Ketogenic diet as a therapeutic intervention for biochemical hyperandrogenism in PCOS women with obesity or overweight: a meta-analytic study

Purpose: To assess whether the Ketogenic Diet (KD) is associated with improvements in biochemical markers of hyperandrogenism in women diagnosed with Polycystic Ovary Syndrome (PCOS) with obesity or overweight. Methods: A systematic search was conducted in five databases: MEDLINE (via PubMed), Embase, Cochrane Central Register of Controlled Trials (CENTRAL), Scopus, and Web of Science. Both RCTs and pre-post studies providing data for direct extraction or calculation of mean (SD) changes in the investigated outcomes, published until March 1, 2025, were considered. Protocols included "Very Low" and "Low" Calories KD, and modified KD regimens. The primary outcome was the effect of the KD on total testosterone (TT), Free T (FT) and Sex Hormone Binding Globulin (SHBG) compared to baseline. Results: The KD was found to be associated with a statistically significant reduction in TT (n = 290, Mean Difference [MD] -0.35, CI -0.46; -0.24) and a significant increase in SHBG levels (n = 144, MD 19.50, CI 6.90; 32.11), independently of study design. When considering only RCTs (4 studies), the KD showed a moderate, statistically significant benefit on TT levels compared to the control intervention, with a high heterogeneity (n = 115 vs. 117, SMD -0.284, CI -0.543, -0.035; p = 0.031). Neither the duration of the KD protocol, the change in BMI, the change in fasting insulin levels nor the change in LH/FSH ratio were able to significantly affect the change in TT levels observed over time. In addition, the KD associated with a significant reduction in LH levels (n = 275, MD -3.48, CI -4.71; -2.26) and in the LH/FSH ratio (n = 275, MD -0.99, CI -1.44; -0.54). Conclusion: Our meta-analysis indicates a reduced biochemical hyperandrogenism-particularly a reduction in TT-and a reduction in the LH/FSH ratio among women with PCOS following a KD. Our findings also suggest a specific effect of ketosis on ovarian function acting alongside metabolic improvements, but a direct causal link cannot be definitively established. Owing to the low quality and heterogeneous nature of the available studies, larger, high-quality RCTs are essential to consolidate this evidence and determine clinical efficacy.