Obesity as a mediator of peripartum cardiovascular complications in polycystic ovary syndrome

Introduction Women with polycystic ovary syndrome (PCOS) face elevated risks of cardiovascular (CV) complications during delivery hospitalizations.Rates of PCOS and obesity have risen in parallel among childbearing women, yet the extent to which obesity mediated PCOS associated peripartum CV remains unclear. Our study aimed to address this by quantifying the proportion of excess CV risk attributable to PCOS that is explained by obesity. Methods Using the National Inpatient Sample (2016–2022), we identified delivery hospitalizations among women aged 12–55 years via ICD-10-CM codes. PCOS was identified by code E28.2 and obesity by E66.x/O99.21x. The primary outcome was a composite acute CV event (preeclampsia, eclampsia, peripartum cardiomyopathy, heart failure, pulmonary oedema, venous thromboembolism, acute myocardial infarction, stroke, arrhythmia, cardiogenic shock, and cardiac arrest). Survey-weighted logistic regression employed a change-in-estimate mediation approach: Model 1 (unadjusted), Model 2 (adjusted for demographics, comorbidities, and hospital characteristics without obesity, showcasing total effect), and Model 3 (additionally adjusted for obesity, showcasing direct effect). Percent attenuation was calculated as (β-total − β-direct) / β-total × 100. Results Among 26,659,071 weighted delivery hospitalizations, 208,620 (0.78%) had PCOS. Women with PCOS were older (mean 30.9 vs 29.0 years, p<0.01) and had higher prevalence of obesity (34.8% vs 12.2%), diabetes (5.6% vs 2.0%), gestational diabetes (17.9% vs 8.3%), chronic hypertension (3.4% vs 1.2%), and dyslipidemia (2.5% vs 0.4%; all p<0.01). Composite acute CV events were more frequent with PCOS (10.1% vs 6.2%, p<0.01). PCOS was associated with 70% higher unadjusted odds of composite CV events (OR 1.70, 95% CI 1.65–1.76). After multivariable adjustment excluding obesity, PCOS remained independently associated (aOR 1.39, 95% CI 1.34–1.44). Upon further adjustment for obesity, the PCOS effect was attenuated but persisted (aOR 1.24, 95% CI 1.19–1.28). Obesity was a strong independent predictor (aOR 1.77, 95% CI 1.75–1.79). The percent attenuation attributable to obesity was 35.7%, indicating that approximately one-third of PCOS-associated excess CV risk is mediated through obesity. Conclusion In this analysis of over 26 million delivery hospitalisations, PCOS was independently associated with significantly elevated peripartum CV risk. Obesity mediated approximately 36% of this excess risk, yet a substantial direct PCOS effect persisted. These findings underscore the dual importance of targeting both PCOS-specific pathophysiology and obesity-related cardiometabolic burden to improve maternal CV outcomes.