Skip to main content
ENDO 2026 Main banner

Reproductive Endocrinology

ORF02-01 - Effect Modification by Sociodemographic Factors in the Association between PCOS Treatments and Cardiovascular Disease Risk

Saturday, June 13, 2026
9:45 AM - 9:50 AM CT
Location: Room W375B
Disclosure(s):
Snigdha Alur-Gupta, MD MSCE: No financial relationships to disclose
Daniel Soffer, Associate Professor: Amgen Inc.: Lipid monitoring committee - Ocean(a) clinical trials (); Heartflow: Advisory Board (); New Amsterdam: Advisory Board (); Regeneron: Advisory Board ()



Objective: Polycystic ovary syndrome (PCOS) increases atherosclerotic cardiovascular disease (ASCVD) risk, yet there is limited data on long-term cardiovascular effects of first-line treatments—metformin and oral contraceptive pills (OCPs). Whether treatment efficacy varies by sociodemographic factors is unknown, potentially contributing to health disparities. We examined effect modification across sociodemographic subgroups when evaluating associations between metformin and OCPs, alone or combined, and ASCVD risk.

Methods: Individuals with PCOS without diabetes or treatment contraindications aged 18-40 were identified in Optum de-identified Clinformatics® DataMart. We emulated a hypothetical randomized controlled trial (target trial framework), classifying individuals into metformin, OCPs, combination, or neither (reference) groups based on prescription dispensations for consecutive months between 12/2000-10/2019. Follow-up occurred until ASCVD event (coronary artery disease, cerebrovascular disease, peripheral artery disease) which were defined using ICD 9/10 codes, disenrollment, or administrative censoring at five years. Five-year ASCVD risk ratios (RR) by treatment were estimated via pooled logistic regression, weighted for baseline and time-varying age, comorbidities, and reproductive health. We estimated the impact of receiving the treatment for at least one year on 5-year risk of ASCVD. Stratified analyses were conducted among subsets with available data for race (White, Black, Asian), ethnicity (Hispanic, Non-Hispanic), baseline age ( < 30, ≥30 years), education (< Bachelor's, ≥Bachelor's), and net worth ( < $250K, ≥$250K). Confidence intervals (CI) were calculated via bootstrapping.

Results: Among 156,861 individuals (41,972 metformin, 40,990 OCPs, 7,084 combination, 66,815 neither), there were a total 1,545 ASCVD events. Metformin (RR 0.56, 95% CI 0.33-0.82) and combination therapy (RR 0.45, 95% CI 0.16-0.71) reduced risk while OCPs showed no association (RR 1.01, 95% CI 0.78-1.74). Metformin demonstrated consistent protective associations across all subgroups. OCPs showed possible heterogeneous associations: increased risk among Black women (RR 1.77, 95% CI 0.53-5.25), younger individuals (RR 1.26, 95% CI 0.80-2.23), and those with higher education (RR 1.39, 95% CI 0.77-3.19), though results were not significant. Combination therapy showed high variability due to limited events but was overall protective.

Conclusions: To our knowledge, this is the first study showing that metformin consistently reduced ASCVD risk among individuals with PCOS over a 5-year period. OCPs showed no change in ASCVD risk, with potential heterogeneous associations across sociodemographic subgroups. These findings support continued use of first line medications and personalized counseling in certain subgroups.