Condition: Polycysteus-ovariumsyndroom (PCOS)

Against placebo or no treatment, metformin improved ovulation (OR 2.55, 95% CI 1.81 to 3.59, 14 studies, 701 women), clinical pregnancy (OR 1.93, 95% CI 1.42 to 2.64, 9 studies, 1027 women) and menstrual frequency (OR 1.72, 95% CI 1.14 to 2.61), with a possible improvement in live birth (OR 1.59, 95% CI 1.00 to 2.51, low-quality). Gastrointestinal side effects were much more common (OR 4.76). Measured in: 48 RCTs (4451 women) of insulin-sensitizing drugs for anovulatory PCOS, 42 involving metformin (4024 women). The evidence quality ran from very low to moderate, and the live-birth signal only just reached the line (lower CI 1.00). For fertility specifically, ovulation-induction agents such as letrozole outperform metformin used alone; metformin's clearest role is metabolic and as an adjunct.

Over up to five treatment cycles, women receiving letrozole had more cumulative live births than those on clomiphene (103 of 374, 27.5%, vs 72 of 376, 19.1%; rate ratio 1.44, 95% CI 1.10 to 1.87, P=0.007) and a higher cumulative ovulation rate (61.7% vs 48.3% of cycles, P<0.001). There was no significant difference in pregnancy loss or overall congenital anomalies. Measured in: 750 women aged 18 to 40 with PCOS (modified Rotterdam criteria) and a fertile male partner, in a double-blind US multicentre trial. This settled letrozole as the first-line ovulation-inducer for PCOS. It was studied in women with a patent tube, a normal uterine cavity and a partner with adequate sperm, so it addresses the ovulation barrier specifically, not every cause of infertility. Letrozole is prescribed and cycle-monitored.

In a factorial trial, live-birth rates did not differ between active and control acupuncture (100 of 458, 21.8%, vs 105 of 468, 22.4%; difference -0.6%, 95% CI -5.9% to 4.7%), while clomiphene clearly beat placebo (28.7% vs 15.4%). There was no interaction between acupuncture and clomiphene. Measured in: 1000 Chinese women with PCOS randomized 1:1:1:1 to active or control acupuncture, each with clomiphene or placebo, across 21 sites. This large, well-conducted trial does not support acupuncture as an infertility treatment in PCOS, and it is included so the fertility claim on this page stays correctly sized. Acupuncture has a long traditional use for menstrual regulation and wellbeing, and this result speaks to the live-birth endpoint it was designed to test, not to those other aims.

Added to standard care, berberine improved the ovulation rate (RR 1.41, 95% CI 1.26 to 1.60), clinical pregnancy rate (RR 1.96, 95% CI 1.59 to 2.41) and endometrial thickness (weighted mean difference 1.62 mm), and lowered luteinizing hormone and total testosterone, compared with standard care alone. Measured in: 10 RCTs involving 713 women with PCOS, testing berberine as adjuvant therapy for reduced fertility. The included trials were mostly small and drawn substantially from Chinese-language databases, and the authors call for further trials before firm conclusions. Berberine acts as a metabolic agent, can interact with prescription medicines, and belongs with a practitioner and a traceable product.

Lifestyle intervention (diet, exercise or behavioral, or a combination) reduced the free androgen index by a mean of 1.11 (95% CI -1.96 to -0.26, 6 RCTs, N=204), lowered weight by a mean of 3.7 lb (1.68 kg) (95% CI -2.66 to -0.70, 9 RCTs, N=353) and reduced BMI by 0.34 kg/m2 (95% CI -0.68 to -0.01, 12 RCTs, N=434), compared with minimal or no treatment. Measured in: 15 randomized trials with 498 women with PCOS, comparing lifestyle treatment against minimal intervention or no intervention. The review graded all of these findings low-quality, mainly from risk of bias and heterogeneity, and no included trial reported live birth, miscarriage or menstrual regularity, so this measures the hormonal and weight changes, not the pregnancy outcomes. The absolute changes are modest averages: for a given woman the effect of losing weight can be considerably larger than the pooled mean.

Lifestyle plus metformin, against lifestyle with or without placebo, produced a lower BMI (mean difference -0.73 kg/m2, 95% CI -1.14 to -0.32) and more menstrual cycles (mean difference 1.06, 95% CI 0.30 to 1.82) at 6 months. There was no clear difference in insulin-resistance markers, glucose, lipids, blood pressure, hyperandrogenism or quality of life between the two. Measured in: 12 RCTs comprising 608 women with PCOS diagnosed by Rotterdam criteria, at any age or BMI. The trials were small, short (mostly 6 months) and at risk of bias, and metformin is well known to be hard to stay on because of stomach side effects, which the pooled figures do not capture. The benefit sits on top of lifestyle, not in place of it.

Compared with usual care, exercise reduced fasting insulin (mean difference -2.44 uIU/mL, 95% CI -4.24 to -0.64), HOMA-IR (-0.57, 95% CI -0.99 to -0.14), total cholesterol, LDL cholesterol and triglycerides, and improved VO2 max, waist circumference and body-fat percentage. Post-intervention analyzes also showed lower BMI (-1.02 kg/m2). Measured in: 18 randomized or quasi-randomized trials (27 papers) of exercise, or exercise plus diet, in women with PCOS. The review graded most of these effects low or very-low quality, the confidence intervals are wide, and several results were sensitive to adding or removing a single trial. Gains were largest in overweight participants and in supervised, aerobic programs. It could not separate exercise from diet where the two were combined.

In the review that informed the 2023 international PCOS guidelines, myo-inositol or D-chiro-inositol (DCI) improved some metabolic measures, with a possible benefit of DCI on ovulation, while showing no clear effect on several other outcomes. Metformin may improve waist-hip ratio and hirsutism more than inositol, with likely no difference in reproductive outcomes, and inositol caused far fewer gastrointestinal side effects than metformin. Measured in: 30 trials (n=2230; 1093 on inositol, 1137 controls), 19 pooled in meta-analyzes, in women with PCOS. The authors concluded the evidence supporting inositol in PCOS is limited and inconclusive, and framed the choice as one for shared decision-making. The reasonable read is that inositol is a well-tolerated first option whose benefit is not certain, not a proven equivalent to metformin across the board.

Over 12 weeks, both a 40:1 myo-inositol plus D-chiro-inositol combination and metformin significantly improved insulin sensitivity (HOMA-IR, p<0.001), SHBG (p=0.021), ovarian volume (p<0.001) and menstrual regularity (p=0.002), along with BMI, quality of life and perceived stress. Metformin was slightly better on some insulin and endocrine markers. Measured in: 60 women with PCOS (Androgen Excess Society criteria) randomized 1:1 to inositol combination or metformin for 12 weeks. This is a single small trial of 60 women over 12 weeks, so it shows both treatments work in the short term, not settling which is better; the authors themselves note the metformin edge may reflect the phenotype mix in that arm. Menstrual regularity and stress improved on both, which matches inositol's appeal as the better-tolerated option.

High dietary fiber and low-glycemic-index eating significantly reduced fasting glucose and insulin resistance in women with PCOS. Both patterns lowered triglycerides and LDL cholesterol, fiber also raised HDL cholesterol, and both raised sex-hormone-binding globulin (SHBG) and reduced the free androgen index. Measured in: A meta-analysis of randomized clinical trials of high-fiber and low-glycemic-index or low-glycemic-load dietary interventions in women with PCOS. Results were reported as standardized mean differences and the trials were heterogeneous, and the authors call for more high-quality studies and personalized, shared-decision dietary plans. The direction is consistent and points at the insulin resistance underneath the condition, which is why carbohydrate quality is a food-first lever.

Spironolactone at 100 mg daily reduced the Ferriman-Gallwey hair score more than finasteride (mean difference -2.43, 95% CI -3.29 to -1.57) and more than cyproterone acetate (-1.18, 95% CI -2.10 to -0.26) in hirsutism, and showed a positive trend in women with PCOS specifically. A 50 mg dose showed no significant difference from metformin on hair score, testosterone or HOMA-IR. Measured in: 24 RCTs of spironolactone in women with PCOS or idiopathic hirsutism. The clearest hair-score benefit was in idiopathic hirsutism; within PCOS the effect was a positive trend, not a decisive result, and spironolactone did not change LH, FSH, menstrual cyclicity, BMI or HOMA-IR. It is an anti-androgen prescribed and monitored by a clinician, and must not be used while trying to conceive.

Across hirsutism treatments, flutamide 250 mg twice daily reduced Ferriman-Gallwey scores more than placebo (mean difference around -7.4), and combined oral contraceptives reduced scores from baseline, though the comparison between two pill types was not statistically clear (mean difference -1.84, 95% CI -3.86 to 0.18). Treatment courses ran six to twelve months. Measured in: 157 RCTs comprising 10,550 women (mean age 25), the majority with PCOS, on interventions for hirsutism excluding laser and light therapy alone. Evidence quality was moderate to very low, most trials were at high risk of bias mainly from lack of blinding, and patient-reported improvement and quality of life were addressed in few studies. Hirsutism treatment works slowly because hair cycles are slow, so results are judged over months. Flutamide carries a liver-toxicity risk and is used cautiously.

Vitamin D supplementation reduced fasting blood glucose (mean difference -2.91 mg/dL, 95% CI -4.78 to -1.04), insulin (-1.98 uIU/mL, 95% CI -3.32 to -0.64), triglycerides, total cholesterol, VLDL and LDL cholesterol in women with PCOS, with no significant change in HDL cholesterol. Measured in: 13 RCTs with 691 women with PCOS. These are modest average changes in surrogate metabolic markers; the review did not assess cycles, fertility or long-term outcomes, and the most effective dose and the durability are not settled. Correcting a measured vitamin D deficiency is reasonable on its own terms; this is a supporting metabolic lever, not a core PCOS treatment.