Educational Benefits of the United States

Schooi Feeding Program:

A Criticai Review ofthe Literature
ERNESTO POLLITT, PHD, MITCHELL GERSOVITZ, MS, AND MARITA GARGIULO, BS
Abstract: In the Child Nutrition Act of 1966 it was
stated that educational progress was an objective of
the United States School Feeding programs. In spite of
this fact no serious attempt has ever been made to
evaluate whether this objective has been met; the few
evaluations that have been conducted lack scientific
rigor. As a whole the studies fail to provide a strong
basis from which to make valid inferences regarding
the long-term effects of the feeding program on school
achievement and adaptation. Studies that have fo-
cused on the short-term effects of hunger or morning
feeding suggest that the provision of breakfast may
both benefit the student emotionally and enhance his
capacity to work on school type tasks. (Am. J. Public
Health 68:477-481, 1978)
The involvement of the United States federal govern-
ment in feeding children in schools may be traced back to the
Agricultural Adjustment Act (P.L. 74-310) of 1935. Under
this provision, the government distributed surplus meat,
dairy products, and wheat to needy families and schools. It
was in 1946, however, with the promulgation ofthe National
School Lunch Program (P.L. 79-396), that the government
decided to institutionalize feeding supplementation within
schools throughout the United States.
Later, the Child Nutrition Act (CNA) of 1966 (P.L. 80-
642) was passed in order to assume control over Breakfast
and Special Milk Programs, Summer and Childcare Pro-
grams, and Maternal and Infant Feeding Programs. The ob-
jectives of this Act, as stated in Section 2, are as follows:
"In recognition ofthe demonstrated relationship between
food and good nutrition and the capacity of children to
develop and learn, based on the years of cumulative suc-
cessful experience under the national school lunch pro-
gram with its significant contributions in the field of ap-
plied nutrition research, it is hereby declared to be the
policy of Congress that these efforts shall be extended,
expanded, and strengthened under the authority of the
Secretary of Agriculture as a measure to safeguard the
health and well-being of the Nation's children and to en-
courage the domestic consumption of agricultural and oth-
er foods, by assisting States, through grants-in-aid and
other means, to meet more effectively the nutritional
needs of our children."
(P.L. 89-642, 11 October 1966, 80 Stat. 885-890.)
From the Department of Nutrition and Food Science, Massa-
chusetts Institute of Technology, Cambridge, MA 02139. Address
reprint requests to Dr. Pollitt. This paper, submitted to the Journal
July 25, 1977, was revised and accepted for publication November
10, 1977.
This review attempts to determine whether educational
progress as an objective ofthe U.S. School Feeding Program
has, in fact, been met. This paper presents a selective review
of the literature bearing on how the behavior of students in a
school setting is affected by short-term hunger, or by school
feeding programs. The inclusiveness of the review attempts
to compensate for the relative lack of evaluative data on the
specific behavioral effects of the school lunch and breakfast
programs. Looking at studies related to, but not directly fo-
cusing on, the issue at hand may prove useful in understand-
ing the eifects of such programs.
Broadening the limits of the review, however, does not
increase the number of scientifically sound studies. It is im-
portant to note at the outset that this area of research is char-
acterized by an absence of specific hypotheses, ambiguity in
the definition of variables, a lack of data on the validity and
reliability ofthe measures used, and a lack of consideration
of relevant moderating variables. These methodological
weaknesses are the strongest evidence for the need for eval-
uative research.
M. S. Read published a review in 1973, entitled "Malnu-
trition, Hunger and Behavior",' a first attempt at forcing
some order into the few available studies. In contrast to
Read's work, this study focuses on investigations conducted
in the United States or other developed countries and makes
no attempt to relate the data obtained within school settings
to the literature on protein-calorie malnutrition and behavior
in developing countries.
The studies are subdivided into two basic categories:
short-term behavioral effects of morning feedings and hun-
ger, and long-term efi'ects (e.g., on school performance) of
school feeding programs. Before beginning the review, it is
AJPH May 1978, Vol. 68, No. 5 477
POLLITT, ET AL.
necessary to define the term hunger as it is used in this re-
port. Currently, there is no universally accepted definition of
hunger because of its multifactorial complexity, which in-
cludes not only metabolic and neurophysiologic factors but
also an emotional component. Moreover, the manifestations
of hunger are likely to interact with the nutritional history
and status of the individual as well as with the frequency
with which it has been experienced. As such, it has been
extremely difficult to quantify hunger, or even to claim that it
is a uniform psychobiological phenomenon across the human
species.
For our purposes, it is not necessary to define the bio-
logical dimensions of hunger. None of the studies reviewed
touch on such issues. It is important to note, however, that
whatever the eifects of hunger on the behavior of school chil-
dren may be, they are surely not mediated by changes in
neural structure. Any behavioral eifects are likely to be asso-
ciated with short-term metabolic and neurohumoral changes.
Short-term Behavioral Effects of Morning
Feeding and Hunger
In this section, the studies focusing on the short-term
behavioral eiFects of taking or not taking breakfast or of hav-
ing a mid-morning snack are reviewed. The section also in-
cludes research on the eifects of the same hunger-satiety
variables on physical activity measures, e.g., endurance.
These data give some insight into the eifects in the school
setting of not eating breakfast.
In 1931, Laird, Levitan, and Wilson^ postulated that
hunger pangs and nervousness in children were related. To
test this hypothesis, 48 children from grades 1, 3, and 5 were
studied for two weeks. Nervous children, selected on the
basis of a behavior checklist completed by the teacher, were
divided into three groups: those receiving no special feeding
(control group, played with toys), those receiving milk, and
those receiving milk plus a calcium supplement. The feeding
was administered at 9:30 am, and teachers were unaware of
the experimental groupings. It was stated that a 6 percent
mean reduction in nervousness occurred with milk feeding
over a two-week period. The teachers reported improve-
ments in behavior, such as, "less abstracted and more wide
awake, more careful and less slovenly in thinking, and less
easily fatigued." Nevertheless, 50 percent of the milk-fed
group showed no improvement or were worse at the end of
the two-week period.
Laird, et al.,^ concluded that nervousness in elementary
school children is related to hunger, and that mid-morning
feedings of milk reduced nervousness. They also concluded
that milk plus a calcium supplement decreased nervousness
further, although no indication of baseline calcium nutriture
was mentioned.
Criticism of this study is difficult because the reported
results make little use of any statistical analysis. Further-
more, the idiosyncrasies of the behavioral terminology do
not fit any theoretical framework and suggest the diificulties
the authors must have had in explaining their own data. This
problem is well exemplified by the categores of behaviors.
e.g., "repulsive bearing and physique," "mentally lethar-
gic," or "average qualities of masculineness."
Keister^ studied the efFects of a mid-morning adminis-
tration of fruit juice on hyperactivity, withdrawal, hostile be-
havior, and nervous habits of 133 childrcin 27-60 months of
age attending a nursery school. Each chid was studied four
times during the yeartwice when receiving fruit juice and
twice when receiving water. The behaviors were assessed by
observation of each child for a 30-secon J interval after the
10:00 am feeding of fruit juice or water.
The results indicate that the target tiehaviors appeared
less frequently among children receiving fruit juice than
among those receiving water. There wen; no significant age
differences. Males, however, showed a greater reduction in
the incidence of "negative" behaviors wilh administration of
fruit juice than females.
Matheson^ assessed the value of a mid-morning orange
juice feeding for 100 fifth grade students irom three different
schools. The study was conducted over a 10-day period, and
performance on arithmetic (addition) and letter symbol de-
coding tests was the outcome variable studied. Because the
children acted as their own controls, they were exposed to
an experimental and a control situation on diiferent days.
The 10:30am orange juice supplementation was associated
with significantly better performance at 9:15, 10:30, and
11:45am on tasks of decoding and addition. Tests taken after
the orange juice feeding at 10:30 showed ihe most significant
differences with respect to decoding tasks.
Matheson-* found that the time at which the arithmetic
or decoding task was given (e.g., 9:15, 10:30, 11:45am) did
not significantly affect the performance of children whose
usual breakfast intake was poor or good (rriid-moming orange
juice feeding was not involved in these comparisons). How-
ever, he did not obtain breakfast intake data on the day of
testing; the quality of usual breakfast intake was determined
by a three-day written food record collected several weeks
after the experiment had been conducted.
Dwyer, Elias, and Warren'^ studied the eifects of an in-
stant breakfast (liquid meal) on 139 males in the first grade.
The children were tested individually or tasks of attention
(slow tapping test, digit test, block test) and for short periods
on eye gaze (to assess attention maintenance within the
classroom). One-half of the pupils were led the breakfast in
the morning and one-half in the afternoor. The investigators
found no between-group differences in p;rformance on any
of the tasks of attention measured between 9:30am and
noon. Although home breakfast intake among the control
group was obtained by dietary recall on the day of testing,
results were reported in terms of "sporadic" breakfast eat-
ers and "always eats breakfast" rather than in terms of in-
take of particular nutrients (e.g., calories] on the day of test-
ing.
A series of studies has been conducted to assess the ef-
fects of different breakfast conditions on physical perform-
ance in children. In one study, by Tuttlc et al.," J2- to 14-
year-old boys alternated between periods of eating cereal
and milk for breakfast and no breakfast for 17 weeks. The
boys' total daily nutrient intake was kept constant. Six cate-
gories of physiologic responses were tested in the late morn-
478 AJPH May 1978, Vol. 68, No. 5
U.S. SCHOOL FEEDING PROGRAM
ing, and the results indicated that the omission of breakfast
had no efiFect on neuromuscular tremor magnitude, choice
reaction time, maximum grip strength, or grip strength en-
durance. Nevertheless, by both individual and group means,
maximum work rate and maximum work output, as mea-
sured by a bicycle ergometer, were lower when breakfast
was omitted.
These same investigators" designed other experiments
to determine the relationship between physiologic responses
and a variety of breakfast types. Twenty-five boys were in-
volved in these tests. The authors concluded that there was
no significant diflFerence in physiologic response in the late
morning between subjects consuming cereal and milk and
those consuming bacon and eggs for breakfast.
Arvedson and associates^ reported that, in the early
1950s, it was believed that breakfast should provide one-
quarter of the total daily protein and energy intake to ensure
maximum physical and mental efficiency in the late morning
hours. To test this assumption, Arvedson and her colleagues
studied 203 children, ages 7-17 years, from various Stock-
holm schools. They found that, in relation to age-group al-
lowances for daily intake, only one-third of these children
consumed a breakfast containing 25 percent of their daily
protein and caloric allowances. To determine whether this
low intake had any efifects on physical capacity, the investi-
gators studied 40 boys, ages 11-17 years. The students were
divided into two groups of similar size, age means, and
ranges. Four breakfasts were designed and exchanged on an
isocaloric basis during the four weeks of the experiment.
(There were two high-carbohydrate breakfasts, one 400 calo-
ries and the other 560 calories [60 percent carbohydrate] and
two high-protein breakfasts, one 400 calories and the other
560 calories [35 percent protein].)
Work tests involving a bicycle ergometer were given
3-3 V2 hours after breakfast and immediately after a finger-
prick test for blood glucose. The authors were unable to find
difiFerences in physical capacity according to the type of
breakfast eaten, although blood glucose was significantly
higher after a high-protein breakfast (88 mg/100 ml) than af-
ter a high-carbohydrate breakfast (79 mg/100 ml), as deter-
mined by a t-test (P < 0.001). Blood glucose for the no-
breakfast condition was 84 mg/100 ml.
In conclusion, the review of studies on the short-term
effects of hunger and feeding does not yield a uniform set of
data. Two researchers examined emotional dimensions of
behavior, two others focused on cognitive components, and
the remaining two concentrated on measures of physical ac-
tivity. The studies on emotionality suggest (with some con-
sistency in the data) that, at least among preschoolers and
children up to the fifth grade, a morning snack or juice may
be beneficialin very general termsto the recipients.
However, it is not possible to specify from the data what
these benefits really are. The researchers use vague terms,
such as "nervousness" or "hyperactivity," and do not pro-
vide clear operational definitions of such variables. Thus, no
conclusive inferences can be drawn.
In connection with the studies on the cognitive com-
ponents of behavior, there is some discrepancy between the
data of Dwyer, et al.,^ and those of Matheson.'' The former
found that breakfast had no detectable eflFect on attention,
whereas the latter observed beneficial efiFects on an arith-
metic and a decoding task. Matheson concluded that "stu-
dents score higher on school type tasks undertaken shortly
after food is given than when no food is given'"* (p. 45).
It is not clear why these two studies yielded con-
tradictory results. One possibility is that they may have
tapped diflFerent mental abilities with diflFerent sensitivities to
the nutrition variable. It is also conceivable that the home
intake difiFered between the populations used in the two stud-
ies. However, such possibilities must remain at the hypo-
thetical level, since the available data are insufficient to clari-
fy the issue. What can be said is that Matheson's study,''
which is one ofthe best in the group, supports the contention
that morning food supplementation in school brings about
some beneficial eflFects on the children's performance in
school-type tests.
In connection with the measures of physical activity,
there again seems to be a contradiction between the two
studies reviewed. It must be noted, however, that one evalu-
ation** involved a breakfast-no breakfast condition, whereas
the other study^ compared the eflFects of various breakfasts
having diflFerent caloric, carbohydrate, and protein levels.
From all the data obtained, perhaps the most meaningful
finding in terms of our present concerns is that the omission
of breakfast interfered with the children's maximum work
rate and output.
All told, the data give some indications that short-term
hunger (due to lack of breakfast) may have some adverse
efiFects on emotional behavior, arithmetic and reading ability,
and physical work output as measured by an ergometer.
School Lunch and Breakfast and Behavior
In this section, the few reports that present data from
evaluations of the long-term behavioral efiFects of school
feeding programs are reviewed. The investigations included
in this section looked for eflFects over months or years. In
contrast to the short-term efiFects reviewed previously, these
long-term eflFects are more meaningful in terms of education-
al progress or successful adaptation to the school setting and
are likely to be enduring.
Lininger* studied the efiFects of a school milk program
on scholastic progress over a two-year period among 4,133
"undernourished" school students ages 6 to 16 years. The
children were enrolled in special health classes in which milk
drinking was one of many practices used to improve their
health. The index of scholastic progress was obtained from
teachers' subjective comments. Over the period of the
study, 45 per cent ofthe children receiving milk improved in
"scholarship," whereas only 24 per cent of those not receiv-
ing milk improved. The teachers, however, presumably
knew which children were and which were not receiving
milk. Thus, for obvious reasons, it is impossible to deter-
mine whether the efiFects reported were caused by the milk
supplement or by the teacher's expectations.
In a year-long study, Kreitzman" looked at attendance
plus school grades in order to determine the efiFects of break-
AJPH May 1978, Vol. 68, No. 5 479
POLLITT, ET AL.
fast programs. His investigation involved third and fifth
grade children from two schools in Atlanta, Georgia, who
lived in a government housing project. One school began its
breakfast program in January, and the other had no breakfast
program. At the end of the school year, there was no dif-
ference between the two third grades in achievement test
scores. This finding, however, may have been related to a
supplementary educational program that was operating for
the third graders in the control school. The fifth graders in
the experimental school did as well or better on every seg-
ment ofthe achievement test than the control group. Kreitz-
man's report of the evaluation he conducted is better de-
scribed as a journalistic efiFort than as a scientific paper. He
gives a cursory description of the methods used, does not
report any statistical treatment of the data collected, and
fails to discuss apparent discrepancies in his results as pre-
sented in a graph. It is unfortunate that we have to use this
type of material in order to make inferences about the pos-
sible efiFects of the feeding program.
Lieberman, et al.,'" studied the efiFects of a breakfast
program among low-income black ghetto children in grades
3-6 over a school year. A breakfast-program school
(N = 281) was compared with an adjacent, non-breakfast-
program school (N = 300). Five psychological tests were ad-
ministered that reportedly measured ability to concentrate,
remember, think abstractly, and work in a classroom. The
authors concluded that the school breakfast program had no
long-term eflFect on performance on psychological tests. It is
noted, however, that the children in both groups were origi-
nally well nourished (reported by dietary evaluation, phys-
ical examination, and anthropometric evaluation). There-
fore, the assumption that malnutrition (defined by the stated
parameters) and low-socioeconomic status go hand in hand
was not upheld. Given the fact that the children were origi-
nally well nourished, one would not expect a breakfast pro-
gram to enhance tests based on the assumption that nutri-
tional status would be improving. Furthermore, 52 per cent
ofthe children participated in less than 35-54 per cent ofthe
breakfasts.
Fellers" studied the efiFects of a breakfast program on
198 tenth grade students on school grades and drop-out rate
over a school year. School grades were used as an index of
achievement. The results showed that participants and non-
participants had similar final grades and similar drop-out
rates. However, as with the Lieberman paper,'" the degree
of participation was less than maximal, and Fellers failed to
consider participation as a moderating variable.
Tisdall, et al.,'^ evaluated over 200 school lunch partici-
pants and non-participant controls, ages 5 Vi to 10 years,
over a three-year period. Indices of school achievement uti-
lized were: school marks as recorded by teachers, scores on
intelligence tests, and scores on objective tests of reading
and arithmetic. Participants and non-participant controls
were reportedly matched exactly for sex, school grade, and
by a medical examination and "as closely as possible" for
classroom, age, height, weight, economic status, dental con-
ditions, mental ability, and school achievement.
The authors concluded that there was no evidence to
indicate the school lunch program accelerated mental or edu-
cational development. However, statistical analyses were
not presented in the report, thus making any critique of the
conclusions diflficult. It is also not quite (;lear whether the
authors controlled for degree of student piirticipation in the
school lunch program with respect to tht^ school achieve-
ment comparisons. The combined impact of the breakfast
and lunch programs was not studied.
Pinkus"* compared breakfast habits, school perform-
ance, and hunger-related behaviors between fourth grade
children participating in the National Breakfast Program and
children ofthe same grade attending non-participating
schools. The eight schools involved were matched for size of
fourth grade, predominant race, and numl^er of teachers in
grades 1-6. All schools were located in Louisiana and met
the government criteria for participation in the National
Breakfast Program. Attendance records from March to Sep-
tember were used in the study. In March, data collection
involved the responses of teachers, parents, and nearly 200
pupils to questionnaires concerning many aspects of break-
fast and breakfast programs. The investigators found that
breakfast was skipped by nearly one-quarter of the pupils in
the schools with no Breakfast Program, and by less than a
tenth ofthe children in participating schools. More students
from schools with a Breakfast Program ate a basic breakfast
than the others. Educational level of the mother was not
found to be related to the children's eatinj; habits.
All pupils recorded the number of i;imes in the past
month they had experienced nausea, headaches, stom-
achaches, and coughs. There was a significant diflFerence be-
tween experimental and control subjects in the reports of
nausea (P < 0.05), stomachaches (P < 0.01), and coughs
(P < 0.01). The students also reported the number of times
during the month they cried, were angry, had been asked to
pay attention or to stop misbehaving. Parsnts and teachers
recorded incidences ofthe same behaviors for one week. Al-
though no significant diflFerences between Ihe groups of chil-
dren were noted by their parents or teachers, a higher pro-
portion of the pupils in non-participating schools reported
having all problems more frequently (e.g., misbehaving was
reported twice as frequently in non-participating schools).
Neither absenteeism nor scholastic achievement, as deter-
mined by the number of D's and F's recorded for one month,
were significantly diflFerent. However, the paper did not in-
dicate whether or not teachers in all schools graded on a
curve. If this were the case, no diflFerences in grades between
the classes should have been expected.
Koonce'^ conducted a study in Anchorage, Alaska, to
determine whether any difiFerences existed between children
who received both breakfast and lunch at school and those
who received lunch only. The students were in the first
through third grades. Those serving as controls received on-
ly lunch. Attendance was scored for each child by sub-
tracting the number of absences in September to December
1970, from the number in September to December 1969, the
year before the program was initiated. No significant dif-
ference was found in attendance. The author suggested that
perhaps the three-month period was inaiiequate for com-
parative purposes.
Increases in weight and height were calculated over the
480
AJPH May 1978, Vol. 68, No. 5
U.S. SCHOOL FEEDING PROGRAM
period from September to December 1970. A trend indicated
greater weight gain in the breakfast and lunch group, but it
was not significant.
This experiment also included a well-organized study of
classroom activity. Teachers rated each student on "general
disposition." Rating sheets were collected in September and
December. The paper made no statement concerning the
teachers' knowledge of the pupils' meal status. The author
concluded that the children who received both breakfast and
lunch obtained higher ratings in all areas assessed than the
children receiving only lunch. The score differentials were
statistically significant, suggesting a better school "dis-
position" on the part of the experimental group.
In summary, like the studies of short-term effects of
morning hunger and feeding, the work reviewed on the long-
term effects of school feeding do not yield a uniform set of
data. In the long-term studies, however, the investigators
had similar notions of what behavioral variables to study.
Most of them focused on school grades, achievement, and
attendance.
The data show that, while two investigators*-" found a
beneficial effect of school breakfast on school performance,
the other five investigators'""'** failed to detect such dif-
ferences. It is impossible, however, to identify the reasons
behind such contradictory findings because most of these re-
ports present only brief descriptions of their samples and
methods. Nonetheless, some of the data suggest that there
are many important moderating variables (e.g., degree of
participation in the feeding programs, teachers' expectations
of success, food intake on the day achievement tests are ad-
ministered) that must be measured and controlled in order to
have a valid assessment of the nature of the correlation be-
tween feeding and achievement. The differences in the de-
signs and the samples of the various studies may also ac-
count for some of the different results obtained.
The study by Lieberman, et al.,' illustrates how the na-
ture of the sample may determine the nature of the results. In
that study, there was no detectable difference between the
experimental and control subjects in a series of school-per-
formance-related measures. However, the recipients of the
program were well nourished before the initiation of the
treatment, and, therefore, the food program may not have
brought additional nutritional benefit. If that were the case, it
is not surprising that the nutritional input brought no addi-
tional educational benefit. An important question that re-
mains unanswered is whether a program that starts with
poorly nourished recipients and brings about nutritional im-
provements would fail to benefit the educational status of the
children.
As a whole, the studies fail to provide a strong basis
from which to make valid inferences regarding the long-term
effects of the feeding programs on the school achievement
and adaptation of the children. It is immediately apparent
that the studies have failed to monitor closely many impor-
tant moderating variables and that their methods were not
sound. Conversely, the studies that focused on the short-
term effects of hunger or morning feeding suggest that the
provision of breakfast may both benefit the student emotion-
ally and enhance his/her capacity to work on school-type
tasks.
REFERENCES
1. Read, M. S. Malnutrition, hunger and behavior. J. Am. Dietet.
Assoc. 63:379-385, 1973.
2. Laird, D. A., Levitan, M., and Wilson, V. A. Nervousness in
school children as related to hunger and diet. Med. J. Rec.
134:494-499, 1931.
3. Kiester, M. Relation of mid-morning feeding to behavior of nur-
sery school children. J. Am. Dietet. Assoc. 26:25-29, 1950.
4. Matheson, N. E. Mid-morning nutrition and its effects on
school type tasks. PhD Dissertation, University of Southern
California, 1970.
5. Dwyer, J. T., Elias, M. F., and Warren, J. H. Effects of an ex-
perimental breakfast program on behavior in the late morning.
Department of Nutrition, Harvard School of Public Health.
Cambridge, MA. Unpublished, 1973.
6. Tuttle, W. W., Daum, K., Larsen, R., Salzano, J., and Roloff,
L. Effect on school boys of omitting breakfast: Physiologic re-
sponses, attitudes and scholastic attainment. J. Am. Dietet. As-
soc. 30:674-677, 1954.
7. Arvedson, I., Sterky, G., and Tjernstrom, K. Breakfast habits
of Swedish school children. J. Am. Dietet. Assoc. 55:257-261,
1969.
8. Lininger, F. Relation of the use of milk to the physical and scho-
lastic progress of undernourished school children. Am. J. Public
Health 23:555-560, 1933.
9. Krietzman, S. W. Evaluation of the Craddock breakfast study.
Atlanta School of Dentistry, Emory University, Atlanta, GA.
Unpublished, 1973.
10. Lieberman, H. M., Hunt, 1. F., Coulson, A. H., Clark, V. A.,
Swendseid, M. E., and Ho, L. Evaluation of a ghetto school
breakfast program. J. Am Dietet. Assoc. 68:132-138, 1976.
11. Fellers, S. A. A study of the effects of breakfast on scholastic
attainment, drop-out rate and knowledge of nutrition. PhD Dis-
sertation, Boston University, Boston, MA, 1967.
12. Tisdall, F. F., Robertson, E. C, Drake, G. H., Jackson, S. H.,
Fowler, H. M., Long, J. A., Brouha, L., Ellis, R. G., Phillips,
A. J., and Rogers, R. S. Canadian Red Cross school meal
study. Can. Med. Assoc. J. 64:477^89, 1951.
13. Pinkus, M. S. A study of pupil breakfast habits and behavioral
patterns in certain Louisiana elementary schools following im-
plementation of the National Breakfast Program. Master's
Thesis, Louisiana State University, 1970.
14. Koonce, T. M. Does breakfast help? School Food Service Jour-
nal 26:51-54, 1972.
ACKNOWLEDGti/IENTS
This study is part of a report prepared at the request of the Ford
Foundation of Modes of Evaluating the Effects of the United States
School Feeding Program on the Behavior of the Recipients.
AJPH May 1978, Vol. 68, No. 5 481