L26/O-011 Metabolomic profiling of Polycystic Ovary Syndrome (PCOS): insights from the UK Biobank
Human Reproduction · 01.07.2026 · 41 (Supplement_1)
DOI: 10.1093/humrep/deag083.011 ↗Kurz gefasst
Fachartikel in der Zeitschrift Human Reproduction, online erschienen am 08.07.2026. Verfasst von S. Vagios und 6 weiteren Autor:innen.
Abstract (Originalsprache)
Study question Do women with PCOS exhibit distinct metabolomic profiles that vary across body mass index (BMI) and free androgen index (FAI) strata? Summary answer Polycystic ovary syndrome (PCOS) was associated with widespread metabolomic differences; however, patients with PCOS and normal BMI showed only minimal disturbances at the metabolomic level. What is known already PCOS is associated with increased cardiometabolic risk, including dyslipidemia, insulin resistance, and hepatic steatosis. However, prior metabolomic studies are often limited by small sample sizes and confounded by age and adiposity, hindering identification of PCOS-specific metabolic effects. PCOS is a heterogeneous disorder, and clustering analyses have described phenotypes with differential metabolic risk, commonly referred to as putative metabolic and reproductive subtypes. The metabolomic features underlying these phenotypes remain incompletely characterized. Stratification by BMI may approximate a metabolic phenotype, while stratification by FAI (hyperandrogenic status) may capture features of a reproductive phenotype. Study design, size, duration Cross-sectional analysis of UK Biobank participants with available Nuclear Magnetic Resonance (NMR)-based metabolomic profiling. Data were collected between 2006 and 2010. Women with PCOS (n = 1,093) were compared with controls (n = 5,465) selected through 1:5 nearest-neighbor matching by age and BMI. Sub-analyses were performed within distinct strata based on BMI (<25 & ≥25 kg/m²) and hyperandrogenic status (FAI ≤5 & >5%) to evaluate effect modification and phenotype-specific metabolomic signatures. Participants/materials, setting, methods The UK Biobank includes women over the age of 40. Participants with available NMR-based metabolomic profiling were included. Associations between PCOS and 251 metabolomic traits spanning multiple metabolic domains were assessed using linear regression. Unadjusted models were followed by multivariable models adjusting for age, BMI, smoking status, exercise, and FAI. Sub-analyses were conducted within BMI strata (<25 vs ≥ 25 kg/m²) and hyperandrogenic status (FAI ≤5 vs > 5%) to evaluate effect modification and phenotype-specific metabolomic signatures. Main results and the role of chance Participants with PCOS and controls were well-balanced for age (mean 48.3 years in both groups) and BMI (mean 30.0 kg/m² in both groups), with similar racial distribution (predominantly White). Most participants were premenopausal. In the total cohort, PCOS was associated with widespread metabolomic differences compared with controls. In unadjusted analyses, 139/251 metabolomic traits differed by PCOS status. After adjustments, 134/251 traits remained associated with PCOS, demonstrating robust persistence of the PCOS-related metabolomic signature. Subgroup analyses revealed marked heterogeneity by BMI. Among women with BMI <25kg/m2, PCOS was associated with substantially fewer metabolomic differences (42/251 traits unadjusted; 11/251 after adjustment), consistent with an attenuated metabolic phenotype in lean PCOS. In contrast, among women with BMI ≥25 kg/m², PCOS was associated with broad metabolomic disruption (136/251 traits unadjusted; 108/251 after adjustment). Stratification by hyperandrogenic status showed a more selective pattern: in women with FAI >5%, 51/251 traits differed unadjusted and 61/251 after adjustment, whereas in women with FAI ≤5%, 93/251 traits differed unadjusted and 117/251 after adjustment. Across analyses, the most consistent PCOS-associated differences involved lipid-related measures, particularly lipoprotein particle concentrations and subclass distributions (VLDL/IDL/LDL/HDL), cholesterol fractions, triglyceride-related traits, and fatty acid composition, while non-lipid traits showed fewer and less consistent differences. Limitations, reasons for caution The cross-sectional design limits causal interpretation and the mean age of included subjects may limit generalizability to younger populations. Additionally, the diagnostic criteria used for PCOS diagnosis were not available. Lastly, the availability of data in a biobank limits the possible confounding factors that can be integrated into subsequent analysis. Wider implications of the findings In a large population cohort matched for age and BMI, PCOS was associated with widespread metabolomic alterations. However, lean PCOS showed only minimal metabolomic impact. These findings support phenotypic risk stratification and may inform individualized metabolic screening and counselling in women with PCOS, as well as motivate future research. Trial registration number No
Zitieren
APA
Vagios, S., Mermin-Bunnell, A., Arkfeld, C., Sassin, A., Souter, I., Polli, F., & Pepin, D. (2026). L26/O-011 Metabolomic profiling of Polycystic Ovary Syndrome (PCOS): insights from the UK Biobank. Human Reproduction, 41(Supplement_1). https://doi.org/10.1093/humrep/deag083.011
Vancouver
Vagios S, Mermin-Bunnell A, Arkfeld C, Sassin A, Souter I, Polli F, et al. L26/O-011 Metabolomic profiling of Polycystic Ovary Syndrome (PCOS): insights from the UK Biobank. Human Reproduction. 2026;41(Supplement_1). doi:10.1093/humrep/deag083.011
BibTeX
@article{Vagios2026,
author = {Vagios, S. and Mermin-Bunnell, A. and Arkfeld, C.K. and Sassin, A.M. and Souter, I. and Polli, F. and Pepin, D.},
title = {L26/O-011 Metabolomic profiling of Polycystic Ovary Syndrome (PCOS): insights from the UK Biobank},
journal = {Human Reproduction},
year = {2026},
volume = {41},
number = {Supplement_1},
doi = {10.1093/humrep/deag083.011}
}
EUVIMED