|Early Supplementary Feeding and Cognition (Society for Research in Child Development, 1993, 123 pages)|
|III. Conceptual rationale for the follow-up|
As noted in the previous chapter, the published reports of the longitudinal study differed in the criteria used to select samples, in the cognitive tests chosen for analyses, and in the definition of the treatment variable. Thus, only tentative inferences concerning the effects of the nutritional intervention on behavioral development were justified.
However, additional support for these inferences emerged from three other major intervention studies (of more than 100 subjects) that were launched at about the same time as the Guatemala study. One of these was conducted in Bogota, Colombia (Luster et al., 1989; Super, Herrera, & Mora, 1991; Waber et al., 1981), the second in Sui Lin, Taiwan (Adair & Pollitt, 1985; Joos, Pollitt, Mueller, & Albright, 1983), and the third in New York City (Rush, Stein, & Susser, 1980). All these investigations tested the proposition that PEM (protein energy malnutrition) limited cognitive development in childhood. However, despite similar objectives, they differed widely in research designs, criteria for sample selection, nutritional status of the subjects, type and timing of supplementation, and behavioral outcomes selected for study (Pollitt & Oh, 1992).
Published findings from these intervention studies are restricted to the infancy period. In two of these, early supplementary feeding had positive effects on the mental (Waber et al., 1981) and the motor (Jogs et al., 1983; Waber et al., 1981) development of infants. Infants exposed to a nutritional supplement during gestation and the first years of postnatal life performed better than control subjects on mental and motor development scales; the sizes of the effects were equivalent to about .20 of a standard deviation. In the third study, no effects were reported (Rush et al., 1980).
Since the 1987 follow-up in Guatemala was begun, two additional reports on the effects of a high-calorie supplement on infant development have also been published. One study was conducted in Kingston, Jamaica (Grantham-McGregor, Powell, Walker, & Himes, 1991), and the other in West Java, Indonesia (Husaini, Karyadi, Husaini, Sandjaja, Karyadi, & Pollitt, 1991). Their findings agreed in part with the previous studies in showing that nutritional intervention had beneficial effects on the motor, but not on the mental, development of infants and young children.
To further test the proposition that, in nutritionally at-risk populations, early supplementary feeding is sufficient to improve infant development, the five studies noted above and the data from Guatemala were pooled in a meta-analysis. This analysis was restricted to comparisons between experimental groups exposed solely to a nutritional supplement and controls. The samples were stratified by age: one group included infants aged from 8 to 15 months and the other those aged from 18 to 24 months.
The results of the meta-analysis showed a significant effect of supplementation on motor development scores in both age groups (p <.01), but significant effects on mental test scores were limited to the older group of infants (Pollitt & Oh, 1992). These data strengthen the hypothesis that early nutritional intervention has an effect on cognitive development, at least during the first 2 years of life.
Results beyond the infancy period were available from two studies. IQ data obtained at age 5 years from the Taiwan study were published by Hsueh and Meyer (1981); the children were no longer being supplemented at that time since the intervention had ceased with weaning. In contrast to the infancy findings, the supplementation was not associated with IQ. Data on preschoolers from the Bogota study have also recently become available (Super et al., 1991). Three and a half years after the end of the intervention (mean age = 6.74 years), the children were assessed on an achievement test containing items of reading readiness, arithmetic, and basic knowledge. Significant effects accounting for about 8 months of cognitive growth were observed in reading readiness; these effects were more likely to be found in families in which mothers had substantial psychological and social resources.
Although studies of the consequences of early supplementary feeding on the development of infants and preschool-aged children are not directly relevant to the plausibility of the hypothesis of long-term effects, they did provide some backing for our contention. Taken together, their results showed that nutritional intervention was sufficient to affect early behavioral development among nutritionally at-risk infants and preschoolers.