PCOS Supplements: What Has Trial Evidence, What Does Not, Ranked

Which PCOS supplements actually have trial evidence?

Four supplements — myo-inositol, NAC, berberine, and vitamin D — have been tested in a combined total of more than 60 randomised controlled trials in women with PCOS. Six others — omega-3, chromium, magnesium, spearmint, vitex, and DIM — have either a handful of small trials or, in DIM’s case, none at all. That gap is the single most useful fact on this page: most of what gets sold as “PCOS support” has never been tested on anyone with PCOS.

This is one page in the supplements and medications section built specifically to replace that guesswork with a ranking. The table below ranks all ten by the honest weight of that evidence, not by how they are marketed. “Multiple RCTs” means an actual body of randomised, controlled human trials in PCOS populations. “One small trial” means exactly that — promising, unconfirmed, and not enough to build a protocol on. “No PCOS trials” means the mechanism might be plausible, but nobody has tested it in this population.

What does “multiple RCTs” actually mean here?

Thirteen separate meta-analyses have now pooled the myo-inositol trials, and a 2026 umbrella review graded the results moderate-quality for testosterone, SHBG, HOMA-IR, and pregnancy and ovulation rates — moderate, not high, because individual trials are still small and inconsistently designed. Inositol reduced LH by 3.43 IU/L, raised SHBG by 36.72 nmol/L, and nearly tripled the odds of live birth (RR 2.29, 95% CI 1.07–4.93) compared with placebo or folic acid. That is the strongest supplement evidence base in PCOS by a wide margin, and it is still graded moderate rather than definitive.

NAC’s evidence is real but thinner. A 2025 meta-analysis of 22 studies and 2,515 women found NAC raised progesterone (SMD 0.95) and endometrial thickness (SMD 0.58) compared with placebo and other drugs, but found no significant effect on estradiol, SHBG, or FSH. It moves some hormonal markers; it does not move all of them.

Berberine’s twelve trials tell an interesting, split story. Compared with metformin, berberine produced a larger drop in total cholesterol (MD −0.44 mmol/L), waist circumference (MD −2.74 cm), and LH:FSH ratio (MD −0.44), with no increase in gastrointestinal side effects. But on live birth, the same review found “no solid evidence” of benefit, and berberine actually produced lower live-birth rates than letrozole (RR 0.61, 95% CI 0.44–0.82). Berberine looks like a genuine metabolic option and a poor fertility one — a distinction most retailers do not make.

The best supplements for insulin resistance, ranked

Four of the ten — myo-inositol, berberine, chromium, and vitamin D — target insulin resistance directly, and they are not interchangeable. Omega-3 also earns a place here: a 2023 meta-analysis of 11 RCTs (816 women) found n-3 fatty acids cut HOMA-IR by 0.45 and waist circumference by 2.76 cm, though only with more than eight weeks of use and with no effect on body weight itself.

Chromium’s ten trials produced a genuinely surprising finding: a 2025 meta-analysis reported chromium outperformed metformin on HOMA-IR and LH in 683 pooled women, alongside improvements in fasting insulin, triglycerides, and total antioxidant capacity. That is a bigger claim than the supplement usually gets credit for, and it deserves the same caveat as everything else on this list: ten trials from mixed-quality sources is not the same evidentiary weight as inositol’s thirteen meta-analyses of meta-analyses.

If you are choosing one supplement for insulin resistance and can only pick one, the trial evidence points to myo-inositol first, berberine second. For the full dose, ratio, and side-effect picture on inositol specifically, see inositol for PCOS — this page ranks the field, that one covers the detail.

What actually works for lean PCOS?

Lean PCOS changes which supplements make sense, because the mechanism most of this list targets — insulin resistance — is present in a smaller share of lean-phenotype cases than in higher-BMI PCOS, though it is far from absent; many lean-PCOS patients still test with elevated fasting insulin or HOMA-IR. Where insulin resistance is confirmed by testing, inositol and berberine’s evidence still applies regardless of body size — none of the trials behind Table 1 selected participants by BMI in a way that would exclude a lean-phenotype reader.

Where lean PCOS differs is in what is more likely to be driving symptoms instead: adrenal androgen excess or hypothalamic-pituitary-ovarian dysregulation rather than hyperinsulinaemia. For those patterns, the insulin-sensitising supplements on this list have less to act on, and the honest answer is that androgen-symptom evidence (spearmint, vitex) is also the weakest tier here. Testing before supplementing is not a formality for lean PCOS — it is the difference between choosing a supplement with a plausible target and choosing one at random.

Do any supplements help hirsutism or acne?

Two supplements have been tested against androgen symptoms specifically, and both sit near the bottom of the evidence tier for a reason. Spearmint tea’s best trial, a 30-day randomised study of 42 women, found free and total testosterone dropped significantly, but the objective Ferriman-Gallwey hirsutism score did not change — only participants’ self-reported impression did. An earlier 5-day trial in 21 hirsute women (12 with PCOS) found the same pattern: hormones moved, and the study was too short to show whether hair growth did.

Vitex has one PCOS-specific randomised trial of note, published in 2026: 60 women took 5.8 mg/day of standardised extract for 12 weeks, and the trial reported improved oxidative stress markers, insulin resistance, and a lower modified Ferriman-Gallwey score. One trial. No ovulation or pregnancy outcome has ever been measured for vitex in PCOS, and it has real interactions with dopamine-affecting and hormonal medications, which is worth knowing before you add it to anything else on this list.

Neither supplement approaches the effect size of a prescription anti-androgen. If hirsutism or hormonal acne is your main concern, spironolactone has a far larger trial base behind it — this page’s job is to tell you that honestly, not to sell you the tea first.

Which supplements have essentially no evidence?

DIM is the clearest case: a direct PubMed search for diindolylmethane and polycystic ovary syndrome returns zero randomised trials. DIM has human data in other contexts — estrogen metabolism, cervical cell changes — but nothing in PCOS. Every dosing claim you see attached to DIM for PCOS is extrapolated from a different population, not measured in this one.

Magnesium’s evidence actively conflicts with itself. A single small trial of 40 women found 250 mg of magnesium oxide daily for two months reduced insulin (p=0.036) and insulin resistance (p=0.032). But when six trials were pooled in 2025, the meta-analysis found no significant effect of magnesium on any cardiometabolic or hormonal marker in PCOS. That is not a supplement with weak evidence; it is a supplement with evidence pointing in two directions at once. Take it for sleep or migraine if that is your reason — do not take it expecting a PCOS-specific metabolic effect the pooled data does not support.

What to check before you buy any of these

Trial dose is the first thing to match, and most shelf products do not. Myo-inositol trials that produced the numbers in Table 1 used 2–4 g daily; a lot of retail capsules deliver 500 mg to 1 g per serving, which means the label dose and the trial dose are not the same product even when the ingredient name matches. The same gap shows up with chromium picolinate, where the trials behind the 200 mcg figure used that specific compound, not chromium chloride or chromium nicotinate at an unstated dose.

Third-party testing matters more here than on most supplement shelves, because PCOS supplements overlap heavily with weight-management and fertility marketing — two categories with a documented history of undisclosed stimulants and contamination. Look for a certificate of analysis or a third-party seal (USP, NSF, or Informed Choice) before price or packaging. None of the trials cited above tested a specific commercial brand; they tested a specific ingredient at a specific dose, and that is the only thing a label can actually promise you.

What this page will not tell you to do

This ranking will not tell you to take all ten. Stacking supplements without a target multiplies cost and side-effect risk without multiplying evidence — none of these trials tested a ten-supplement combination, so there is no data on what happens when you take one. It will not tell you a supplement can replace metformin for confirmed insulin resistance, replace a fertility specialist for anovulatory infertility, or replace spironolactone for hirsutism that is actually affecting your quality of life. It will not tell you a supplement produces the weight change some people hope for, either — see honestly what metformin itself achieves for weight loss specifically before assuming a supplement does better. And it will not tell you these are risk-free: NAC, berberine, and high-dose magnesium all carry gastrointestinal side effects, vitex interacts with hormonal medications and dopamine-affecting drugs, and “natural” is not the same claim as “tested.”

You may see PCOS referred to as polyendocrine metabolic ovarian syndrome (PMOS) after a 2026 global consensus of more than 50 organisations renamed it. Nothing about the evidence above changed with the name — this article uses PCOS because that is still what people search.

Affiliate disclosure: some links to products discussed elsewhere on this site are affiliate links, meaning we may earn a commission if you buy through them, at no extra cost to you. This does not change how supplements are ranked here — the ranking above is ordered strictly by trial evidence.

Common questions

Not sure which pattern of PCOS you’re actually working with? Take the quiz before you spend money on any of the ten supplements above — it changes which ones are worth trying first.