Interactions of pre-pregnancy body mass index with early- and mid-pregnancy gestational weight gain in relation to gestational diabetes mellitus: a retrospective cohort study

Abstract Background Pre-pregnancy body mass index (BMI) and gestational weight gain (GWG) are associated with gestational diabetes mellitus (GDM), but their stage-specific interactions remain unclear. We examined the independent, interactive, and joint associations of pre-pregnancy BMI and early- and mid-pregnancy GWG with GDM. Methods This retrospective cohort study included 1,661 pregnant women. Univariable and multivariable logistic regression models examined the associations of pre-pregnancy BMI and early- and mid-pregnancy GWG with GDM, adjusting for maternal age, mode of conception, parity, polycystic ovary syndrome (PCOS), previous GDM, and family history of diabetes. Pairwise multiplicative and additive interactions were assessed, followed by a three-factor joint analysis. Results GDM was diagnosed in 372 women (22.4%). Pre-pregnancy overweight/obesity was associated with higher odds of GDM (adjusted odds ratio [aOR] 2.213, 95% confidence interval [CI] 1.655–2.958). Pre-pregnancy BMI showed a significant multiplicative interaction with early-pregnancy GWG ( P < 0.001). A negative additive interaction was observed between overweight/obesity and insufficient early-pregnancy GWG (RERI − 1.011, 95% CI − 1.773 to − 0.249; AP − 0.808, 95% CI − 1.597 to − 0.019). Six joint exposure categories were associated with significantly higher odds of GDM; all involved pre-pregnancy overweight/obesity. These comprised insufficient early-pregnancy GWG with excessive mid-pregnancy GWG; appropriate early-pregnancy GWG with insufficient or excessive mid-pregnancy GWG; and excessive early-pregnancy GWG with any mid-pregnancy GWG category. Conclusions Pre-pregnancy overweight/obesity was associated with higher odds of GDM, and the strength of this association varied by early-pregnancy GWG category. These findings support BMI-informed preconception counseling and early-pregnancy weight monitoring; they do not establish the effects of weight-loss interventions.