Serum phoenixin-14 and phoenixin-20 concentrations in polycystic ovary syndrome: Associations with anti-Müllerian hormone and diagnostic utility.

Abstract

Polycystic ovary syndrome (PCOS) is a complex metabolic disorder characterized by altered peptide signaling that may contribute to reproductive dysfunction, insulin resistance, and metabolic dysregulation. This case-control study investigated serum phoenixin (PNX) isoforms, phoenixin-14 (PNX-14) and phoenixin-20 (PNX-20), in women with PCOS to evaluate their associations with biochemical and hormonal markers, including anti-Müllerian hormone (AMH), and assess their diagnostic utility. Serum PNX-14 and PNX-20 levels were measured in 90 women with PCOS and 60 age-matched healthy controls. Demographic, metabolic, and hormonal parameters were compared between the groups. Statistical analyses included receiver operating characteristic (ROC) analysis, logistic regression, correlation analysis, principal component analysis (PCA)-based clustering, mediation analysis, and exploratory Random Forest analysis. Results indicated that PNX-14 and PNX-20 levels were significantly higher in women with PCOS compared to controls (p < 0.001). Both isoforms correlated positively with AMH, insulin, homeostatic model assessment for insulin resistance (HOMA-IR), and luteinizing hormone (LH), while demonstrating negative correlations with sex hormone-binding globulin (SHBG) and high-density lipoprotein (HDL) cholesterol. ROC analysis showed moderate discriminatory performance, with area under the curve (AUC) values of 0.834 for PNX-14 and 0.819 for PNX-20. Mediation analysis suggested an associative pattern among PNX isoforms, AMH, and PCOS status. PCA-based clustering identified two distinct phenotypic profiles. Random Forest analysis ranked AMH, SHBG, fasting insulin, and body mass index (BMI) as the most influential variables, with PNX isoforms demonstrating lower importance scores. In conclusion, circulating PNX-14 and PNX-20 were elevated in PCOS and associated with key reproductive and metabolic parameters, suggesting a potential complementary role within multi-marker assessment frameworks. However, given the cross-sectional design and exploratory nature of the analyses, these findings should be considered hypothesis-generating rather than causal. Future longitudinal, mechanistic, and externally validated multicenter studies are essential to further elucidate these relationships.