BackgroundThe specialization of human fat deposits is an inquiry of special importance in the study of fetal growth. It has been theorized that maternal lower-body fat is designated specifically for lactation and not for the growth of the fetus. ObjectiveOur goal was to compare the contributions of maternal upper-body versus lower-body adiposity to infant birth weight. We hypothesized that upper-body adiposity would be strongly associated with infant birth weight and that lower-body adiposity would be weakly or negligibly associated with infant birth weight—after adjusting for known determinants. Study designIn this prospective cohort study, 355 women initiated medical pre-natal care during the first trimester of pregnancy at The University of Oklahoma Health Sciences Center during 1990–1993. Maternal anthropometric measurements were assessed at the first clinic visit: (a) height; (b) weight; (c) circumferences of the upper arm, forearm, and thigh; and, (d) skin-fold measurements of the bicep, subscapular region, and thigh. ResultsInfant birth weight was regressed on known major determinants to create the foundational model. Maternal anthropometric variables subsequently were added one at a time into this multiple regression model. The highest contribution by a single anthropometric variable to infant birthweight was, in order: subscapular skin-fold, forearm circumference, and thigh circumference. With one upper-body (subscapular skin-fold) and one lower-body (circumference of the thigh) adiposity measure in the model, the z-score regression coefficient (s.e.) was 85.7g (30.8) [p=0.0057] for maternal subscapular skin-fold and 19.0g (31.6) [p=0.5477] for circumference of the thigh. When the second-best upper-body contributor to infant birthweight (circumference of the forearm) was entered with one lower-body measure into the model, the z-score regression coefficient (s.e.) was 77.5g (38.5) [p=0.0451] for maternal forearm circumference and 14.1g (38.5) [p=0.7146] for circumference of the thigh. When both subscapular skinfold and forearm circumference were added to the model in place of BMI, the explained variance (r2=0.5478) was similar to the model using BMI (r2=0.5487). ConclusionUpper-body adiposity – whether operationalized by subscapular skin-fold or circumference of the forearm – was a markedly larger determinant of infant birth weight than lower-body adiposity.