Sex- and Menopause-Specific Hormone-Related ECG Signatures and Incident Cardiovascular Risk in UK Biobank

Abstract

Aims: Sex differences in cardiac electrophysiology and cardiovascular (CV) risk are well established, but the extent to which circulating hormonal markers contribute to electrocardiogram (ECG) morphology remains unclear. We investigated whether hormone-related ECG components are associated with incident CV events and whether these associations differ by sex and menopausal status. Methods: In a prospective cohort (N=8,056 UK Biobank participants; median follow-up: 12.8 years), we first evaluated direct associations between sex hormones and incident CV disease. Based on initial risk signals, two-stage analyses focused on premenopausal females and males. Linear regressions evaluated associations between standardized hormonal markers and residualized lead-I ECG parameters after confounders adjustment. Cox proportional hazards models then tested whether the hormone-predicted ECG components were associated with incident CVD events. Results: In premenopausal females (N=980), higher free androgen index (FAI) was positively associated with QRS amplitude ({beta}=19.17V, p-value=0.038) and the FAI-predicted QRS amplitude component was associated with incident CVD (Hazard Ratio [HR]=1.41, p-value=0.02). In males (N=3,997), FAI was inversely associated with QTc duration ({beta}=-0.95ms, p-value=0.005), and its predicted component with increased CVD risk (HR=1.32, p-value<0.001). Conversely, higher sex hormone-binding globulin (SHBG) prolonged QTc, with its predicted component also conferring risk (HR=1.15, p-value=0.004). Additionally, FAI directly associated with ST segment deviation ({beta}=19.17V), whose predicted component was protective (HR=0.76, p-value<0.001). Conclusions: This study shows that SHBG and FAI subtly modulate specific depolarization and repolarization ECG parameters. Relative androgen excess in premenopausal females and lower androgenic status in males appear to be associated with subclinical electrophysiological variations linked to long-term CV risk.