Comparison of Clinical, Metabolic, and Hormonal Profiles in Obese Versus Lean Women With Polycystic Ovary Syndrome: A Systematic Review and Meta-Analysis.

Introduction: Obesity substantially influences clinical and metabolic manifestations of polycystic ovary syndrome (PCOS). However, differences in clinical, metabolic, and hormonal profiles between obese and lean women with PCOS have not been explored in large-scale meta-analyses.

Methods: Databases were searched from inception through 31 December 2025 for studies reporting clinical, metabolic, and hormonal profiles in obese and lean PCOS phenotypes. Random-effects meta-analyses in R yielded mean differences (MDs) or risk ratios (RRs) with 95% confidence intervals.

Results: Forty-two studies (N = 5363; 3078 obese, 2285 lean PCOS) met inclusion criteria. Menstrual irregularity (84.7% vs. 70.9%; RR 1.19), hirsutism (65.9% vs. 58.8%; RR 1.10), acne (44.9% vs. 36.8%; RR 1.25), acanthosis nigricans (52.7% vs. 22.6%; RR 1.94), and biochemical hyperandrogenism (45.6% vs. 31.8%; RR 1.49) were more prevalent in obese than in lean PCOS. Obese PCOS was associated with a less favourable cardiometabolic profile, with higher HOMA-IR (MD 2.03), fasting (MD 5.12 mg/dL), and 2-h (MD 9.91 mg/dL) plasma glucose, and greater risks of prediabetes (25.3% vs. 12.1%; RR 1.84), diabetes (8.6% vs. 2.7%; RR 2.58), and metabolic syndrome (51.2% vs. 11.9%; RR 3.59). Mean lipid levels were worse in obese PCOS, with higher total cholesterol, triglycerides, and LDL cholesterol, and lower HDL cholesterol. Nonetheless, lean PCOS exhibited substantial burdens of menstrual dysfunction, hyperandrogenism, insulin resistance, dysglycemia, and dyslipidemia.

Conclusion: Obesity in PCOS is associated with more severe reproductive, metabolic, and hormonal abnormalities, but lean PCOS still carries substantial reproductive and metabolic burdens. However, the observational design of the included studies and substantial heterogeneity across several outcomes warrant cautious interpretation.

Trial registration: The meta-analysis was registered with PROSPERO under registration number CRD420251240521. The review protocol summary is available on the PROSPERO website.