Causal adiposity and clinical validation of regional fat distribution in polyendocrine metabolic ovarian syndrome (PMOS): a multi-modal analysis integrating Global Burden of Disease (GBD) 2021, Mendelian randomization, and machine learning
Authors' conclusion
Adiposity, particularly regional fat distribution, appears to be a clinically relevant determinant of PMOS. The convergence of global epidemiological patterns, genetic evidence, and clinical prediction supports the importance of body-composition phenotypes beyond conventional obesity measures in understanding PMOS risk.
Study context
- Population
- Women with PMOS (previously polycystic ovary syndrome (PCOS)); global burden (GBD 2021), genetic data and a clinical cohort with detailed body-composition measurements
- Outcome
- PMOS burden, causal adiposity traits and machine-learning (ML) prediction
Methodological detailsmachine-learning models internally validated in a clinical cohort, SHAP interpretation
- Machine-learning models internally validated in a clinical cohort
- SHAP interpretation
Main finding
As reported in the abstract
- Women affected by PMOS globally, 2021
- ≈69.5 million
- Age-standardized prevalence rate
- 1,757.8 per 100,000
- Highest modeled incidence age group
- 10-14-year
- Infertility cases attributable to PMOS worldwide
- ≈12.5 million
- Mendelian randomization (MR): BMI and PMOS, OR
- 2.60 (1.84-3.66) (95% CI)
- MR: body fat percentage and PMOS, OR
- 3.32 (1.97-5.58) (95% CI)
- MR: trunk fat mass and PMOS, OR
- 2.54 (1.79-3.60) (95% CI)
- MR: left-leg fat mass and PMOS, OR
- 3.44 (2.19-5.40) (95% CI)
- XGBoost best predictive performance, testing set AUC
- 0.701
Observational data: associations, not effects.
Original publication
Causal adiposity and clinical validation of regional fat distribution in PMOS: a multi-modal analysis integrating GBD 2021, Mendelian randomization, and machine learning. Int J Obes (Lond). Published 7 September 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →