journal of Human Ecology,
26(1), 31-39.
Moore, A.C., Akhter, S., & Aboud, F.E. (2006). Responsive complementary feeding in
rural Bangladesh. Social Science & Medicine, 62, 1917–30.
Nussbaum, M. (2004). Women’s education: A global challenge. Signs: Journal of Women in
Culture and Society, 29(2), 325-355.
Patel, A., Badhoniya, N., Khadse, S., Senarath. U., Agho, K.E., & Dibley, M.J. (2010).
Infant and young child feeding indicators and determinants of poor feeding
practices in India: secondary data analysis of National Family Health Survey 2005-
06. Food and Nutrition Bulletin, 31, 314-33.
Sastry, N., (1997). What explains rural-urban differentials in child mortality in Brazil?
Social Science and Medicine, 44(7), 989-1002.
Sen, A. (1999). Development as freedom. New York, NY: Anchor Books.
Sethuraman, K., Lansdown, R. & Sullivan, K. (2006). Women’s empowerment and
domestic violence: The role of sociocultural determinants in maternal and child
undernutrition in tribal and rural communities in South India. Food and Nutrition
Bulletin, 27(2), 128-143.
Shetty, P.S. & James, W.P.T. (1994). Body mass index: a measure of chronic energy
deficiency in adults. Rome, Italy: Food and Agriculture Organization of the United
Nations, (FAO food and nutrition paper 56).
Smith, L.C., & Haddad, L.J., (2000). Explaining Child Malnutrition in Developing
Countries, Technical report. International Food Policy Research Institute (IFPRI),
Washington, DC .
Subramanyam, M.A., Kawachi, I., Berkman, L.F. & Subramanian, S.V. (2010).
Socioeconomic inequalities in childhood undernutrition in India: Analyzing trends
between 1992 and 2005. PLoS One, 5(6), e11392.
Svedberg, P. (2008). Why malnutrition in shining India persist. 4th Annual Conference
on Economic Growth and Development, New Delhi, 2008.
Tarozzi, Alessandro & Mahajan, Aprajit (2007). Child nutrition in India in the nineties.
Economic Development and Cultural Change. University of Chicago Press, 55(3), 441-486.
98 A logistic regression analysis of determinants of child malnutrition in Uttar Pradesh, India

Thomas, D., Strauss, J., & Henriques, M.H. (1991). How does mother’s education
affect child health? Journal of Human Resources, 26(2), 183-211.
United Nations (2019). Transforming Our World: The 2030 Agenda for Sustainable
Development.
https://sustainabledevelopment.un.org/post2015/transformingourworld.
Waihenya, E.W., Kogi-Makau, W. & Muita, J.W. (1996). Maternal nutritional
knowledge and the nutritional status of preschool children in a Nairobi slum. East
African Medical Journal, 73, 419-423.
Walker, S.P., Wachs, T.D., Gardner, M., Lozoff, B., Wasserman, G.A., Pollitt, E. &
Carter, J.A. (2007) Child development: risk factors for adverse outcomes in
developing countries. The Lancet, 369(9556), 145-157.
Wemakor, A., Garti, H., Azongo, T., Garti, H., & Atosona, A., (2018). Young maternal age
is a risk factor for child undernutrition in Tamale Metropolis, Ghana. BMC Research Note,
11: 877.
WHO (1978). A Growth Chart for International Use in Maternal and Child Health Care.
WHO: Geneva
WHO (2010), Nutrition Landscape Information System (NLIS) Country Profile
Indicators. Geneva, Switzerland.
World Bank (2006), Repositioning Nutrition as Central to Development A Strategy for
Large-Scale Action. Washington DC, USA.

Acknowledgements
This study was conducted by Reena Kumari, and Rekha Gupta

Funding
The authors received no financial support for the research, authorship, and/or
publication of this article.

Declaration of conflicting interests

We declare, on our own responsibility, that there is no conflict of interest in the
production and publication of this article..