Abstract
We evaluated relationships between preconception adiposity and human offspring sex and sex ratio. Using data from a prospective preconception cohort nested within a randomized controlled trial based at 4 US clinical sites (2006-2012), we used logistic regression to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for male:female sex ratio, and log-identity regression to estimate risk differences (RDs) and 95% CIs for male and female livebirth according to preconception adiposity measures. Inverse-probability weights accounted for potential selection bias. Among 603 women attempting pregnancy, there were meaningful reductions in sex ratio for the highest category of each adiposity measure. The lowest sex ratios were observed for obesity (body mass index of ≥30, calculated as weight (kg)/height (m)2, OR = 0.48, 95% CI: 0.26, 0.88) relative to normal body mass index, and the top tertiles (tertile 3) of serum leptin (OR = 0.50, 95% CI: 0.32, 0.80) and skinfold measurements (OR = 0.50, 95% CI: 0.32, 0.79) relative to the lowest tertiles. Reductions were driven by 11-15 fewer male livebirths per 100 women (for obesity, RD = -15, 95% CI: -23, -6.7; for leptin tertile 3, RD = -11, 95% CI: -20, -3.2; and for skinfolds tertile 3, RD = -11, 95% CI: -19, -3.3). We found that relationships between preconception adiposity measures and reduced sex ratio were driven by a reduction in male births.