Sex-specific Associations of Maternal Birthweight with Offspring Birthweight in the Omega Study: The role of pre-pregnancy body mass index

Background: Birth weight is an indicator of fetal growth and development, important determinants of life course health. Maternal birthweight (BW) is one of the strongest predictors of offspring BW, perpetuating generational cycles of poor health and risk for adverse birth outcomes. However, potential non-linear relationships or modifying or mediating factors of maternal and offspring BW associations have not been fully described. Objective: We examined (1) the overall maternal-offspring BW association, (2) offspring sex-specific differences in the association, as well as (3) the role of pre-pregnancy body mass index (BMI) as modifier or mediator of the association. Methods: The study was conducted among N = 3745 participants of the Omega pregnancy cohort study in Washington State (1996-2008). The exposure variable was maternal BW, modeled as a continuous variable, a linear spline with knots at 2500 grams and 4000 grams (macrosomia), and a binary variable indicating low birthweight (LBW) status (< 2500 grams vs. ≥ 2500 grams). Outcomes were offspring BW and LBW status. Linear and logistic regression models, adjusted for potential confounders, were used to estimate differences in mean offspring BW and odds of LBW, respectively, and corresponding 95% confidence intervals. Effect modification (by offspring sex or pre-pregnancy BMI) was assessed using stratified analyses and interaction terms. Mediation of associations by pre-pregnancy BMI was examined using the potential outcomes framework of causal analysis. Results: The average offspring BW in the study population was 3450.8 grams. Overall, 4.2% of the offspring had LBW. An increase in maternal BW was positively associated with offspring BW among mothers with normal BW (β =22.05 grams offspring BW per 100 grams maternal BW, 95% CI: 17.32, 26.79) or macrosomia (β =23.86 grams offspring BW per 100 grams maternal BW, 95%CI: 7.26, 40.46). LBW mothers had a two-fold higher risk of having LBW infants (OR = 1.96, 95% CI: 1.24, 3.09). Across the distribution of maternal BW we observed a non-linear slope for males (likelihood ratio test (LRT) p-value for improvement in fit by linear spline vs continuous model < 0.001) and a linear slope for females (LRT p-value = 0.9231). The relationship between maternal and offspring BW, as represented by the linear spline, differed by offspring sex (p-value for interaction = 0.0149). Mothers who were LBW and had male offspring had a statistically insignificant 1.2-fold (95% CI: 0.55, 2.72) higher risk of having a LBW infant while mothers who were LBW and had female offspring had a 2.67-fold (95% CI: 1.51, 4.75) higher risk of having a LBW infant (p-value for interaction = 0.120). In sex-specific analyses, maternal-offspring BW associations were similar among women of normal pre-pregnancy BMI or overweight/obese pre-pregnancy BMI groups. Maternal pre-pregnancy BMI mediated 2.57% (95% CI: 2.16%, 3.21%) of the associations between maternal and offspring BW, overall. However this mediation was not evident in offspring sex-specific models. Conclusion: We found that the association between maternal and offspring BW may differ by offspring sex, with a non-linear relationship among males, but not females. Pre-pregnancy BMI did not appear to modify maternal-offspring BW associations and it mediated only a small proportion of the associations. Our findings, if supported by other replication studies in diverse populations, provide a new paradigm to understand factors involved in transgenerational BW transmission and guide future investigations of potential mechanisms accounting for maternal-offspring BW associations.