P463 - LBA_ECE_1389 - High-resolution metabolomics reveals a BMI-independent metabolic signature in polycystic ovary syndrome
Abstract
Polycystic ovary syndrome (PCOS) is a heterogeneous endocrine disorder with metabolic alterations not fully explained by obesity or androgen excess. To define a plasma metabolomic signature of PCOS independent of BMI and to evaluate associations with measures of androgen excess. Forty women with PCOS (mean age: 22.9 ± 2.7 years; mean BMI: 23.4 ± 4.6 kg/m2) and 40 age- and BMI-matched healthy controls underwent standardized clinical, biochemical, and hormonal assessment. Plasma metabolomic profiling was performed using integrated GC–MS and LC–MS platforms. Data were analyzed using principal component analysis (PCA), partial least squares–discriminant analysis (PLS-DA) with cross-validation and permutation testing, false discovery rate (FDR)-adjusted univariate analysis, receiver operating characteristic (ROC) analysis, pathway enrichment (MetaboAnalyst), and Spearman correlations with clinical parameters. PCA and PLS-DA demonstrated clear separation of PCOS and control groups. PLS-DA achieved high discrimination (accuracy = 0.95, R2 = 0.92, Q2 = 0.78) and was validated by permutation testing (P = .002). Thirty-one metabolites were significantly altered after FDR correction. Threitol, a four-carbon sugar alcohol, emerged as the top discriminatory metabolite (fold change = 5.86, FDR = 3.59×10⁻14), achieving near-perfect discrimination between PCOS and controls (multivariate ROC AUC = 0.993; 95% CI: 0.962-1.000), highlighting its potential as a novel metabolic marker. Pathway analysis revealed significant perturbations in branched-chain amino acid (BCAA: valine, leucine, isoleucine)) biosynthesis (P = 5.49×10⁻5, FDR = 0.004), alanine–aspartate–glutamate metabolism (a central amino acid–energy pathway) (P = 1.56×10⁻4, FDR = 0.006), and the citrate cycle (P = .00103, FDR = 0.028), indicating disrupted core energy metabolism. Threitol correlated inversely with total testosterone (r = −0.46), free androgen index (r = −0.56), and anti-Müllerian hormone (AMH) (all P < .001). Palmitoylcarnitine correlated positively with total testosterone (r = 0.35, P = .001). Other key metabolites, including BCAA and energy metabolites, showed no significant association with androgen or insulin resistance measures. High-resolution metabolomics identifies a robust, BMI-independent metabolic signature in PCOS characterized by BCAA and energy metabolism alterations. Selected metabolites show targeted associations with androgen levels, supporting refined metabolic phenotyping in PCOS.