Determinants of plasma cholesterol in children--a family study.

Twin and family studies have revealed genetic and maternal influences on plasma cholesterol. This study was designed to identify familial and environmental variables related to cholesterol levels in children. Volunteer families (n = 74) were obtained from a previous study of cord blood cholesterol. When the children were 2.5 years old, blood samples, dietary histories, heights, weights, and skinfold thicknesses were obtained. Dietary variables were significantly correlated with plasma cholesterol levels in children but not in their parents. Stepwise multiple regression was used to identify sets of variables predictive of plasma cholesterol fractions of these children. For plasma total cholesterol, the set included previous history of breast feeding (versus formula), child's current dietary poly-unsaturated/saturated fatty acid ratio, maternal alcohol and fat consumption, paternal cholesteryl esters, and maternal hematocrit (r = 0.53). Free cholesterol had a similar predictive variable set (r = 0.52). More variation in esterified cholesterol of children was predicted (r = 0.69) by history of breast feeding, child's current protein and caloric intake, maternal total and esterified cholesterol, and proteinuria during pregnancy. Cord blood cholesterol was not predictive of any later cholesterol measurement. These correlations of children's plasma cholesterol with multiple factors emphasize the importance of focusing on very early childhood for study of both the normal developmental pattern of human cholesterol and intervention to prevent adult hypercholesterolemia.