Depressive Symptoms and Central Obesity Over the Menopausal Transition and the Development of Type 2 Diabetes Later in Life: Decomposing Direct and Indirect Effects

Objective: This study investigated the association between depressive symptoms during the menopausal transition and subsequent incident type 2 diabetes (T2D), with particular focus on: (1) quantifying the mediating role of central obesity in this association, and (2) examining differential effects by depressive symptom subscales (somatic vs. cognitive/affective symptoms). Methods: We analyzed data from 1719 participants in the Study of Women’s Health Across the Nation, a multi-ethnic cohort study of midlife women. Depressive symptoms were assessed repeatedly using the Center for Epidemiologic Studies Depression Scale (CES-D). We calculated the total and subscales of depressive symptom burden during the menopausal transition. Causal mediation analysis within a counterfactual framework was employed to decompose the total effect between depressive symptoms and incident T2D during postmenopausal follow-up into natural direct effects (NDE) and indirect effects (NIE) mediated through central obesity measured at the first postmenopausal visit, adjusting for demographic, lifestyle, and health-related covariates. Results: During a mean postmenopausal follow-up of 11.60 (SD 3.58) years, 170 women (9.89%) developed T2D during follow-up. Central obesity mediated 21.4-23.5% of the total effect of depressive symptoms burden on T2D risk. Somatic subscales of CES-D scores showed strongest associations (OR=1.272 (95% CI:1.041-1.567) and NIE OR=1.100 (1.040-1.165)) in unadjusted models, with 21.3-25.0% mediated through central obesity after full adjustment. Cognitive/affective subscale showed weaker mediation effects (19.4-26.9%). Sensitivity analyses confirmed robustness to unmeasured confounding, with somatic symptom effects remaining significant under moderate confounding (NIE OR= 1.057 (95% CI:1.004–1.115)). Conclusions: Greater cumulative depressive symptom burden during the menopausal transition, particularly somatic subscales, was associated with higher odds of incident T2D after postmenopause, with part of the association mediated through central obesity at postmenopause. These findings highlight the menopausal transition as a critical period for interventions targeting both mental health and metabolic risk factors.