Machine Learning-Based Steroid Metabolome Analysis Reveals Three Distinct Subtypes of Polycystic Ovary Syndrome and Implicates 11-Oxygenated Androgens as Major Drivers of Metabolic Risk

    May 2023 in “ Endocrine Abstracts ”
    Eka Melson, Thais P. Rocha, Roland J. Veen, Lida Abdi, Tara McDonnell, Veronika Tandl, James Hawley, Laura B. L. Wittemans, Amarah V. Anthony, Lorna C Gilligan, Fozia Shaheen, Punith Kempegowda, Caroline D. T Gillett, Leanne Cussen, Cornelia Missbrenner, Fannie Lajeunesse‐Trempe, Helena Gleeson, Rees D. Aled, Lynne Robinson, Channa Jayasena, Harpal Randeva, Georgios K. Dimitriadis, Larissa Garcia Gomes, Alice Sitch, Eleni Vradi, Angela E. Taylor, Michael O’Reilly, Barbara Obermayer-Pietsch, Michael Biehl, Wiebke Arlt
    TLDR PCOS has three subtypes, with 11-oxygenated androgens increasing metabolic risk.
    The study analyzed 488 women with polycystic ovary syndrome (PCOS) to identify subtypes based on androgen profiles and their associated metabolic risks. Using machine learning, researchers identified three distinct subgroups: one with gonadal-derived androgen excess (GAE), another with adrenal-derived androgen excess (AAE), and a third with mild androgen excess (MAE). The AAE cluster, comprising 21.7% of participants, showed the highest rates of hirsutism and female pattern hair loss, as well as increased insulin resistance compared to the other clusters. This suggested that 11-oxygenated androgens, prevalent in the AAE cluster, were major contributors to metabolic risk in PCOS.
    Discuss this study in the Community →