Myo-Inositol: The Insulin Sensitiser for PCOS and Beyond

Key Takeaways

Myo-inositol occupies an unusual position among insulin sensitisers. It is a sugar alcohol your kidneys already make, sold at four grams a day, and in one specific population — women with polycystic ovary syndrome — it has been tested against placebo and against metformin in enough randomised trials to be pooled twice. In that population it lowers fasting insulin, lowers the HOMA index, raises SHBG, and restores regular cycles at rates the meta-analyses call non-inferior to the drug.

That is a real result, and it deserves to be stated plainly before the qualifiers. The qualifiers are also real: the founding trials came from a single group and one has been retracted; the fertility outcome that matters most, live birth, is unproven; the pregnancy-prevention claim was tested properly in 2025 and failed; and outside the ovary, the case rests on mechanism rather than outcome.

This article puts both halves together, then explains how the 40:1 ratio, the dose and the population determine whether myo-inositol is worth taking.

What Myo-Inositol Does at the Receptor

Insulin binding its receptor does not directly move glucose. It sets off a cascade, and one branch of that cascade releases inositol phosphoglycan second messengers from the membrane. These are built from two inositol isomers. The myo-inositol family activates enzymes involved in glucose uptake and oxidation; the D-chiro-inositol family activates glycogen synthesis. A cell short of either precursor receives insulin's signal and transmits it poorly.

The body makes D-chiro-inositol from myo-inositol using an epimerase that is itself insulin-dependent. In an insulin-resistant tissue, that conversion falls — which is why D-chiro-inositol was, for a time, the isomer of interest.

Then came the ovarian paradox. In 2011 an Italian group proposed that the ovary is the exception: in PCOS it remains insulin-sensitive while the rest of the body does not, so the epimerase there runs hard, over-converting myo-inositol into D-chiro-inositol. The result is a local myo-inositol deficiency in the very tissue that needs it for FSH signalling and oocyte quality. If that is right, giving a woman with PCOS large doses of D-chiro-inositol makes her ovary worse, not better — and the clinical data on ratios, covered below, support the idea.

The PCOS Evidence

Three things stand out. First, the insulin result is robust: in the 2017 analysis, trial sequential analysis showed the cumulative evidence crossed the monitoring boundary, which is the statistical way of saying more trials would not overturn it. Second, the androgen effect is smaller and slower — testosterone only trended down in 2017 and reached significance in the larger 2023 pool, while the SHBG rise needed at least 24 weeks to appear. Third, cycle regularity is the endpoint a woman can actually observe, and the 79% increase in the odds of a regular cycle is the number that justifies a trial of it.

Against Metformin

Metformin is the gold-standard insulin sensitiser in PCOS and its main limitation is the gut. That is the framing of the head-to-head meta-analysis, and its finding was that across eight trials myo-inositol and metformin did not differ on any metabolic or hormonal parameter measured: BMI (SMD 0.16, P = 0.24), fasting insulin (SMD 0.00, P = 0.97), fasting glucose (P = 0.60), HOMA index (SMD 0.09, P = 0.50), LH/FSH (P = 0.37).

That is not a demonstration that inositol is as good as metformin in every respect — these were modest trials, heterogeneity was moderate, and non-inferiority on surrogate markers is not the same as equivalence on pregnancy or diabetes prevention. But for a woman who has stopped metformin because of nausea and diarrhoea, it establishes that the supplement route is not a large step down on the markers that matter for PCOS management.

Berberine sits in the same conversation. Its own head-to-head with metformin, in type 2 diabetes rather than PCOS, produced matching HbA1c reductions, and it has been compared with inositol and metformin in the same network meta-analysis of oral insulin sensitisers in PCOS. The two are different levers — berberine activates AMPK, inositol supplies a second-messenger precursor — and they are the first two layers of The Insulin Sensitivity Protocol.

The 40:1 Ratio

The ratio question was settled, as far as it has been, by a seven-arm trial of 56 women with PCOS — eight per arm — given 2 g of inositols twice daily for three months at ratios of D-chiro-inositol alone, 1:3.5, 2.5:1, 5:1, 20:1, 40:1 and 80:1 myo:D-chiro. The primary outcome was ovulation.

The 40:1 ratio — which matches healthy plasma — was best at restoring ovulation and normalising the secondary markers (FSH, LH, SHBG, oestradiol, free testosterone, insulin, HOMA index). The other ratios were less effective, and the authors were specific that activity fell as the ratio shifted toward D-chiro-inositol. That is the ovarian paradox in clinical form.

Eight women per arm is a small trial and it should be read as one. But it agrees with the mechanistic prediction, it agrees with the mouse model that preceded it, and it is the reason the formulation on the shelf that says "40:1" is not marketing.

Gestational Diabetes: The Two Answers

Because myo-inositol improves insulin sensitivity and is safe, it was a natural candidate for preventing gestational diabetes. The 2023 Cochrane review pooled seven randomised trials and 1,319 women who started supplementation at 10–24 weeks' gestation. Myo-inositol may reduce gestational diabetes (RR 0.53, 95% CI 0.31–0.90; six trials, 1,140 women), hypertensive disorders of pregnancy (RR 0.34) and preterm birth (RR 0.35). The certainty was rated low to very low: the trials were small, six of the seven were Italian, and doses and timing varied.

A properly powered trial then arrived. The Dutch multicentre randomised, double-blind, placebo-controlled trial gave 464 pregnant women with PCOS myo-inositol 2 g twice daily (with folic acid) or placebo from 8–16 weeks to delivery. The primary composite of gestational diabetes, pre-eclampsia or preterm birth occurred in 25.0% on myo-inositol and 26.8% on placebo — RR 0.93, 95% CI 0.68–1.28, P = 0.67.

The two results are not strictly contradictory: the Cochrane pool was largely general obstetric populations, the JAMA trial was PCOS. But PCOS is exactly the group in whom the mechanism should work best, and the trial that tested it was larger and better controlled than any in the pool. The honest position now is that myo-inositol has not been shown to prevent pregnancy complications in women with PCOS, and that its benefit in other pregnant women remains a low-certainty signal awaiting an equivalent trial.

What Myo-Inositol Has Not Shown

Live birth. The 2018 Cochrane review of 13 trials in 1,472 subfertile women with PCOS was uncertain whether myo-inositol improves live birth (OR 2.42, 95% CI 0.75–7.83; two trials, 84 women) or clinical pregnancy (OR 1.27, P = 0.22). An apparent reduction in miscarriage disappeared when a single trial with an unusually high control-group miscarriage rate was removed. The endpoint a subfertile woman cares about is, so far, unproven.

Anything outside PCOS. There is no randomised trial showing that myo-inositol reduces progression to type 2 diabetes, cardiovascular events, or liver fat in people with metabolic syndrome who do not have PCOS. The mechanism is intact; the outcome trials have not been run. That is why it carries a B for PCOS endpoints and a C for general glycaemic use.

Weight. The BMI effect in the 2023 pool was a mean difference of −0.45 kg/m² — statistically significant and clinically marginal.

A clean trial base. The 2007 placebo-controlled trial that anchored the early enthusiasm was retracted in 2023 after the journal found it duplicated the same group's 2003 paper, which now carries its own expression of concern. The larger meta-analyses were built on later trials and their conclusions stand, but a field whose founding papers have been withdrawn should be held to a higher standard for its next claims.

Dosing

2 g twice daily — 4 g/day — as powder dissolved in water. That is the dose in nearly every PCOS trial and in the pregnancy trials. There is no evidence that more helps for metabolic endpoints, and the safety review found gastrointestinal side effects (nausea, flatulence, diarrhoea) only at 12 g/day, with severity not increasing with dose beyond that. Food does not matter.

For ovulation, use a 40:1 formulation — roughly 4 g myo-inositol with 100 mg D-chiro-inositol daily. For metabolic endpoints alone, plain myo-inositol is fine.

Give it time. Insulin and HOMA-IR respond within 12 weeks; the SHBG effect and the androgen effect need 24 weeks. A cycle diary is a better measuring instrument than memory.

How to Judge a Trial of It

Because myo-inositol is safe and cheap, the usual failure is not harm but a trial that was never set up to be judged. The endpoints move on different clocks, and a woman who checks one of them at the wrong time will conclude it does nothing.

A practical sequence: baseline fasting glucose and insulin plus a cycle diary in month one; the same draw and the diary at 12 weeks; androgens and SHBG at 24 weeks. If the insulin markers and the cycle diary have both moved by 12 weeks, the SHBG result at 24 is confirmation rather than the decision. If neither has moved at 12 weeks despite four grams a day taken consistently, myo-inositol is unlikely to be the answer, and the conversation should turn to metformin or to the berberine layer that sits below it in the protocol.

Two things will contaminate the read. Weight change of more than a few kilograms in either direction moves fasting insulin on its own, and should be noted alongside the bloods. And starting a combined oral contraceptive during the trial will raise SHBG by a pharmacological route that has nothing to do with inositol — the SHBG endpoint becomes uninterpretable from that point.

Who Should Take It

A woman with PCOS who wants an insulin sensitiser and either cannot tolerate metformin or prefers to try the non-pharmaceutical route first has the clearest case. She should expect improved fasting insulin and HOMA-IR, more regular cycles, and a slow drift downward in free androgen — not a guaranteed pregnancy.

A woman with PCOS trying to conceive should discuss it with her fertility clinic in the knowledge that cycle regularity is demonstrated and live birth is not.

A man with metabolic syndrome, or a woman without PCOS, is taking it on mechanism. It is cheap and safe enough that the trial costs nothing; it is also not the layer to build a plan around. Berberine has the outcome data in that population, and magnesium repletion has a pooled HOMA-IR effect of similar size for a fraction of the cost.

The stack that puts all of this in order — measurement first, magnesium, berberine, then inositol as the PCOS fork — is set out in The Insulin Sensitivity Stack.

Frequently Asked Questions

Is myo-inositol as effective as metformin for PCOS?

On the surrogate markers measured — BMI, fasting insulin, fasting glucose, HOMA index, LH/FSH — eight head-to-head trials found no significant difference, and the 26-trial meta-analysis called inositol non-inferior for cycle regularity. That is not the same as equivalence on pregnancy or long-term outcomes, which have not been compared.

Should I take D-chiro-inositol instead?

No. In the seven-ratio trial, D-chiro-inositol alone was the poorest at restoring ovulation, and activity fell as the ratio moved toward it. The 40:1 myo:D-chiro formulation was best. High-dose D-chiro-inositol is a different product with a worse ovarian result.

Does myo-inositol prevent gestational diabetes?

In women with PCOS, a 464-woman double-blind trial found no reduction in gestational diabetes, pre-eclampsia or preterm birth. In general obstetric populations, Cochrane's pool of seven small trials suggests it may (RR 0.53), at low certainty. It is not a proven prevention.

How long before it works?

Fasting insulin and HOMA-IR move within 12 weeks. SHBG rose only in trials of 24 weeks or more, and the androgen effect follows the same slower timeline. Judge cycles over at least three months, not one.

Are there side effects?

Very few. The safety review found mild gastrointestinal effects only at 12 g/day — three times the usual dose. At 4 g/day the trials report tolerability comparable to placebo, which is the main argument for it over metformin.

Can I combine it with berberine?

Yes, with monitoring. They act through different mechanisms and both have been trialled in PCOS. In someone not on diabetes medication, the combination is reasonable; in someone on insulin or a sulfonylurea, stacking glucose-lowering agents needs a prescriber.

Related Research

• Myo-Inositol
• The Insulin Sensitivity Protocol
• The Insulin Sensitivity Stack: Berberine + Inositol + ALA + Chromium
• Berberine vs. Metformin
• Female Hormonal Optimization