Serum retinol levels among preschool children in Central Java: demographic and socioeconomic determinants.

Abstract

Vitamin A deficiency is a significant problem in many countries in the developing world. Reports have noted demographic and socioeconomic risk factors for vitamin A deficiency. 'Deficiency' has usually been defined by clinical signs and symptoms which represent only a small proportion of those children at increased risk for vitamin A preventable morbidity and mortality. As part of a population-based trial of vitamin A to prevent childhood morbidity, we collected census data (n = 666), baseline socioeconomic data (n = 636) and sera (n = 666) from children aged 6-48 months in 25 adjacent villages in a rural area in Central Java, Indonesia; there was more than 95% participation. We used t tests, ANOVA, and a multiple variable linear regression model in our analyses. Differences in mean retinol level were detected for the following variables: village (P < 0.001), child's age (P = 0.03), size of sibship (P < 0.001), mother's occupation (P < 0.01), mother's education (P = 0.05), father's education (P = 0.03), monthly household earnings (P = 0.02), land ownership (P = 0.03), possession of ducks (P = 0.06), radio or tape player (P = 0.02), or a watch or clock (P = 0.07), and presence of a natural well (P = 0.09). Our regression model verified the predictive value of village, age, sibship, land ownership and earnings. We found that owning land and that the highest and lowest categories of reported household income were associated with higher serum retinol levels. We also noted clustering of serum retinol levels by village and discovered that children from larger sibships and infants had significantly lower serum retinol levels. Vitamin A supplementation of lactating mothers, particularly of high parity, and/or their infants should be considered.