![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
University of North Carolina at Chapel Hill, Chapel Hill, NC 27516
3To whom correspondence should be addressed. E-mail: mchris{at}email.unc.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT SUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
7% for boys and 9% for girls. Improving early childhood nutrition may have long-lasting educational benefits, increasing the likelihood of high school completion in developing countries.
KEY WORDS: • growth • height • nutrition • school • education
Studies accumulating over the past few decades have shed light on the relation between early childhood nutrition and success in school. Height for age is the anthropometric nutritional indicator most strongly linked with school grade and achievement level (1). Studies in Ghana, Tanzania, and the Philippines support the idea that parents often delay initial school enrollment when children are short in stature (2–4). In countries where school participation is not mandatory (e.g., Ghana, Tanzania, Guatamala, and Nepal) low height for age appears to diminish a child’s overall probability of enrolling in school at any age (3,5–7). Studies have also linked low height for age to grade repetition in China (8), Nepal (7), and the Philippines (4). The present study investigated the effect of low height for age on grade school and high school completion, because this was not adequately addressed in earlier reports.
School patterns of enrollment, grade repetition, and dropout are influenced by a complex web of factors ranging from school quality to economic constraints and student readiness. Although height for age is associated with retention and promotion in school, a causal relation remains unproven. The quality of much of the current research is limited by the use of cross-sectional data and inadequate control of potential confounders. The extent and nature of this relation, the pathways through which its most influential effects occur, and the extent to which early childhood nutrition interventions improve schooling outcomes are unclear. The undeniable link between schooling outcomes and economic progress make such knowledge particularly relevant to international researchers.
The present study examined the relation between early childhood malnutrition and schooling outcomes. The study used detailed longitudinal data collected in Cebu, Philippines, on >2000 children. It clarified the effect of height for age on student progress through both grade school and high school, assessed a large variety of potential confounders of this relation, and explored IQ and age at enrollment as potential pathways through which height for age may affect school outcomes. These results offer important contributions to the current literature.
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
The present analysis used the height-for-age Z-score (HAZ; based on WHO reference) at 24 mo. HAZ data were available for 2663 children; some schooling data were also available for 2175 children in this group. Incomplete schooling data further limited sample sizes for the analysis of individual schooling outcomes. Analysis samples included 2191 children whose age at entrance into grade 1 was known, 1997 who answered a question about repeating grades, and 1952 who had a clear final schooling outcome as of 2002 (80 children who were not available for the 2002 survey, but who had clearly dropped out of school or were enrolled but behind during a 2000 tracking survey, were also included, giving a total of 2032 with final schooling outcomes). All of the above data were available for 1970 children.
Schooling outcomes (dependent variables).
We examined multiple outcomes, including age at initial enrollment into school, grade repetition, completion of primary school, and final grade attainment at age 18 y. According to the United Nations Educational, Scientific, and Cultural Organization (UNESCO), the official age at entry in the Philippines in 1990 was 7 y, and large numbers of children entered school late or early at that time (12). We created a 3-level school entrance variable, designating early and late entrance as enrollment before age 6.5 y or after age 8 y, respectively. Children were scored for grade repetition if they reported that they ever repeated a grade in any of the surveys. Children were also divided into 5 categories of ultimate grade attainment: 1) completed < grade 6, 2) completed grade 6 and left school, 3) completed some high school, 4) graduated from high school, and 5) still enrolled in primary or secondary school in 2002 (i.e., in school, but behind). Categories 2 and 3 were combined during analysis because the results did not differ between these categories.
Independent variables.
HAZ was calculated using WHO reference data. The data for HAZ at 24 mo existed for 2504 children and were calculated from the average of measurements at 18 to 24 mo for another 159 children, giving a total of 2663 children. Potential confounders were chosen through literature review to identify factors associated with both low HAZ and schooling outcomes. Those assessed at birth included birth order (number of live births prior to the birth of the child), maternal and paternal education, and maternal height. Paternal educational status (highest grade completed) was assessed in conjunction with the father’s presence in the home at the birth of the child, or at age 8 y if data were unavailable at birth. Approximately 3.6% of the children had no paternal schooling grade recorded because the father was not in the home.
Variables assessed from the 1991 survey (at age 8.5 y) included household assets, number of siblings, deflated household income (assessed both in tertiles and as a logarithmic continuous variable), place of residence, presence of electricity, and environmental cleanliness. Household assets were assessed as a summary count of the presence of an air conditioner, refrigerator, and television, plus a score ranking residential construction quality [0–2]. Number of siblings was assessed as 2 variables: 1) the number of in-house siblings 
5 y old in 1991 (because parents may delay enrolling older children to keep their help in the home), and 2) the number 
15 y old (because older, more capable children may help share the household workload and make it easier for younger children to attend school). Place of residence was classified in 2 ways: 1) as urban or rural, and 2) more finely as urban, urban squatter, periurban, rural, or rural remote. Environmental cleanliness was measured by an index variable created by summing the presence of 4 component variables: 1) an indoor in-house toilet (vs. a latrine or no toilet), 2) piped water [vs. a borehole or some other less sanitary water source (13)], 3) the presence of little or no excreta in the yard and little or no garbage in the neighborhood (vs. some or much), and 4) a food area ranked by the interviewer as "very clean" (vs. "not clean" or "filthy").
IQ at 8 y and age at enrollment (derived above as a dependent variable) were not categorized with the confounders, but were explored separately as potential pathways explaining the effects of HAZ on repetition and long-term retention in school. IQ was measured using the Philippines nonverbal intelligence test developed by Guthrie et al. (14) to measure fluid abilities (i.e., analytic and reasoning skills), which requires children to discriminate differences between 5 pictures on each of 100 cards. The test was standardized by the developers at a public elementary school in Manila, using data from 100 children in each grade, but it is not nationally representative or culturally transferable and is intended only for within-sample comparisons.
Statistical methods.
To assess the potential for bias related to missing data, children were stratified by the amount of schooling data available (none, some, or complete), and differences in sociodemographic variables were assessed by comparison of either means (continuous variables) or percentages (dichotomous variables). Schooling outcome differences by sex and HAZ were similarly tabulated. Mean differences in binary comparisons were assessed with two-sided t tests and in multiple comparisons with one-way ANOVA overall F-tests and post-hoc Bonferroni comparisons. Percentile differences were assessed with Pearson’s chi-square test. Values of 
= 0.05 were considered to be significant.
Logistic regression models of each schooling outcome regressed on HAZ were used as a basis for sensitivity analysis of the confounders. For simplicity, dichotomous versions of multinomial schooling outcomes were used at this stage, including the following variables: on-time or early vs. late enrollment, ever vs. never repeated a grade, dropped out of grade school vs. completed at least grade 6, and completed high school vs. dropped out or were still enrolled in school. HAZ coefficients from 2 models—a sex-specific univariate logistic model that regressed each schooling outcome on HAZ and an identical model that also included the individual potential confounder—were compared. Independent variables altering the HAZ coefficient by >10% of its original value were controlled for in further regression analyses. Any variables that were identified as confounders in either males or females were retained for further analysis.
Regular and multinomial logistic regression models were generated as appropriate for each final schooling outcome variable. HAZ coefficients were estimated with both crude and adjusted models. Adjusted models predicting grade repetition and completion were further expanded to include IQ and age at school entrance, both individually and in combination. All models were stratified by gender.
Models predicting the change in each of the schooling outcomes associated with a change in overall height from –2 to 0 SD of HAZ were simulated and graphically represented. All analyses were performed using Stata software (15).
|
RESULTS |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
|
|
|
|
Models including IQ at 8 y, age at enrollment, or both, as potential pathways through which HAZ might affect schooling outcomes were also calculated (Table 5). These models were otherwise identical to the adjusted models (without the HAZ change variable; Table 4). We hypothesized that if the effects of low HAZ acted through either or both of these pathways, the associations between HAZ and schooling would decline or disappear after adjustment for these variables.
The association between grade repetition and HAZ was not affected by age at enrollment, but was attenuated by IQ. Dropout patterns among girls were also more completely explained by IQ than by age at enrollment. Among girls, the inclusion of IQ attenuated the associations between HAZ and dropping out in grade school or being in school but behind to nonsignificance (P = 0.462 and P = 0.090, respectively). However, greater height at 2 y still appeared to protect against dropping out in high school after adjustment for IQ. Inclusion of age at enrollment did not generally explain the apparent protective effect of HAZ at 2 y on dropping out or being in school and behind, although the association between HAZ and dropping out in grade school became nonsignificant (P = 0.115).
Among boys, IQ explained dropout patterns less well than age at enrollment. After adjustment for IQ, boys who were taller at 2 y remained less likely to drop out in grade school or to be in school but behind vs. graduate from high school. With adjustment for age at enrollment, the likelihood of being enrolled but behind in school vs. graduating from high school became borderline equivalent regardless of HAZ, but HAZ still appeared to protect against dropping out in grade school.
Four bar charts (Fig. 1) illustrate the adjusted predicted probability of selected schooling outcomes, where HAZ = –2 or 0 for all children. The probability of late enrollment, repeating a grade, being behind in school, and dropping out prior to graduation all declined with increased HAZ among both boys and girls.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Very few published studies focus on grade repetition as an outcome measure for malnutrition (17). Grade repetition (ever vs. never) in the present study cohort was predicted by HAZ, as in a previous study on these data that focused on the repetition of grade 1 (4). In the present study, the percentage of boys ever repeating a grade was nearly double that of girls (58% vs. 34%), although the decrease in repetition rate among children who were taller at 2 y was greater for girls than for than boys (Table 4). Inclusion of IQ attenuated the relation between HAZ and grade repetition for both genders, and nutrition-related IQ effects appeared to be an important pathway affecting the successful completion of grade levels.
The effect of low height for age on the likelihood of dropping out vs. completing high school has not been investigated previously, to our knowledge. Early childhood height for age is clearly associated with the probability of dropping out of high school or being enrolled but behind vs. graduating on time. Boys and girls who were taller at 2 y were markedly less likely to drop out in grade school or to be behind in school and were therefore more likely to graduate from high school on time. Both boys and girls who were taller at 2 y were also less likely to drop out during high school, although this association persisted only among girls after controlling for confounding factors. Because total schooling is commonly used as a predictor for socioeconomic outcomes, understanding the role of malnutrition as a determinant of high school completion is of value in social planning and intervention.
Low IQ and low height for age occur simultaneously within the context of early childhood malnutrition, and previous research shows that low height for age predicts depressed IQ scores among Philippine children (20) and that IQ scores in grade 3 are associated with later retention in school among American children (21). We interpreted HAZ at 2 y as an indicator of childhood malnutrition and IQ at 8 y as a mediator through which the long-term effects of malnutrition may operate. Thus, we expected that IQ would mitigate the effects of height for age on schooling outcomes, and the data supported this prediction. We further hypothesized that late school enrollment would affect progress and attainment in school. Enrollment in school helps stunted children overcome IQ deficits (20). Late enrollment, then, would broaden the gap in achievement capacity between children of different heights and might affect ultimate capacity for grade attainment.
Although we acknowledge the value of previous studies, we also recognize the need to explore the relation between height for age and schooling further. Earlier research was often limited by the type and quality of data available. Few published studies are longitudinal. Cross-sectional data, although valuable for positing associations, cannot be used to verify the temporality of critical events, and actual outcome trajectories are not often clear. For instance, in a study in China, Jamison (8) reported an association between a 1-SD reduction in height for age and an additional 4 mo delay in schooling achievement. However, the analysis was unable to differentiate between the effects of late enrollment and grade repetition, although the author cited the high prevalence of repetition of grade 1 (25%) and felt this was the main cause of the trend. Moock and Leslie (7) reported valuable findings from a study in Nepal, showing that height for age predicted the difference between actual and expected grade attainment, but they were similarly unable to assess the cause of this difference. Perfect estimation of these relations is still well outside our reach, but the use of longitudinal data improves these estimates markedly by allowing the tracking of children through time.
The present analysis has several strengths. First, the data were well suited for this analysis. The Philippines have high levels of nutritional inadequacy among children [aggravated by an economic recession during the 1980s when these data were collected (22)] paired with strong national support for education (i.e., mandatory primary school enrollment for all Filipino children). Alhough nearly all Filipinos enroll in grade school, their schooling trajectories vary markedly. Early and late enrollment, grade repetition, dropout, and school completion are prevalent (4), making the Philippines an ideal setting to investigate the effects of early childhood nutrition on actual progress through school. Second, the study was designed to identify factors affecting the specific relation between growth retardation and schooling outcomes and to investigate the extent and veracity of those effects. A great number of studies in the literature simply identify those factors of greatest explanatory importance with respect to schooling (1,3,16,23–25). Through these and other studies, nutritional status has become a recognized concern, but few studies have subjected the relation between HAZ and schooling outcomes to rigorous control of confounding factors. We hypothesized that the apparent effects of early childhood nutritional insults on schooling outcomes could not be entirely dismissed even after carefully accounting for other childhood insults. Although it was impossible to investigate every potential confounder of this relation given the limitations inherent in the data, many important associations were explored. Third, although many studies have explored the effects of undernutrition on success in grade school (see previous citations), and a few have looked at secondary school enrollment (26,27), we were unable to locate any that evaluate the effects of early growth retardation on schooling outcomes through high school completion. This was an important component of the present analysis.
The present study also has weaknesses. Because schooling outcomes are the result of a complex web of causes, the true relations between nutrition and school achievement may be largely confounded and difficult to isolate. We used the classical definition of a confounder—something that 1) may cause the outcome (i.e., schooling) independent of the main exposure and 2) is associated with, but not caused by, the exposure (e.g., HAZ) (28)—in our literature search and located numerous potential candidates. Although we adjusted for all of these for which CHLNS data were available (birth order, maternal and paternal education, maternal height, household assets, number of siblings, household income, place of residence, presence of electricity, and environmental cleanliness), there were several items for which we could not adjust and which therefore introduced the potential for residual confounding. Differences in parental attitudes, the child’s general health, school and teacher quality, school location and availability (16), verbal ability at enrollment (6), blood lead levels (29,30), parasitic infection rates (5,31,32), in-home stimulation including attention (33,34), available reading materials (27), and travel time required to get to school (35) are all predictors of schooling success and could also help explain early childhood nutrition patterns. These factors may further explain the relation between linear growth and schooling outcomes. Further research examining the interplay of these factors in the relation between nutritional status and schooling outcomes—specifically, whether early nutritional insults continue to affect schooling outcomes once other risk factors are accounted for—would be valuable to those seeking to improve schooling outcomes in the less-developed world.
Another drawback to the current study was the nature of the IQ test. In most cases, the test was administered after children had been enrolled in school (age 8 y), and it contained some items that appeared to us to test abilities learned in school. Therefore, the test may have been responsive to the amount of previous schooling a child had rather than being a simple measure of a child’s innate ability. The proof of our hypothesis (that malnutrition affects IQ, which in turn affects school outcomes) is thus clouded by the effect that schooling outcomes may have had on IQ. This would result in an overestimation of the explanatory power of IQ and an underestimation of the explanatory power of HAZ with regard to schooling outcomes. It is also important to note that differentials in even purely measured IQ scores are related to different behavioral characteristics in children. Parents may treat children with lower IQs differently, in which case IQ would affect both the nature of early care the child receives and the child’s ultimate success in school, and thereby confound the relation between the two. We felt that IQ was more likely to act as a mediator than a confounder and constructed our hypothesis accordingly. Considering both of these possibilities, the appropriateness of controlling for IQ at all comes into question. Therefore we excluded IQ from the first round of analyses. The role that IQ plays in the relation between malnutrition and school outcomes clearly needs further investigation.
Although it is the classically preferred indicator in nutrition and schooling studies, height for age has the drawback of being a composite measure. Growth retardation may be caused by a variety of insults, such as both pre- and postnatal nutritional deficits (36), infection or illness that results in an unmet increase in nutrient needs (37), and inadequate stimulation from caregivers (34). Accordingly, the connection between malnutrition and schooling implied in the present study must be interpreted cautiously, with recognition that in most cases, growth retardation is probably due to a combination of these factors.
The results of the present analysis underscore the importance of height for age as a predictor of enrollment, grade repetition, and retention in Filipino schools. Children receiving good nutrition in early life are facilitated in their schooling not only through more optimal cognitive development. Low HAZ is associated with biological and environmental changes that increase the resistance between children and their educational achievements. Such changes include increased stress levels (damaging adaptability and problem-solving skills) (38,39), increased rates of infection (increasing nutritional needs as well as damaging attendance capability) (37), increased rates of conduct difficulties (40), differential caregiver behavior (poorer development leading to fewer positive interactions with caregivers), and lower activity levels (41), although more research is needed to verify that activity levels affect development. Height for age thus becomes a marker for a host of malnutrition effects and predicts schooling success due to this combination of effects. Nutritional status measured via HAZ at age 2 y is not sufficient to explain all of the later schooling patterns of these children, but it appears to have real and lasting effects that persist through high school.
|
FOOTNOTES |
|---|
2 The Cebu Longitudinal Health and Nutrition Survey is a collaborative project with the Office of Population Studies of the University of San Carlos. Follow-up surveys were supported by grants from the World Development Bank and the Asian Development Bank (1991, in collaboration with Paul Glewwe and Elizabeth King), the Thrasher Research Fund (1991), the U.S. Agency for International Development (1994 and 1998), the Mellon Foundation (1998), and the Fogarty International Center of the National Institutes of Health (R01 TW05596–01).
4 Abbreviations used: CLHNS, Cebu Longitudinal Health and Nutrition Survey; HAZ, height-for-age Z-score; OR, odds ratio; UNESCO, United Nations Educational, Scientific, and Cultural Organization.
Manuscript received 12 January 2004. Initial review completed 7 February 2004. Revision accepted 22 March 2004.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
1. Grantham-McGregor, S. M. & Walker, S. P. (1998) Health and nutritional determinants of school failure. Nutrition, Health, and Child Development: Research Advances and Policy Recommendations 1998 Pan American Health Organization, World Bank, & Tropical Metabolism Research Unit, University of the West Indies Washington, DC.
2. The Partnership for Child Development (1999) Short stature and the age of enrolment in primary school: studies in two African countries. Soc. Sci. Med. 48:675-682.
3. Fentiman, A., Hall, A. & Bundy, D. (2001) Health and cultural factors associated with enrolment in basic education: a study in rural Ghana. Soc. Sci. Med. 52:429-439.
4. Glewwe, P., Jacoby, H. G. & King, E. M. (2001) Early childhood nutrition and academic achievement: a longitudinal analysis. J. Public Econ. 81:345-368.
5. Beasley, N. M., Hall, A., Tomkins, A. M., Donnelly, C., Ntimbwa, P., Kivuga, J., Kihamia, C. M., Lorri, W. & Bundy, D. A. (2000) The health of enrolled and non enrolled children of school age in Tanga, Tanzania. Acta Trop. 76:223-229.[Medline]
6. Balderston, J. B., Wilson, A. B., Friere, M. E. & Simonen, M. S. (1981) Malnourished Children of the Rural Poor: The Web of Food, Health, Education, Fertility, and Agricultural Production 1981 Auburn House Boston, MA.
7. Moock, P. R. & Leslie, J. (1986) Childhood malnutrition and schooling in the Terai region of Nepal. J. Dev. Econ. 20:33-52.
8. Jamison, D. T. (1986) (1986) Child malnutrition and school performance in China. J. Dev. Econ. 20:299-309.
9. Adair, L. S., Popkin, B. M., VanDerslice, J., Akin, J., Guilkey, D., Black, R., Briscoe, J. & Flieger, W. (1993) Growth dynamics during the first two years of life: a prospective study in the Philippines. Eur. J. Clin. Nutr. 47:42-51.[Medline]
10. The Cebu Study Team (1991) Underlying and proximate determinants of child health: the Cebu Longitudinal Health and Nutrition Study. Am. J. Epidemiol. 133:185-201.
11. The Cebu Longitudinal Health and Nutrition Survey: Survey Procedures (1989) [Online]. Office of Population Studies, University of San Carlos & Carolina Population Center, University of North Carolina at Chapel Hill http://www.cpc.unc.edu/projects/cebu/docs/clhn1.pdf [accessed February 17, 2004], pp. 1–73.
12. Philippines Country EFA 2000 Assessment Report, Part I. Descriptive Section (2000) [Online]. United Nations Educational, Scientific, and Cultural Organization http://www2.unesco.org/wef/countryreports/philippines/rappport_1.html [accessed July 23, 2003].
13. VanDerslice, J. & Briscoe, J. (1993) All coliforms are not created equal: a comparison of the effects of water source and in-house water contamination on infantile diarrheal disease. Water Resour. Res. 29:1983-1995.
14. Guthrie, G. M., Tayag, A. H. & Jacobs, P. J. (1977) The Philippine Non-verbal Intelligence Test. J. Soc. Psychol. 102:3-11.
15. StataCorp (2003) Stata Statistical Software Release 8.0 2003 Stata College Station, TX.
16. Jamison, D. T. & Lockheed, M. E. (1987) Participation in schooling: determinants and learning outcomes in nepal. Econ. Dev. Cult. Change 35(2):279-306.
17. Glewwe, P. & Jacoby, H. G. (1995) An economic analysis of delayed primary school enrolment in a low income country: the role of early childhood nutrition. Rev. Econ. Stat. 77:156-169.
18. Agarwal, D. K., Upadhyay, S. K. & Agarwal, K. N. (1989) Influence of malnutrition on cognitive development assessed by Piagetian tasks. Acta Paediatr. Scand. 78:115-122.[Medline]
19. Fentiman, A., Hall, A. & Bundy, D. (1999) School enrolment patterns in rural Ghana: a comparative study of the impact of location, gender, age and health on children’s access to basic schooling. Comp. Educ. 35:331-339.
20. Mendez, M. A. & Adair, L. S. (1999) Severity and timing of stunting in the first two years of life affect performance on cognitive tests in late childhood. J. Nutr. 129:1555-1562.
21. Jimerson, S., Egeland, B., Sroufe, L. A. & Carlson, B. (2000) A prospective longitudinal study of high school drop-outs: examining multiple predictors across development. J. Sch. Psychol. 38:525-549.
22. Heaver, R. A. & Hunt, J. M. (1995) Improving early child development: an integrated program for the Philippines 1995 World Bank Washington, DC.
23. Glick, P. & Sahn, D. E. (2000) Schooling of girls and boys in a West African country: the effects of parental education, income, and household structure. Econ. Educ. Rev. 19:63-87.[Medline]
24. Gomes-Neto, J. B. & Hanushek, E. A. (1994) Causes and consequences of grade repetition: evidence from brazil. Econ. Dev. Cult. Change 43:117-148.
25. Gomes-Neto, J. B., Hanushek, E. A., Raimundo, H. L. & Frota-Bezzera, R. C. (1997) Health and schooling: evidence and policy implications for developing countries. Econ. Educ. Rev. 16:271-282.[Medline]
26. Walker, S. P., Grantham-McGregor, S. M., Himes, J. H. & Williams, S. (1996) Adolescent Kingston girls’ school achievement: nutrition, health and social factors. Proc. Nutr. Soc. 55:333-343.[Medline]
27. Walker, S. P., Grantham-McGregor, S. M., Himes, J. H., Williams, S. & Duff, E. M. (1998) School performance in adolescent Jamaican girls: associations with health, social and behavioural characteristics, and risk factors for dropout. J. Adolesc. 21:109-122.[Medline]
28. Kelsey, J. L., Whittemore, A. S., Evans, A. S. & Thompson, W. D. (1996) Methods in Observational Epidemiology 1996 Oxford University Press New York, NY.
29. Clarke, N.M.A., Grantham-McGregor, S. M. & Powell, C. (1991) Nutrition and health predictors of school failure in Jamaican children. Ecol. Food Nutr. 26:47-57.
30. Fergusson, D. M., Horwood, L. J. & Lynskey, M. T. (1997) Early dentine lead levels and educational outcomes at 18 years. J. Child Psychol. Psychiatry 38:471-478.[Medline]
31. Connolly, K. J. & Kvalsvig, J. D. (1993) Infection, nutrition and cognitive performance in children. Parasitology 107:187S-200S.
32. Stoltzfus, R. J., Albonico, M., Tielsch, J. M., Chwaya, H. M. & Savioli, L. (1997) Linear growth retardation in Zanzibari school children. J. Nutr. 127:1099-1105.
33. Walker, S. P., Powell, C. A., Grantham-McGregor, S. M., Himes, J. H. & Chang, S. M. (1991) Nutritional supplementation, psychosocial stimulation, and growth of stunted children: the Jamaican study. Am. J. Clin. Nutr. 54:642-648.
34. Grantham-McGregor, S. M., Powell, C. A., Walker, S. P. & Himes, J. H. (1991) Nutritional supplementation, psychosocial stimulation, and mental development of stunted children: the Jamaican study. Lancet 338:1-5.[Medline]
35. Glewwe, P. & Ilias, N. (1996) The determinants of school attainment in sub-Saharan Africa: a case study of Ghana. J. Int. Dev. 8:395-413.
36. Fall, C. H., Yajnik, C. S., Rao, S., Davies, A. A., Brown, N. & Farrant, H. J. (2003) Micronutrients and fetal growth. J. Nutr. 133:1747S-1756S.
37. Stephensen, C. B. (1999) Burden of infection on growth failure. J. Nutr. 129:534S-538S.
38. Fernald, L. C. & Grantham-McGregor, S. M. (2002) Growth retardation is associated with changes in the stress response system and behavior in school-aged Jamaican children. J. Nutr. 132:3674-3679.
39. Fernald, L. C. & Grantham-McGregor, S. M. (1998) Stress response in school-age children who have been growth retarded since early childhood. Am. J. Clin. Nutr. 68:691-698.[Abstract]
40. Chang, S. M., Walker, S. P., Grantham-McGregor, S. & Powell, C. A. (2002) Early childhood stunting and later behaviour and school achievement. J. Child Psychol. Psychiatry 43:775-783.[Medline]
41. Gardner, J. M. & Grantham-McGregor, S. (1998) Activity levels and maternal-child behavior in undernutrition: studies in Jamaica. Nutrition, Health, and Child Development: Research Advances and Policy Recommendations 1998 Pan American Health Organization, World Bank, & Tropical Metabolism Research Unit, University of the West Indies Washington, DC.
This article has been cited by other articles:
|
|
 |
|
  S. P. Walker, S. M. Chang, C. A. Powell, E. Simonoff, and S. M. Grantham-McGregor Early Childhood Stunting Is Associated with Poor Psychological Functioning in Late Adolescence and Effects Are Reduced by Psychosocial Stimulation J. Nutr., November 1, 2007; 137(11): 2464 - 2469. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. N. Coly, J. Milet, A. Diallo, T. Ndiaye, E. Benefice, F. Simondon, S. Wade, and K. B. Simondon Preschool Stunting, Adolescent Migration, Catch-Up Growth, and Adult Height in Young Senegalese Men and Women of Rural Origin J. Nutr., September 1, 2006; 136(9): 2412 - 2420. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. D. Wachs, H. Creed-Kanashiro, S. Cueto, and E. Jacoby Maternal Education and Intelligence Predict Offspring Diet and Nutritional Status J. Nutr., September 1, 2005; 135(9): 2179 - 2186. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S Pearce, I. J Deary, A. H Young, and L. Parker Growth in early life and childhood IQ at age 11 years: the Newcastle Thousand Families Study Int. J. Epidemiol., June 1, 2005; 34(3): 673 - 677. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
