Study Type
Meta-analysis
Population
Women with PCOS, mostly with overweight or obesity (mean BMI above 25 kg/m²). Age: not reported.
Size
15 studies in the review; 10 pooled per the text (Section 3.1); 335 on ketogenic diet and 192 on other diets, 527 total (calculated; Table 1); largest pooled outcomes 250 participants before vs after (weight; Table 2) and 364 vs other diets (HOMA-IR; Table 3).
Who Was Studied
PCOS definition used by included studies: not reported; ages not reported, with subgroups split at mean age 30 (Section 2.5); Europe 6, Asia 5, Americas 4 studies (Section 3.1), from Italy, China, Iran, Pakistan, Mexico and the USA (Table 1); any design eligible, with case reports, case series and IVF-outcome studies meant to be excluded from pooling (Section 2.2); diets mostly under 20–50 g carbohydrate a day for 45–180 days, follow-up up to 8 months (Table 1; Section 4).
Main Findings
(1) Before vs after the ketogenic diet, BMI fell by a mean of 3.38 kg/m² (95% CI 2.53 to 4.23; 9 studies, 210 participants; I² 0%) and weight by 10.77 kg (95% CI 8.73 to 12.81; 9 studies, 250 participants; I² 0%) (Section 3.3; Table 2; Fig. 2a–b). (2) Compared with other diets, the ketogenic diet was associated with lower BMI (MD −1.65 kg/m², 95% CI −2.76 to −0.55; I² 0%), weight (MD −4.98 kg, 95% CI −9.05 to −0.91; 3 studies, 264 participants; I² 7%) and waist circumference (MD −6.81 cm, 95% CI −10.63 to −2.99; 4 studies, 294 participants; I² 12%) (Section 3.4; Table 3; Fig. 3a–c). (3) Before vs after, LH fell by 4.07 IU/L (95% CI 3.36 to 4.79; I² 0%), testosterone by 0.31 nmol/L (95% CI 0.18 to 0.44; I² 48%) and HOMA-IR by 2.43 (95% CI 1.16 to 3.69; I² 95%) (Table 2); compared with other diets, HOMA-IR was lower (MD −1.71, 95% CI −2.98 to −0.43; 5 studies, 364 participants; I² 90%) and testosterone did not differ significantly (MD −0.36 nmol/L, 95% CI −0.90 to 0.18; I² 94%) (Table 3); clinical importance not established. Also reported: menstrual cycle duration changed by 26.06 days before vs after (95% CI 2.28 to 49.85; 2 studies, 20 participants; I² 68%) (Table 2; Fig. 2k); triglycerides were lower than with other diets (MD −0.10 mmol/L, 95% CI −0.19 to −0.01; I² 9%) (Table 3); adverse events were reported in only 2 studies and were mild and short-lived (Section 4.1, p. 9); trial sequential analysis found too little data to cross the benefit or futility boundaries (Section 3.4).
Limitations
(1) Authors reported: most evidence comes from observational studies with moderate to high risk of bias, confounders, unmatched comparators and differing diet protocols (Section 4.2).
(2) Authors reported: there were few studies and participants and high heterogeneity, hormone tests were timed inconsistently or the timing was not reported, and exercise was not accounted for (Section 4.2).
(3) Authors reported: all participants had BMI above 25, so results do not apply to lean PCOS, and diets were followed for up to 8 months only (Sections 4, 4.2).
(4) Lucky Vera observation: the main analysis has no control group, so effects of the diet cannot be separated from weight loss in general; some comparison groups also took the combined contraceptive pill (Table 1); and no outcome had enough studies (more than 10) for the planned publication-bias check (Section 2.5).
What Does the Study Not Establish
(1) Cause and effect: the main analysis has no control group, and the comparison with other diets mixes randomized and non-randomized studies.
(2) Whether the changes are clinically meaningful or improve ovulation, fertility or symptoms; fertility outcomes were not pooled (Section 4.1).
(3) Long-term safety or effects beyond 8 months, or effects in lean PCOS.
(4) Whether any benefit comes from ketosis itself rather than from weight loss.
Author’s Conclusion
The authors say the ketogenic diet may be a promising short-term option for women with PCOS and BMI above 25, with potential to improve body measures, insulin resistance and some reproductive hormones; they caution that heterogeneity and generally low study quality limit confidence, and call for larger long-term randomized trials, including in lean PCOS (Section 5).
Arsenaki E, Stathi D, Triantafyllidis KK, et al. The effects of ketogenic diet on polycystic ovary syndrome: A systematic review and meta-analysis. Clin Nutr. 2026;56:106535. doi:10.1016/j.clnu.2025.11.019