Clustering analysis uncovers four reproducible PCOS subtypes with distinct clinical outcomes

Back to news list

Source: Nature Medicine

Original: https://www.nature.com/articles/s41591-025-04040-8...

Published: 2025-11-25

An international study published in Nature Medicine identified four distinct subtypes of polycystic ovary syndrome (PCOS) by analysing clusters of clinical data.1[2] The research analysed data from more than 11,900 women diagnosed with PCOS over a period of 6.5 years, using sophisticated analysis of hormonal and metabolic parameters.[3] The subtypes identified are HA-PCOS (high androgen), OB-PCOS (obese), SHBG-PCOS (high sex hormone-binding globulin), and LH-PCOS (high luteinizing hormone).[2] Each subtype exhibits distinct biochemical profiles and clinical characteristics, with SHBG-PCOS having the most favorable metabolic profile with the lowest incidence of diabetes and hypertension.3 The study was validated in different ethnic populations with high cluster stability (Jaccard score above 0.79).1][2] These findings form the basis for a personalized approach in the management of PCOS and allow for the development of more accurate diagnostic criteria and treatment strategies tailored to individual subtypes[1][2].