Kramer M S, Barr R G, Leduc D G, Boisjoly C, McVey-White L, Pless I B
J Pediatr. 1985 Jan;106(1):10-4. doi: 10.1016/s0022-3476(85)80456-x.
To overcome methodologic defects (failure to control for confounding factors, univariate statistical analyses) in previous studies of etiologic determinants of childhood adiposity, we carried out a prospective cohort study of 462 healthy, full-term infants observed from birth to 12 months. Postpartum, we obtained sociodemographic data and administered two recently validated scales of maternal attitudes toward feeding and infant body habitus. Parental heights and weights and infant feeding variables were determined by interview, and at 6 and 12 months we measured height and weight and triceps, subscapular, and suprailiac skinfolds. Multiple regression analysis was used to determine independently predictive factors for weight, body mass index (BMI = weight/height), and the sum of the three skinfold measurements. Birth weight, sex, age at introduction of solid, and duration of breast-feeding were all significant predictors of weight at 12 months (r2 = 0.296, P less than 0.0001). Significant determinants for BMI included birth weight, duration of breast-feeding, sex, and IBH (r2 = 0.125, P less than 0.0001); those for total skinfold were age at introduction of solid foods and birth weight (r2 = 0.038, P = 0.002). Similar results were obtained at 6 months, although slightly less of the variance was explained. We conclude that the ability to predict which babies will be heavy or obese during the first year is limited. Breast-feeding and delayed introduction of solid foods do offer some protective effect, however, and thus efforts to encourage these practices may be reaping some benefit.