Myo-inositol is one of the most-googled supplements in women's health, largely on the back of its reputation for polycystic ovary syndrome. The research is genuinely interesting, but it is still developing, so it deserves an honest read rather than a hype reel.
What is inositol?
Inositol is a sugar-like compound your body makes and that also turns up in foods like beans, grains, nuts, and citrus fruit. It is not a vitamin in the strict sense, though it is sometimes loosely grouped with the B vitamins. Inside cells, its main job is acting as a messenger, including in the pathways that help insulin do its work.
There are several forms. The two that matter most for supplements are myo-inositol (by far the most abundant in the body) and D-chiro-inositol. The two have different roles in tissues, which is why the ratio between them comes up in research and on product labels. The body normally keeps myo-inositol and D-chiro-inositol in roughly a 40-to-1 balance in the bloodstream, and many PCOS studies deliberately copy that 40:1 ratio rather than using D-chiro-inositol on its own.
Why inositol is studied for insulin and hormones
Because inositol participates in insulin signaling, researchers have looked at whether supplementing it can help conditions linked to insulin resistance. When cells respond poorly to insulin, the body tends to make more of it to compensate, and persistently higher insulin can nudge the ovaries toward producing more androgens (hormones such as testosterone). That chain is part of why inositol became a candidate for PCOS, where insulin resistance is common. The theory is that improving how cells read the insulin signal could ease some of the downstream hormonal effects, which is exactly what trials have tried to measure.
The best-supported use: PCOS
The strongest evidence for myo-inositol benefits is in PCOS, and it comes mainly from systematic reviews and meta-analyses that pool many small randomized trials. Together they suggest myo-inositol may support insulin sensitivity and, in some people, more regular cycles and improved ovulation. That makes it one of the better-studied non-prescription options for the condition. For the bigger picture, see our guides to PCOS symptoms and PCOS and insulin resistance.
Two honest caveats. First, "may" is doing real work here: the underlying trials vary in size, quality, dose, and how long they ran, and many were small, so effects are not guaranteed for everyone and the certainty of the evidence is best described as moderate rather than strong. Second, inositol is not a replacement for the foundations of PCOS care, which include medical treatment where appropriate plus dietary and lifestyle changes such as those in our PCOS diet guide.
Other areas: earlier or mixed evidence
Beyond PCOS, the picture is thinner. Treat these as areas of interest, not settled benefits, and raise them with a clinician before acting.
- Metabolic markers. Some studies look at blood sugar and lipids, but the results are mixed and often come from small, short studies.
- Anxiety and mood. Older, limited research explored inositol for mood and anxiety; the evidence is early and inconsistent, and it is not a basis for treating a diagnosed mental-health condition or replacing prescribed treatment.
- Pregnancy-related blood sugar. Inositol has been studied for lowering the risk of gestational diabetes, mostly in small trials of higher-risk women, and reviewers have called the evidence promising but not yet strong enough to recommend routinely. Pregnancy is exactly when self-supplementing is risky, so this is a conversation to have with your obstetric clinician, not a do-it-yourself decision.
Inositol evidence at a glance
| Use | What the evidence suggests | Confidence |
|---|---|---|
| PCOS: insulin sensitivity | May help | Moderate |
| PCOS: menstrual regularity and ovulation | May help in some people | Moderate |
| Metabolic markers (blood sugar, lipids) | Mixed, often small effects | Low |
| Anxiety and mood | Early, inconsistent | Low |
| Pregnancy-related blood sugar | Promising but unsettled; clinician-led only | Low to moderate |
How myo-inositol is usually taken
Inositol is generally well tolerated. When side effects occur, they tend to be mild and digestive, such as nausea, gas, or loose stools, and are more likely at higher amounts.
In studies, it is most often given as myo-inositol, sometimes paired with a small amount of D-chiro-inositol in that fixed 40:1 ratio. We give the following only as amounts studied, not as a recommendation or a prescription: PCOS trials have frequently used roughly a few grams of myo-inositol per day, commonly split into two doses, often over a few months before reassessing. The right choice for you, including whether to use it at all, depends on your health, your other medications, and your goals, which is why this belongs in a clinician conversation rather than a label or a forum thread.
Honest caveats about supplements
Supplements are not regulated as tightly as prescription medicines, so quality, dose, and label accuracy can vary between products, and a few have been found to contain more or less than they claim. Inositol is not a cure and not a substitute for medical care or for the basics that move the needle most, like sleep, movement, and nutrition. We take the same stance across our coverage; our broader take on choosing supplements wisely is in best supplements for menopause. If your real concern is a wider sense of "hormone imbalance," the better first step is a proper clinical evaluation with specific tests, not a supplement bought on a hunch.
When professional care makes sense
Talk to a clinician before starting myo-inositol, especially if you are pregnant, trying to conceive, breastfeeding, or taking any medication, since interactions and individual factors matter. See a clinician promptly if you have irregular or absent periods, are struggling to conceive, or have symptoms of a metabolic or hormonal condition, so you can get a real diagnosis rather than guessing. Supplements like inositol are meant to complement, not replace, medical care, and the most useful version of "is this right for me?" is one answered with a professional who knows your history.
What the pooled evidence actually found
Inositol has an unusually large number of small, short, single-centre trials behind it — which is exactly the situation where cherry-picking one study tells you nothing. A 40-woman trial from a single clinic can show a dramatic result by chance, and supplement marketing tends to quote precisely those. Pooled analyses combine every eligible trial, weight them by size, and report how wide the uncertainty really is. For inositol the honest summary is that pooling narrows the claims considerably — though the pooled reviews do not fully agree with each other.
Three syntheses matter here. The largest and most methodologically careful is the review by Fitz and colleagues in the Journal of Clinical Endocrinology & Metabolism (2024), commissioned to inform the 2023 International Evidence-based PCOS Guideline; it screened 30 trials covering 2,230 women and pooled 19 of them[s1]. Greff and colleagues in Reproductive Biology and Endocrinology (2023) pooled 26 RCTs in 1,691 women[s2]. And an umbrella review by Duan and colleagues in Frontiers in Endocrinology (2026) sits one level higher again, pooling 13 separate meta-analyses and grading 85 individual evidence items[s3].
They disagree in an informative way. Greff found inositol improved menstrual cycle normalisation against placebo (RR 1.79, 95% CI 1.13–2.85) and lowered total testosterone, but its insulin numbers against placebo did not reach significance — fasting insulin MD −2.78 µU/mL and HOMA-IR MD −0.23, both straddling zero[s2]. Fitz, comparing myo-inositol plus folic acid against folic acid alone across 8 trials, found the opposite emphasis: clear improvement in fasting insulin (MD −4.17 µU/mL, −5.14 to −3.20) and HOMA-IR (MD −1.24, −1.50 to −0.99), but no significant change in BMI or clinical pregnancy (OR 1.24)[s1]. One reason the reviews differ is that they included different trials: the Fitz team excluded 13 studies flagged at moderate risk under the RIGID research-integrity framework, which other recent meta-analyses retained[s1].
On the metformin question — the one most buyers are actually asking — the picture is mostly, but not entirely, one of equivalence. Across 10 trials (roughly 700 women), Fitz found myo-inositol and metformin statistically indistinguishable on ovulation, menstruation, pregnancy, fasting insulin, HOMA-IR and BMI. Metformin did better on hirsutism and waist-hip ratio; myo-inositol caused far fewer gastrointestinal side effects (OR 0.09)[s1]. The umbrella review agrees on ovulation versus metformin (RR 1.33, 0.99–1.77, just failing significance) and on HOMA-IR (MD −0.12, −0.31 to 0.08) — but it parts company on fertility, reporting clinical pregnancy favouring inositol over metformin (RR 1.43, 1.14–1.79) at moderate certainty[s3]. That contradiction is unresolved, and worth knowing before anyone tells you the question is settled either way.
| Buyer's question | What the pooled data show | Source and size | How solid is it |
|---|---|---|---|
| Does it restore ovulation? | Overall RR 1.36 (1.04–1.77). The signal is against placebo/folic acid (RR 2.75, 1.71–4.41), not against metformin (RR 1.33, 0.99–1.77). | Duan 2026 umbrella review, 13 meta-analyses[s3] | Low certainty for the overall figure; moderate for the two sub-comparisons |
| Is the D-chiro-inositol ovulation claim real? | OR 11.50 (3.40–38.91) versus placebo — a huge-looking number from only 2 trials totalling 64 women. | Fitz 2024, JCEM[s1] | Very weak — the confidence interval spans a more-than-tenfold range |
| Does it improve insulin resistance? | Yes vs folic acid alone: fasting insulin MD −4.17 µU/mL (−5.14 to −3.20), HOMA-IR MD −1.24 (−1.50 to −0.99). But Greff found no significant placebo effect on either marker. | Fitz 2024, 8 trials[s1]; Greff 2023, 26 RCTs n=1,691[s2] | Contested — the two reviews reach opposite conclusions |
| Does it regulate periods? | Vs placebo, RR 1.79 (1.13–2.85). Vs metformin, no significant difference (RR 1.42, 0.80–2.53; OR 1.85, 0.68–5.01). | Greff 2023[s2]; Fitz 2024[s1] | Modest — the metformin comparison rests on very wide intervals |
| Is it as good as metformin? | Fitz: no pooled difference on ovulation (OR 1.28), pregnancy (OR 1.18), fasting insulin, HOMA-IR or BMI. Duan disagrees on one point — clinical pregnancy favoured inositol (RR 1.43, 1.14–1.79). | Fitz 2024, 10 trials[s1]; Duan 2026[s3] | Broadly equivalent, with one unresolved conflict on pregnancy |
| Where does metformin win? | Hirsutism (Ferriman-Gallwey MD 2.42) and waist-hip ratio (MD 0.04). | Fitz 2024[s1] | Reported as favouring metformin |
| Where does inositol win? | Tolerability — gastrointestinal adverse events OR 0.09 versus metformin. Also modestly higher SHBG (MD 3.85). | Fitz 2024[s1] | Consistent in direction, but 23 of 29 trials reported no adverse-event data at all |
The honest bottom line for someone about to spend money. The most reproducible finding across these syntheses is a negative one: inositol is not clearly better than metformin on ovulation, menstrual regularity or insulin markers — and it is markedly gentler on the gut. That makes tolerability, rather than superior efficacy, the strongest argument for it, and the guideline review's own framing is that inositol may be worth considering where metformin cannot be tolerated. One caveat cuts the other way: the 2026 umbrella review did find clinical pregnancy favouring inositol over metformin (RR 1.43, 1.14–1.79) at moderate certainty, which the guideline review did not — so the fertility question specifically is genuinely open rather than closed[s3]. Where the evidence is weakest: the eye-catching D-chiro-inositol ovulation figure rests on 64 women; the two large reviews contradict each other on whether inositol moves insulin markers against placebo at all; 23 of 29 trials in the guideline review never reported adverse events, so the safety picture is thinner than it looks; most trials ran only 3–24 weeks with no follow-up beyond a year; and 15 of 30 trials came from a single country[s1][s3]. Of 85 graded evidence items in the umbrella review, none reached high quality. The guideline authors' own conclusion is the fairest summary available: the evidence "is limited and inconclusive," and the decision belongs in a conversation with your clinician rather than to a supplement label. This page is reference information, not a diagnosis or a treatment plan.
