Limited retinol (ret) stores at birth contribute to infant ret deficiency, a condition common in developing nations. Maternal (M) and infant (I) serum ret levels are being measured in 100 M-I pairs from different socioeconomic classes in Quito, Ecuador to identify determinants of newborn ret status.Methods: M and I blood were collected within 24 hours of birth. M were interviewed and classified into mid to upper class (class 1; n=25) and mid to lower class (class 2; n=28). A linear regression model which included sociodemographic and clinical variables was used to identify predictors of I and M serum ret levels. Results: 53 M-I pairs have been studied to date. Class 1 mothers were more likely to have taken prenatal vitamins (100% vs 68%, p=0.037). 96% of I ret levels were in the deficient range. M serum ret, parity, weight gain, age, and I birthwts and gestational ages did not differ between groups. I serum ret levels were significantly lower among class 2 infants. Linear regression identified only social class as a significant predictor of I serum ret levels. Conclusion: Deficient levels of ret were found in most I, but in none of the mothers. I but not M ret levels were significantly associated with social class. These data suggest that factors other than M ret, such as ethnicity, may affect newborn serum ret levels. Table
