The Journal of Nutrition

Community and International Nutrition

Early Childhood Stunting Is Associated with

Poor Psychological Functioning in Late
Adolescence and Effects Are Reduced
by Psychosocial Stimulation1,2
Susan P. Walker,3* Susan M. Chang,3 Christine A. Powell,3 Emily Simonoff,4
and Sally M. Grantham-McGregor5
3
Epidemiology Research Unit, University of the West Indies, Kingston 7, Jamaica; 4Institute of Psychiatry, London SE5 8AF, UK;
and 5Institute of Child Health, London WC1N 1EH, UK

Downloaded from jn.nutrition.org at Pakistan: ASNA Sponsored on March 16, 2017

Abstract
Stunting is associated with deficits in cognition and school achievement from early childhood to late adolescence; how-
ever, there has been little investigation of emotional and behavioral outcomes. The objective of this study was to deter-
mine whether linear growth retardation (stunting) in early childhood is associated with poorer psychological functioning in
late adolescence. The study was a prospective cohort study of stunted and nonstunted children. Participants were iden-
tified at age 924 mo by a survey of poor neighborhoods in Kingston, Jamaica, and a 2-y intervention trial of supple-
mentation and stimulation was conducted in the stunted children. Psychological functioning was assessed at age 17 y in
103 of 129 stunted children enrolled and 64 of 84 nonstunted participants. Anxiety, depressive symptoms, self-esteem,
and antisocial behavior were reported by participants using interviewer-administered questionnaires and attention deficit,
hyperactivity, and oppositional behavior were reported by parent interviews. The stunted participants reported significantly
more anxiety (regression coefficient 3.03; 95% CI 0.99, 5.08) and depressive symptoms (0.37; 95% CI 0.01, 0.72)
and lower self-esteem (21.67; 95% CI 20.38, 22.97) than nonstunted participants and were reported by their parents
to be more hyperactive (1.29; 95% CI 0.12, 2.46). Effect sizes were 0.40.5 SD. Participants who received stimulation in
early childhood differed from the nonstunted group in hyperactivity only. Children stunted before age 2 y thus have poorer
emotional and behavioral outcomes in late adolescence. The findings expand the range of disadvantages associated with
early stunting, which affects 151 million children ,5 y old in developing countries. J. Nutr. 137: 24642469, 2007.

Introduction
Growth retardation or stunting in children in developing coun-
tries occurs primarily as a result of chronic undernutrition and
infectious diseases and affects 30% of children ,5 y old (1).
Stunting is associated with poor development in early childhood
(2) and with deficits in cognition and school achievement com-
pared with nonstunted children through late adolescence (3,4).
Children hospitalized for severe malnutrition in early child-
hood have been reported to have problems with aggressive be-
havior (5), attention deficits (6), and poor social relationships at
school age (7). Although stunted children have been found to
have behavior changes in early childhood, such as less happiness
and more apathy and fussiness (8), there has been little investi-
gation of whether their later emotional and behavioral function-
ing is affected.
1
Supported by a grant from the Wellcome Trust (no. 066088).
2
Author disclosures: S. P. Walker, S. M. Chang, C. A. Powell, E. Simonoff, and
S. M. Grantham-McGregor, no conflicts of interest.
* To whom correspondence should be addressed. E-mail: susan.walker@
uwimona.edu.jm.
2464
Manuscript received 14 March 2007. Initial review completed 10 May 2007. Revision accepted 29 August 2007.
We conducted a prospective cohort study of stunted and non-
stunted Jamaican children beginning at age 924 mo. In the first
2 y, the stunted children participated in a randomized trial of
supplementation and psychosocial stimulation. Each treatment
benefited concurrent development (2). Stimulation had sustained
benefits to cognitive functioning at ages 7, 11, and 17 y (911),
whereas benefits from supplementation were not sustained. Nei-
ther intervention affected teacher- and parent-reported behavior
problems at age 11 y (12). There were no benefits of supplemen-
tation to psychological functioning at age 17 y; however, stim-
ulation had significant benefits, including fewer symptoms of
anxiety and depression and better self-esteem (13).
Stunted children who did not receive stimulation continued
to have poorer levels of cognitive functioning than nonstunted
children (11). Here, we compare emotional and behavioral out-
comes of stunted participants with nonstunted participants at
age 1718 y. As stimulation had significant benefits to psycholog-
ical functioning, we compared the nonstunted participants with
the stunted participants according to whether they received stim-
ulation. There are considerable data showing that many of
the functions measured (symptoms of depression, self-esteem,
0022-3166/07 $8.00 2007 American Society for Nutrition.
attention deficit, and oppositional and antisocial behavior) are
related to poverty; however, general anxiety is sometimes asso-
ciated with better socioeconomic conditions (14). We therefore
measured extensive socioeconomic variables to allow for any dif-
ferences between the groups when assessing the effect of stunting.
Participants and Methods
Initial study. In 19861987, we identified children aged 924 mo by a
house-to-house survey of poor neighborhoods of Kingston, Jamaica (2).
All 129 stunted children [length-for-age , 22 SD of the National Center
for Health Statistics references (15)] identified were assigned to 1 of 4
groups: control, supplementation, stimulation, or both interventions.
The order of group assignment was determined randomly. Supplemen-
tation comprised 1 kg milk-based formula provided weekly (16). Stim-
ulation comprised weekly 1-h home visits by trained community health
workers who conducted play sessions with the mother and child. Both
interventions were provided for 2 y. Thirty-two nonstunted children
(length-for-age . 21 SD) were enrolled matched to the control group for
age, sex, and neighborhood.
The children were reexamined at ages 7 and 11 y. To increase the
power of comparisons between stunted and nonstunted groups, an ad-
ditional 52 nonstunted children were studied from age 7 y. They had
been identified in the original survey when they were aged 924 mo and
fulfilled all inclusion criteria. They lived in the same neighborhoods as
the study children and were comparable to the original group of non-
stunted children in social background (9).
Follow-up at 1718 y. Psychological functioning was assessed when
participants were aged 1718 y. The aim was to assess emotional and
behavioral symptoms and no attempt was made to diagnose specific dis-
orders. Written informed consent was obtained from the participants
and their parents. Ethical approval for the study was given by the Ethics
Committees of the University of the West Indies and the Institute of Child
Health, University College London.
Psychological functioning. Self-esteem was assessed with the How I
Think about Myself questionnaire (17), anxiety symptoms were assessed
with the Manifest Anxiety (What I Think and Feel) Questionnaire (18),
and depressive symptoms were assessed with the Short Mood and Feel-
ings Questionnaire (19). Antisocial behavior was assessed with the Be-
havior and Activities Checklist (20). Questionnaires were administered
in a private interview at our research unit. All were obtained by self re-
port.
Although young people are considered valid informants for most
behaviors by age 17 y, attention deficit and hyperactivity symptoms are
exceptions. Therefore, the Conners Parent Rating Scale (short form)
(21) was administered to the mothers (or primary caregiver if not the
mother) to obtain information on attention deficit. The Conners Scale
also provides scores for cognitive problems/inattention, hyperactivity,
and oppositional behavior.
All questionnaires were piloted and reworded as necessary to ensure
comprehension. This involved replacing words or short phrases not easily
understood with others more familiar to the participants. For example,
the item often I feel sick in my stomach was changed to you often feel
sick/upset in your belly. No items were omitted from any of the ques-
tionnaires. All questionnaires were interviewer administered; therefore,
for 3 of the scales (How I Think about Myself, What I Think and Feel,
and Short Mood and Feelings), which were written to allow self-
administration, we changed portions of the questions from first- to
second-person (i.e. I felt . to You felt.). stimulated stunted group did not participate in the follow-up
Test-retest reliabilities (intraclass correlation coefficient) after a 2-wk
interval, in 18 nonstudy subjects recruited from schools attended by the
study subjects, were $0.7 and internal consistency (Cronbachs a) ranged
from 0.55 to 0.85 (Table 1). Participant interviews were administered by
one interviewer and parent questionnaires by another interviewer. Inter-
viewers were unaware of the participants group. Inter-observer agree-
ment with another trained interviewer in 23 interviews was $97% for all
items.
TABLE 1 Reliability and internal consistency of emotional
and behavioral measures
Measure Items, n Test retest1 (R) Cronbach's a
Anxiety 28 0.92 0.84
Depression 13 0.70 0.83
Self-esteem 8 0.80 0.55
Antisocial behavior 18 0.82 0.68
Attention Deficit 12 0.70 0.85
Cognitive problems/inattention 6 0.81 0.77
Hyperactivity 4 0.74 0.69
Oppositional behavior 6 0.85 0.83
1
Intraclass correlation coefficient, retest after a 2-wk interval, n 18.
Socioeconomic background. Participants were asked about the fre-
quency of hunger because of lack of food in the home during the previous
year. They were also asked to report if they had ever been a victim of
violent crime (e.g. robbery, stabbing, or shooting) and if they had
witnessed violent acts (involving knives, guns, or other weapons) during
Downloaded from jn.nutrition.org at Pakistan: ASNA Sponsored on March 16, 2017
the past year. If yes, they were asked to say how many times they had
witnessed such acts. The homes were visited and information obtained
on water and toilet facilities (each rated on a scale of 16), the number of
persons per room, and the number of household possessions from a list
of 11 items. A housing score was derived by factor analysis of these
variables. The mothers or primary caregivers occupation was recorded.
Mothers verbal intelligence had been assessed previously with the Pea-
body Picture Vocabulary Test (PPVT) when the children were aged 11 y.
Statistical analysis. Outcome variables were screened for normality
and the depressive symptoms and antisocial behavior scales were nor-
malized using square root transformations. Associations between par-
ticipant characteristics and psychological outcomes were determined by
Pearsons product-moment correlation coefficient (age, housing, mothers
PPVT) or Spearman rank correlation (hunger, occupation, and witness-
ing and victim of crime). Gender differences were determined by t test.
An a of P , 0.05 was used to assess significance. Because stimulation but
not supplementation had significant benefits on psychological functioning,
to determine the effect of stunting, the nonstunted group was compared
with the stunted children who had not participated in the stimulation
program (nonstimulated stunted: control and supplement only groups)
and with those who had received stimulation (stimulated stunted: stimu-
lation only and stimulated and supplemented groups). Group was coded
as 2 dummy variables to compare each of the stunted groups with the
nonstunted group. The dummy variables were: nonstimulated stunted 1,
other groups 0; and stimulated stunted 1, other groups 0. Our
interest in this report is on the impact of early childhood stunting. We
report the unadjusted effects of stunting followed by the effects after
adjusting for possible confounders. In multiple regression analyses,
participants age, sex, and social background variables that either dif-
fered between stunted and nonstunted groups at follow-up or were cor-
related with the outcomes were offered stepwise before entering the 2
dummy variables for group. Social background variables offered in the
analyses were hunger, housing score, mothers PPVT and occupation,
father present in the home, and whether the participants had witnessed
or been a victim of violent crime in the past year.
Results
Loss from study. Ten of 65 children (15.4%) from the non-
study, 16 of 64 (25.0%) children from the stimulated stunted
group did not participate, and 20 of 84 (23.8%) from the
nonstunted group did not participate. The proportion of subjects
lost was not significantly different by group. Most of the loss was
due to migration, which accounted for 70% of children lost.
Participants and those who did not take part in the study did not
differ in any enrollment measure (initial age, developmental
Stunting affects psychological functioning 2465
level, height-for-age, home stimulation, mothers verbal intelli-
gence and education, and housing quality), except within the
nonstimulated stunted group, children who did not participate
had younger mothers (P , 0.001) than those assessed.
Participant characteristics. The nonstunted children had
higher birth weight (P , 0.001), height-for-age (P , 0.001),
and better housing (P 0.002) on enrollment than the stunted
groups (Table 2). The nonstunted participants remained taller
than the stunted groups (P , 0.001) at follow-up. The mothers/
caregivers of the nonstunted group had better verbal scores than
the combined stunted groups (P 0.03). There were no other
differences between the nonstunted and stunted groups in any
measured social background variables. The nonstimulated stunted
participants reported more hunger than the stimulated stunted
group (P 0.03).
Psychological functioning. The emotional and behavioral
outcomes are shown in Table 3. Several of the participants
they had witnessed violence (P , 0.001) or had been a victim of
crime (P , 0.05), although it is possible that being a victim or
witnessing violence could be due to the children themselves being
antisocial. Girls reported less antisocial behavior than boys (P ,
0.05) and were reported by their mothers/caregivers to have
fewer problems with attention (P , 0.05). Mothers/caregivers
with better verbal scores reported fewer problems with attention
in their children (P , 0.01) and higher occupation level was
associated with better self-esteem (P , 0.02) and less opposi-
tional behavior (P , 0.02). Poorer housing was associated with
more hyperactivity (P , 0.001), poorer attention (P , 0.01),
and more oppositional behavior (P , 0.004).
Effect of stunting. Multiple regression analyses were used to
determine the effects of stunting on psychological functioning
(Table 4). Nonstunted participants reported significantly fewer
symptoms of anxiety and depression and better self-esteem than
nonstimulated stunted participants. The nonstunted participants
did not differ from the stimulated stunted group for these out-

Downloaded from jn.nutrition.org at Pakistan: ASNA Sponsored on March 16, 2017

characteristics were associated with their psychological func-
tioning. Older adolescents reported better self-esteem (P ,
0.001). Participants who reported experiencing hunger and those
who had been victims of crime reported more anxiety (P , 0.005).
Participants whose fathers lived with them had better self-esteem
(P , 0.05). Participants reported more antisocial behavior if
comes. Parents of the nonstunted group also tended to report less
oppositional behavior than in the nonstimulated stunted group
(P 0.06) but similar to the stimulated stunted group. Both
stimulated and nonstimulated stunted participants were reported
by their parents to be more hyperactive than the nonstunted chil-
dren. Because the childrens behavior might influence the likelihood

TABLE 2 Participant characteristics at enrollment (924 mo) and follow-up (1718 y)1

Nonstimulated stunted, Stimulated stunted, Nonstunted,

Characteristics n 55 n 48 n 642

Enrollment
Male, n, % 32, 41.8 26, 54.2 32, 50.0
Age, mo 18.6 6 4.2 18.9 6 3.9 19.0 6 4.6
Birth weight, kg 2.95 6 0.45 2.80 6 0.52 3.26 6 0.48*
Height-for-age (Z-score) 22.9 6 0.6 23.0 6 0.6 0.1 6 0.6*
Housing rating3 7.5 6 1.8 7.2 6 1.5 8.7 6 1.4y
HOME4 16.7 6 4.6 15.8 6 4.0 17.6 6 3.2
Mothers' age #19 y, n, % 11, 20.0 13, 27.1 5, 21.7
Follow-up
Age, y 17.5 6 0.3 17.5 6 0.3 17.5 6 0.2
Height-for-age (Z-score) 20.7 6 1.0 20.9 6 0.8 0.7 6 0.8*
Housing score5 20.06 6 0.9 0.10 6 1.0 20.03 6 1.1
Mother's PPVT 86.7 6 20.2 86.5 6 23.9 95.6 6 27.6z
Mother's occupation, n, %
None/unskilled 17, 31.5 8, 16.7 14, 21.9
Semi-skilled 23, 42.6 26, 54.2 28, 43.8
Skilled 14, 25.9 14, 29.2 22, 34.4
Father present in home, n, % 17, 32.1 16, 30.2 20, 37.7
Hunger in past year, n, %
Never 29, 52.7 30, 62.5 45, 70.3
,once/wk 12, 21.8 15, 31.3 12, 18.8
$once/week 14, 25.5 3, 6.2 7, 10.9
Witnessed violent crime in past year, n, %
Never 29, 52.7 26, 54.2 27, 42.2
16 times 19, 34.5 15, 31.3 19, 29.7
$7 times 7, 12.7 7, 14.6 18, 28.1
Participant ever victim of crime, n, % 10, 18.2 6, 12.5 12, 18.8
1
Values are means 6 SD or n, %. *Different from stunted groups, P , 0.001. yDifferent from stunted groups, P 0.002. zDifferent from
combined stunted groups, P 0.03. Different from nonstimulated stunted, P 0.03.
2
For enrollment data, n 23.
3
Sum of ratings of toilet and water facilities, crowding, and possessions.
4
Caldwell Home Observation for Measurement of the Environment.
5
Housing score derived from factor analysis of toilet and water facilities, crowding, and possessions.

2466 Walker et al.

TABLE 3 Emotional and behavioral outcomes at age 1718 y mained significant (coefficient 2.57, 95% CI 0.47, 4.67; P
in nonstunted participants and in stunted 0.02). Birth weight was not a significant predictor of any of the
participants who did or did not receive other outcomes.
psychosocial stimulation1

Nonstimulated Stimulated The role of IQ. Most psychosocial problems are more common
Measure2 stunted stunted Nonstunted P4 in children with low IQ (14) and we previously reported that the
stunted children had lower IQ than the nonstunted children (11).
n 55 48 64 IQ scores were significantly correlated with all the psychological
Anxiety 16.4 6 5.7 13.6 6 5.5 13.0 6 6.0 0.004 scores. We therefore examined whether cognitive ability medi-
Depressive symptoms3 7.1 6 0.8 5.0 6 0.8 5.1 6 1.2 0.035 ated the effect of stunting by repeating the multiple regressions
Self-esteem 24.1 6 3.6 25.6 6 4.1 25.8 6 3.8 0.044 entering IQ before the dummy variables for group. The change
Antisocial behavior3 5.3 6 0.7 4.4 6 0.7 4.8 6 0.8 0.429 in the regression coefficients comparing the nonstimulated stunted
Attention deficit 14.3 6 8.6 11.0 6 7.1 11.7 6 8.2 0.085 group and the nonstunted group was tested using the method of
Cognitive problems/inattention 6.7 6 4.7 5.6 6 3.9 5.9 6 4.9 0.464 Freedman and Schatzkin (22). The coefficients decreased (P ,
Hyperactivity 4.7 6 3.4 4.5 6 3.5 3.4 6 3.2 0.075 0.05) for anxiety (24.8%), depressive symptoms (43%), and
Oppositional behavior 7.6 6 4.9 6.0 6 5.0 5.9 6 5.3 0.125 self-esteem (33.1%). The change in the regression coefficient for
1
Values are means 6 SD. For all scales, higher scores indicate poorer psychological hyperactivity was not significant. The difference between stunted
functioning except for self-esteem, where higher scores indicate better self-esteem. and nonstunted children in anxiety remained significant (P ,
2
Participant report using interviewer-administered questionnaires: anxiety, Manifest 0.05), but the groups no longer differed in depressive symptoms

Downloaded from jn.nutrition.org at Pakistan: ASNA Sponsored on March 16, 2017

Anxiety (What I Think and Feel) questionnaire (18); depressive symptoms, Short Mood
and Feelings questionnaire (19); self-esteem, How I Think about Myself questionnaire
(17); antisocial behavior, Behavior and Activities Checklist (20). Attention deficit,
cognitive problems, hyperactivity, and oppositional behavior assessed by parent report
with Conners Parent Rating scale, short form (21).
3
Variables normalized using square root transformation, values shown are square of
the transformed data.
4
For ANOVA by group.
of their being exposed to violence, the analyses were repeated
omitting the covariates victim of crime and witnessing vio-
lent crime. The results did not change.
Early childhood stunting is associated with poor home en-
vironments. Although the HOME scores for the stunted groups
were not significantly lower than that of the nonstunted group,
the scores tended to be lower. We therefore repeated the analyses
offering the HOME score. (These analyses included the stunted
groups and the nonstunted participants enrolled at age 924 mo
only, as HOME scores were not available for nonstunted par-
ticipants enrolled at age 7 y). The HOME was not a significant
predictor of any of the psychological outcomes.
The analyses were repeated entering birth weight. Higher birth
weight was associated with less anxiety (coefficient 21.77,
95% CI 23.52, 20.02; P 0.05). The difference in anxiety
between nonstunted and nonstimulated stunted participants re-
or self-esteem.
Discussion
Children stunted in the first 2 y of life had poorer psychological
functioning in late adolescence than nonstunted participants.
This included more symptoms of anxiety and depression, poorer
self-esteem, and more hyperactive behavior. There was also a
tendency for higher levels of oppositional behavior.
A critical question is whether the association between early
childhood stunting and poor psychological functioning in late
adolescence is causal or whether stunting is a proxy for poor
environments that independently affect psychological outcomes.
The stunted and nonstunted children came from the same neigh-
borhoods and we had measures of their socioeconomic status,
maternal characteristics, and exposure to violence. Several of
these measures were related to the outcomes, but after adjusting
for their effects, significant differences remained between stunted
and nonstunted participants, with effect sizes ranging from 0.37
SD for depressive symptoms to 0.51 SD for anxiety. The HOME
score did not predict psychological functioning; however, we did
not have this measure for the nonstunted participants enrolled at
age 7 y. Therefore, the power of the analyses for the HOME
score was less than for the other measures included. It thus

TABLE 4 Effect of stunting in early childhood on emotional and behavioral outcomes at age 1718 y in participants
who did or did not receive psychosocial stimulation1

Nonstimulated Stunted Stimulated Stunted

unadjusted adjusted for covariates unadjusted adjusted for covariates Covariates in the
Measure B 95% CI B 95% CI B 95% CI B 95% CI adjusted model

Anxiety 3.40 1.31, 5.48*** 2.98 0.85, 4.94** 0.62 1.55, 2.79 0.78 1.30, 2.86 Hunger, victim of crime
Depressive symptoms2 0.42 0.06, 0.77* 0.37 0.01, 0.72* 20.01 0.38, 0.36 20.01 0.38, 0.35 Hunger
Self-esteem 21.65 3.03, 20.27* 21.63 2.94, 20.33* 20.19 1.62, 1.24 20.17 1.52, 1.18 Age, father present
Antisocial behavior2 0.12 0.19, 0.43 0.17 0.13, 0.48 20.10 0.43, 0.22 20.01 0.32, 0.31 Witnessed violence, housing
Attention deficit 2.62 0.31, 5.56 1.89 1.01, 4.78 20.70 3.73, 2.34 21.34 4.33, 1.64 Mother's PPVT, sex
Cognitive problems/inattention 0.81 0.85, 2.49 0.43 1.22, 2.08 20.28 2.00, 1.45 20.61 2.31, 1.09 Mother's PPVT, sex
Hyperactivity 1.31 0.09, 2.54* 1.29 0.12, 2.46* 1.11 0.16, 2.37 1.24 0.03, 2.45* Housing score
Oppositional behavior 1.79 0.06, 3.64 1.76 0.05, 3.57 0.16 1.75, 2.08 0.31 1.57, 2.18 Housing score
1
Regression coefficients represent the difference between the nonstunted group and the stunted groups that received or did not receive stimulation. For the unadjusted model,
2 dummy variables were entered: nonstimulated stunted 1, else 0; stimulated stunted 1, else 0. This approach compares each of the stunted groups with the nonstunted
group. For the adjusted model, covariates were offered stepwise before entering the 2 dummy variables. Covariates offered were participants age and sex, hunger, housing,
mothers PPVT, occupation, father present in home, victim of crime, witnessed violence. *P , 0.05, **P , 0.01, ***P , 0.005.
2
Square root transformation used in analyses.

Stunting affects psychological functioning 2467

remains possible that differences in the quality of the home
environment, or residual confounding from unmeasured aspects
of socioeconomic status, may contribute to differences between
the nonstimulated stunted group and the nonstunted group.
Nonetheless, it is likely that at least part of the differences in
psychological functioning is due to undernutrition in the first 2 y
of life. Increased internalizing problems were also found in
children aged 1114 y treated for iron deficiency anemia in
infancy compared with nonanemic children (23), suggesting that
psychological functioning may also be affected by specific
nutrient deficiencies.
The stunted children had a more negative affect in early
childhood (8) and were more inhibited at 910 y (24). Thus,
differences in affect may have persisted since early childhood.
The stunted participants also had higher conduct disorder at age
11 y (12) and the tendency to higher levels of oppositional
behavior at age 17 y is consistent with this.
We cannot be certain of the mechanism whereby stunting is
associated with poor psychological functioning. Animal research
shows that early undernutrition can affect brain structure and
function, leading to lasting cognitive and emotional effects (25).
Thus, alterations to the brain as a result of childhood undernu-
trition could be 1 mechanism for the observed differences in
psychological functioning. Lower IQ is a well-recognized risk
factor for a range of child psychopathology (26). IQ was lower
in the stunted participants, and when IQ was controlled for in
the analyses, stunting was no longer significantly associated with
depressive symptoms or self-esteem. It is possible that low IQ led
to the children having lower self-esteem and more depressive
symptoms. However, the stunted children had a more negative
affect in early childhood (8). It is therefore also possible that
stunting could have affected cognitive and psychological func-
tioning concurrently.
Unlike the other behaviors that were more prevalent in the
stunted children, hyperactivity was not affected by intervention,
suggesting the effects may be difficult to reverse. In adequately
nourished children, hyperactivity is associated with altered func-
tional anatomy of several areas of the brain, including the frontal
lobe and basal ganglia (27). It is possible that stunting in the first
2 y also affects these areas. Whereas hyperactivity is genetic in
general populations, there is also substantial evidence for longer
term effects of inadequacies in the early rearing environment on
attention and hyperactivity (28).
Raised cortisol levels are associated with depression and
anxiety in adequately nourished subjects (29,30). Alterations in
the stress response were found in a sub-sample of the stunted
children at 910 y of age (24) and could be another mechanism
by which childhood undernutrition might affect later psycho-
logical functioning. We plan to investigate this possibility in the
future.
We previously reported that an early childhood stimulation
program had significant benefits to the stunted childrens psycho-
logical functioning (13). We have shown here that the stunted
children who received stimulation reported levels of anxiety,
depressive symptoms, and self-esteem similar to those of the non-
stunted participants. This is very encouraging and emphasizes
the importance of early interventions. The only detectable re-
maining deficit in the stimulated stunted group was in hyperac-
tive behavior.
There is some evidence of associations between growth in
utero and adult psychological functioning. Exposure to famine
in utero increased the risk of schizophrenia (31) and antisocial
behavior (32), and birth weight was associated with psycholog-
ical distress, depression, and suicidal behavior (3335). In the
2468 Walker et al.
present study, birth weight was related to anxiety but not to any
other psychological outcomes; however, few participants had
birth weights ,2500 g. There is less information relating postna-
tal growth to later psychological functioning. In the UK, weight
at 1 y was related to incidence of suicide (36) and weight gain
from birth to 7 y was related to psychological distress (33).
A limitation to this study is that the instruments used have
not been validated in Jamaica. However, the associations be-
tween the scores of the scales and socioeconomic background,
gender, and IQ were theoretically sensible, the test retest reliabil-
ity was good, and the internal reliability of the scales acceptable.
We made minor modifications to the wording to facilitate better
understanding by the participants and gave the questionnaires
by interview because the participants reading ability was not
always adequate. We did not remove any items and it is likely
that the external validity of the revised instruments would be
similar to that of the original scales; however, it would be de-
sirable to have this confirmed with clinical assessments in the
future.
Downloaded from jn.nutrition.org at Pakistan: ASNA Sponsored on March 16, 2017
We used a conservative approach to examine differences be-
tween the nonstunted and stunted groups, including as covari-
ates any variables that were correlated with the outcomes even if
they did not differ among the groups. To achieve the most par-
simonious model, we included covariates using stepwise anal-
yses. Thus, although the associations found between covariates
and the outcomes are plausible and consistent with the litera-
ture, it is possible that some of these may be due to chance.
Maternal depression has been shown to be associated with
poor growth and undernutrition in children (37,38). Maternal
psychological status, including depression and anxiety, is also an
important contributor to poor adolescent mental health (39,40).
We did not measure maternal affect or depression; however, it is
unlikely to have been the main cause of stunting in this study. In
a subsequent study of undernourished and adequately nourished
Jamaican children (41), mothers of undernourished children
were more likely to report depressive symptoms. However, de-
pression did not predict the likelihood of the child being under-
nourished once socioeconomic factors were considered.
In conclusion, we have previously shown that stunting in
early childhood is associated with deficits in cognition and edu-
cational achievement in late adolescence (11). This study provides
probably the first evidence, to our knowledge, linking stunting
and poor psychological functioning, extending the range of dis-
advantages attributed to early childhood stunting. Deficits were
seen in emotional outcomes and in some aspects of behavior.
Poor psychological functioning is likely to affect the partici-
pants quality of life and may influence their social competence
and parenting. The etiology of stunting may vary between set-
tings with low or high prevalence. Further studies of the asso-
ciation between stunting and later psychological functioning are
needed, particularly from countries where early childhood stunt-
ing is more prevalent than in Jamaica. Considering that 151 mil-
lion children ,5 y old in developing countries are estimated to
be stunted (42), the findings further emphasize that stunting in
early childhood is a serious public health concern.
Acknowledgments
We thank Amika Wright and Sydonnie Shakespeare-Pellington
for conducting the interviews.
Literature Cited
1. United Nations System SCN. Fifth report on the world nutrition situ-
ation. Geneva: SCN; 2004.
2. Grantham-McGregor SM, Powell CA, Walker SP, Himes JH. Nutri-
tional supplementation, psychosocial stimulation, and mental develop-
ment of stunted children: the Jamaican Study. Lancet. 1991;338:15.
3. Daniels MC, Adair LS. Growth in young Filipino children predicts
schooling trajectories through high school. J Nutr. 2004;134:143946.
4. Martorell R, Rivera J, Kaplowitz H, Pollitt E. Long-term consequences of
growth retardation during early childhood. In: Hernandez M, Argente
J, eds. Human growth: basic and clinical aspects. Amsterdam: Elsevier
Science Publishers B.V.;1992. p. 1439.
5. Galler JR, Ramsey F. A follow-up study of the influence of early mal-
nutrition on development: behavior at home and at school. J Am Acad
Child Adolesc Psychiatry. 1989;28:25461.
6. Richardson SA, Birch HG, Grabie E, Yoder K. The behavior of children
in school who were severely malnourished in the first two years of life.
J Health Soc Behav. 1972;13:27684.
7. Richardson SA, Birch HG, Ragbeer C. The behaviour of children at
home who were severely malnourished in the first 2 years of life.
J Biosoc Sci. 1975;7:25567.
8. Gardner JM, Grantham-McGregor SM, Himes J, Chang S. Behaviour
and development of stunted and nonstunted Jamaican children. J Child
Psychol Psychiatry. 1999;40:81927.
9. Grantham-McGregor SM, Walker SP, Chang SM, Powell CA. Effects of
early childhood supplementation with and without stimulation on later
development in stunted Jamaican children. Am J Clin Nutr. 1997;66:
24753.
10. Walker SP, Grantham-McGregor SM, Powell CA, Chang SM. Effects of
growth restriction in early childhood on growth, IQ, and cognition at
age 11 to 12 years and the benefits of nutritional supplementation and
psychosocial stimulation. J Pediatr. 2000;137:3641.
11. Walker SP, Chang SM, Powell CA, Grantham-McGregor SM. Effects of
early childhood psychosocial stimulation and nutritional supplementa-
tion on cognition and education in growth-stunted Jamaican children:
prospective cohort study. Lancet. 2005;366:18047.
12. Chang SM, Walker SP, Grantham-McGregor S, Powell CA. Early child-
hood stunting and later behaviour and school achievement. J Child
Psychol Psychiatry. 2002;43:77583.
13. Walker SP, Chang SM, Powell CA, Simonoff E, Grantham-McGregor
SM. Effects of psychosocial stimulation and dietary supplementation in
early childhood on psychosocial functioning in late adolescence: follow-
up of randomised controlled trial. BMJ. 2006;333:472.
14. Goodman R, Scott S. Child psychiatry. Oxford: Blackwell Science;
1997.
15. Hamil P, Drizd T, Johnson C, Reed R, Roche A. Growth curves for children,
birth-18 years, 165 (DHEW # 781650) Vital and Health Statistics Series
11. Hyatsville (MD): National Center for Health Statistics; 1977.
16. Walker SP, Powell CA, Grantham-McGregor SM, Himes JH, Chang
SM. Nutritional supplementation, psychosocial stimulation, and growth
of stunted children: the Jamaican study. Am J Clin Nutr. 1991;54:6428.
17. Warr P, Jackson P. Self-esteem and unemployment among young workers.
Trav Hum. 1983;46:35566.
18. Reynolds CR, Richmond BO. What I think and feel: a revised measure
of childrens manifest anxiety. J Abnorm Child Psychol. 1978;6:27180.
19. Costello EJ, Angold A. Scales to assess child and adolescent depression:
checklists, screens, and nets. J Am Acad Child Adolesc Psychiatry. 1988;
27:72637.
20. Olweus D. Persistence and prevalence in the study of antisocial behav-
iour: definitions and measurement. In: Klein MW, ed. Cross-national re-
search in self reported crime and delinquency. Dordrecht (The Netherlands):
Kluwer; 1989.
21. Conners CK, Sitarenios G, Parker JD, Epstein JN. The revised Conners
Parent Rating Scale (CPRS-R): factor structure, reliability, and criterion
validity. J Abnorm Child Psychol. 1998;26:25768.
22. MacKinnon DP, Lockwood CM, Hoffman JM, West SG, Sheets V. A
comparison of methods to test mediation and other intervening variable
effects. Psychol Methods. 2002;7:83104.
23. Lozoff B, Jimenez E, Hagen J, Mollen E, Wolf AW. Poorer behavioral
and developmental outcome more than 10 years after treatment for iron
deficiency in infancy. Pediatrics. 2000;105:E51.
24. Fernald LC, Grantham-McGregor SM. Stress response in school-age
children who have been growth retarded since early childhood. Am
J Clin Nutr. 1998;68:6918.
25. Strupp BJ, Levitsky DA. Enduring cognitive effects of early malnutri-
tion: a theoretical reappraisal. J Nutr. 1995;125:S222132.
26. Rutter M. The role of cognition in child development and disorder. Br
J Med Psychol. 1987;60:116.
27. Swanson J, Casellanos X, Frith U, Pennington B, Shafer D, Spitzer M,
Spence MA. Developmental psychopathology. Dev Sci. 2001;4:34558.
28. Croft C, Beckett C, Rutter M, Castle J, Colvert E, Groothues C,
Hawkins A, Kreppner J, Stevens SE, et al. Early adolescent outcomes of
institutionally-deprived and non-deprived adoptees. II: language as a
protective factor and a vulnerable outcome. J Child Psychol Psychiatry.
2007;48:3144.
29. de Kloet ER, Vreugdenhil E, Oitzl MS, Joels M. Brain corticosteroid
Downloaded from jn.nutrition.org at Pakistan: ASNA Sponsored on March 16, 2017
receptor balance in health and disease. Endocr Rev. 1998;19:269301.
30. Michael RP, Gibbons JL. Interrelationships between the endocrine
system and neuropsychiatry. Int Rev Neurobiol. 1963;5:243302.
31. Susser ES, Lin SP. Schizophrenia after prenatal exposure to the Dutch
Hunger Winter of 19441945. Arch Gen Psychiatry. 1992;49:9838.
32. Neugebauer R, Hoek HW, Susser E. Prenatal exposure to wartime fam-
ine and development of antisocial personality disorder in early adult-
hood. JAMA. 1999;282:45562.
33. Cheung YB, Khoo KS, Karlberg J, Machin D. Association between
psychological symptoms in adults and growth in early life: longitudinal
follow up study. BMJ. 2002;325:749.
34. Gale CR, Martyn CN. Birth weight and later risk of depression in a
national birth cohort. Br J Psychiatry. 2004;184:2833.
35. Mittendorfer-Rutz E, Rasmussen F, Wasserman D. Restricted fetal
growth and adverse maternal psychosocial and socioeconomic condi-
tions as risk factors for suicidal behaviour of offspring: a cohort study.
Lancet. 2004;364:113540.
36. Barker DJ, Osmond C, Rodin I, Fall CH, Winter PD. Low weight gain
in infancy and suicide in adult life. BMJ. 1995;311:1203.
37. Harpham T, Huttly S, De Silva MJ, Abramsky T. Maternal mental health
and child nutritional status in four developing countries. J Epidemiol
Community Health. 2005;59:10604.
38. Patel V, DeSouza N, Rodrigues M. Postnatal depression and infant
growth and development in low income countries: a cohort study from
Goa, India. Arch Dis Child. 2003;88:347.
39. Beardslee WR, Versage EM, Gladstone TR. Children of affectively ill
parents: a review of the past 10 years. J Am Acad Child Adolesc Psy-
chiatry. 1998;37:113441.
40. Hammen C, Brennan PA. Severity, chronicity, and timing of maternal
depression and risk for adolescent offspring diagnoses in a community
sample. Arch Gen Psychiatry. 2003;60:2538.
41. Baker-Henningham H, Powell C, Walker S, Grantham-McGregor S.
Mothers of undernourished Jamaican children have poorer psychosocial
functioning and this is associated with stimulation provided in the
home. Eur J Clin Nutr. 2003;57:78692.
42. Grantham-McGregor S, Cheung YB, Cueto S, Glewwe P, Richter L,
Strupp B. Developmental potential in the first 5 years for children in
developing countries. Lancet. 2007;369:6070.
Stunting affects psychological functioning 2469