.png){kind=icon}
.png){kind=icon}
Return
Data-driven endocrine–metabolic phenotypes in young women with polycystic ovary syndrome and associations with cardiometabolic risk markers
.png){kind=icon}
OA
AI
N
K
M
G
A
DOI：10.1007/s40618-026-02998-x.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Polycystic ovary syndrome (PCOS) is a heterogeneous endocrine disorder associated with reproductive dysfunction and increased cardiometabolic risk. Conventional diagnostic phenotypes may not adequately capture endocrine-metabolic variability, particularly in young women. We aimed to identify data-driven endocrine–metabolic phenotypes in young women with PCOS and assess their associations with selected cardiometabolic risk markers. This cross-sectional study included 1032 women aged 16–25 years with PCOS diagnosed according to the Rotterdam criteria, selected from 1300 consecutive patients evaluated at a tertiary gynecological endocrinology center in Poland between 2018 and 2025. Principal component analysis (PCA) followed by Gaussian mixture model clustering was used to identify latent endocrine–metabolic subgroups. Cardiometabolic outcomes included impaired glucose tolerance (2-hour plasma glucose ≥ 140 mg/dL), elevated triglyceride/high-density lipoprotein cholesterol (TG/HDL) ratio, elevated non–high-density lipoprotein cholesterol (≥ 130 mg/dL), and a composite endpoint. PCA retained 10 components explaining 81.9% of total variance. The optimal clustering solution identified two subgroups: a predominant mixed endocrine–metabolic subgroup (n = 954; 92.4%) and a smaller thyroid/autoimmune-enriched subgroup (n = 78; 7.6%) characterized by higher anti-thyroid peroxidase antibody and thyroid-stimulating hormone levels. Cluster assignment certainty within the selected model was high, although sensitivity analyses suggested that the latent structure was partially driven by thyroid/autoimmune-related variables. The thyroid/autoimmune-enriched subgroup showed a higher prevalence of an elevated TG/HDL ratio, whereas differences in glucose tolerance and non–high-density lipoprotein cholesterol were less pronounced. These findings suggest the presence of exploratory endocrine–metabolic subgroup structure in young women with PCOS, including a thyroid/autoimmune-enriched subgroup associated with modest differences in lipid-related cardiometabolic risk markers. The observed heterogeneity extends beyond conventional diagnostic classifications, although external validation and longitudinal studies are required to determine clinical relevance and reproducibility. This study was not registered as a clinical trial, as it was an observational cross-sectional analysis.
Keywords:
Polycystic ovary syndrome
Endocrine phenotypes
Machine learning
Clustering
Thyroid autoimmunity
Cardiometabolic risk
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
J
IF:
3.5
Papers:
6.0K
Citations:
8.9K
Organization
Cited Papers