Manjula Nandakumar, Sana Waris, Thozhukat Sathyapalan, Alexandra E Butler, Stephen L Atkin
Endothelial dysfunction has been reported in polyendocrine metabolic ovarian syndrome (PMOS) and is associated with obesity. We hypothesised that circulating biomarkers of endothelial dysfunction would not differ between non-obese women with PMOS and age- and BMI-matched controls in the absence of major between-group differences in insulin resistance (IR), systemic inflammation or vitamin D status. Plasma concentrations of 20 endothelial-associated proteins were measured using Slow Off-rate Modified Aptamer (SOMAscan) technology in 29 non-obese women with PMOS and 29 age- and BMI-matched controls. Women with PMOS had higher Ferriman-Gallwey scores, free androgen index (FAI) and anti-Müllerian hormone (AMH), whereas HOMA-IR, C-reactive protein (CRP), vitamin D metabolites and endothelial-associated biomarkers did not differ significantly between groups. Within the PMOS cohort, FAI was strongly positively associated with E-selectin (Pearson r = 0.718, p < 0.001, FDR-adjusted q = 0.027), and HOMA-IR was positively associated with ICAM-2 (r = 0.670, p = 0.001, q = 0.046). The FAI-E-selectin association remained robust in Spearman sensitivity analysis (ρ = 0.720, p < 0.001, q = 0.009), whereas the HOMA-IR-ICAM-2 association remained nominally significant (ρ = 0.579, p = 0.007) but not after FDR correction. A positive association between 24,25(OH)2D3 and tumour necrosis factor (TNF) was observed using Pearson correlation (r = 0.648, p = 0.002, q = 0.042), but was not retained after FDR correction in Spearman sensitivity analysis. Thus, non-obese women with PMOS did not demonstrate a generalised alteration in circulating endothelial biomarkers compared with matched controls. Hyperandrogenism, in particular, was robustly associated with endothelial activation, whereas associations with IR were less consistent and vitamin D metabolites showed no robust relationship with the endothelial biomarker profile. Collectively, the findings suggest that variation in endothelial activation in non-obese PMOS is more closely associated with metabolic dysfunction and androgen excess than with circulating vitamin D status, while the cross-sectional design precludes inference regarding causality.