History of miscarriage and type 2 diabetes

Abstract

Abstract We examined a history-based measure of miscarriage propensity in relation to the future occurrence of miscarriage, gestational diabetes mellitus (GDM), and type 2 diabetes. For each of 42,313 gravid Sister Study participants, we estimated propensity for miscarriage, accounting for maternal age and smoking, and propensity for GDM, accounting for body mass index, using an empirical Bayes method. For pregnancy-level predictive analyses, these propensities were updated over time based on prior pregnancy outcomes. Associations with subsequent miscarriage and GDM were estimated using within-cluster-resampled logistic regression. Incident type 2 diabetes was assessed prospectively using adjusted Cox models for 39,468 gravid women who enrolled after completing reproduction and had no prior diabetes diagnosis. Over 532,364 person-years of follow-up, 2874 cases were observed. Per 0.1-unit increase, prior miscarriage propensity was associated with subsequent miscarriage (odds ratio [OR] 5.14; 95% CI 4.20–6.28) and with subsequent GDM (OR 1.51; 95% CI 1.02–2.22). Pregnancy-history-based propensities for miscarriage and GDM were both associated with incident type 2 diabetes (hazard ratios per 0.1 increase were 1.38; 95% CI 1.13–1.69, and 1.35; 95% CI 1.29–1.41, respectively). Thus, when the full reproductive history is considered, even with adjustment for propensity to GDM, history-based propensity to miscarriage is associated with subsequent risk of type 2 diabetes.

The authors' abstract, as published at the source. Scientific Reports, 2026 · DOI ↗