Inositol: what myo-inositol and D-chiro-inositol actually do

Inositol is one of the few supplements on the shelf that a mainstream clinical guideline actually discusses by name — and it is also one of the most confidently oversold. The honest version is narrower than the marketing: a sugar-like compound found in ordinary food, with a real but low-certainty case in polycystic ovary syndrome, an encouraging but geographically narrow signal for preventing gestational diabetes, and a psychiatric literature that is three decades old, used enormous doses, and did not hold up when pooled. It is also two different molecules sold as if they were one, at a ratio almost nobody selling it can source properly.

What inositol actually is

Inositol is a six-carbon cyclic polyol — a sugar alcohol, structurally a close relative of glucose. Nine stereoisomers exist; the one that dominates both food and biology is myo-inositol, which turns up in cereals, corn, legumes, fruit and meat, and which the body also synthesises from glucose. Its day job is signalling: inositol is the backbone of the phosphatidylinositol system, and inositol-containing glycans act as second messengers downstream of insulin.^[1] That is the whole mechanistic pitch in one sentence — if insulin signalling depends on inositol species, then topping up inositol might improve insulin sensitivity. Everything below is a test of how far that idea travels.

Myo- versus D-chiro-inositol: different molecules, different jobs

This is the distinction the category is built on and the one most product pages blur. An insulin-dependent epimerase converts myo-inositol into D-chiro-inositol by inverting a single hydroxyl group, and the enzyme’s activity — and therefore the local MI:DCI ratio — differs from tissue to tissue, presumably because different tissues need different amounts of each.^[4] DCI sits closer to the glucose-disposal end of insulin signalling: in women with PCOS, the amount of DCI-containing inositolphosphoglycan mediator released during a glucose load correlated with measured insulin sensitivity.^[5] Myo-inositol, by contrast, is the species the ovary appears to need most.

Hence the “D-chiro-inositol paradox”: the hypothesis that in PCOS ovaries, epimerase activity is too high, so myo-inositol is over-converted and the follicle ends up locally myo-inositol-deficient — which the authors proposed as a cause of poor oocyte quality.^[4]If that is right, loading DCI is not simply a stronger version of loading MI. It could push the ratio further in the wrong direction inside the one tissue you were trying to help. This is the honest reason to care about which isomer is in the bottle.

Where the 40:1 ratio actually comes from

Nearly every combination product uses 40 parts myo-inositol to 1 part D-chiro-inositol, and the stated rationale is specific: 40:1 is meant to reproduce the reported physiological MI:DCI ratio in human plasma. The papers that set out that rationale argue that DCI alone, especially at high dose, worsens oocyte quality, and that the plasma-mimicking blend is therefore the better bet.^[6] Two things are worth saying plainly about this.

First, the “too much DCI is bad” half of the argument has real supporting data — in mice. Feeding young adult female mice DCI alone at doses corresponding to roughly 1,200 mg/day in humans altered ovarian histology, raised serum testosterone and reduced ovarian aromatase, essentially producing an androgenic PCOS-like picture; higher doses (human-equivalent 2,400–4,800 mg/day) produced lesions resembling aged ovaries.^[7] That is a legitimate reason to be wary of high-dose DCI monotherapy sold as a PCOS product.

Second, and less comfortably: the head-to-head human evidence that 40:1 beats myo-inositol alone is essentially absent. The ratio is a mechanistic argument, published largely by a cluster of authors with commercial and society affiliations in the inositol field, not the output of randomised comparisons of 40:1 against MI alone at matched dose. The 2023 international evidence-based PCOS guideline is explicit on this point: no specific inositol type, dose or combination can be recommended, because the quality of evidence is insufficient to distinguish them.^[8] A ratio can be biologically sensible and still be commercially convenient and clinically unproven at the same time.

PCOS: the strongest indication, and still low certainty

This is where inositol has the most data by a wide margin. A 2023 systematic review pooled 26 randomised trials covering 1,691 women and found that inositol raised the chance of having a regular menstrual cycle by a factor of 1.79 (95% CI 1.13 to 2.85) versus placebo, and showed non-inferiority to metformin on that outcome. It also reported modest advantages over placebo in BMI (mean difference −0.45 kg/m²), free testosterone (−0.41), total testosterone (−20.39), fasting glucose (−3.14) and insulin AUC, plus a rise in sex-hormone-binding globulin.^[9] Directionally, that is a coherent picture: slightly better cycles, slightly lower androgens, slightly better glucose handling.

Now the counterweight. The systematic review commissioned to inform the 2023 guideline update assessed 30 trials in 2,230 participants and reached a deliberately deflating conclusion: the evidence supporting inositol in PCOS is “limited and inconclusive,” with benefits for some metabolic measures, a possible DCI effect on ovulation, and no effect on much else. It also found that metformin may beat inositol for hirsutism and waist-to-hip ratio, while myo-inositol likely causes fewer gastrointestinal adverse events than metformin — those events being mild and self-limiting in both arms.^[10] The guideline itself lands on a recommendation that reads almost apologetically: inositol in any form could be considered based on individual preference, at very low certainty of evidence, with metformin preferred where hirsutism and central adiposity are the targets.^[8]

The reason for that gap between “statistically significant pooled effects” and “inconclusive” is trial quality, not trial direction. The underlying studies are small, typically run for three to six months, frequently open-label, and a substantial share come from research groups with commercial links to inositol products. Positive pooled estimates built on that base are worth taking seriously and worth not overstating.

Gestational diabetes: the most encouraging non-PCOS signal

Here the effect sizes are the largest anywhere in the inositol literature. A 2023 Cochrane review pooled 7 randomised trials in 1,319 pregnant women and found that antenatal myo-inositol may reduce the incidence of gestational diabetes (RR 0.53, 95% CI 0.31 to 0.90) and of hypertensive disorders of pregnancy (RR 0.34, 95% CI 0.19 to 0.61). The certainty was rated low to very low, only one trial reported large-for-gestational-age infants (RR 1.40, 95% CI 0.65 to 3.02 — consistent with both benefit and harm), and no included trial reported perinatal mortality or a morbidity composite.^[13]Six of the seven trials were run in Italy and one in Ireland, which is a real external-validity limit, not a footnote.

The single most-cited trial gives a feel for the raw numbers: in 220 obese pregnant Italian women randomised from the first trimester to 2 g myo-inositol plus 200 micrograms folic acid twice daily or folic acid alone, gestational diabetes occurred in 14.0% versus 33.6% (odds ratio 0.34, 95% CI 0.17 to 0.68), with a greater fall in HOMA-IR in the inositol arm.^[14] That is a large effect from a cheap intervention — and it is one open-label trial at two centres in one country, which is exactly why Cochrane graded the pooled result the way it did.

Pregnancy is also the context where self-prescribing is least appropriate. Dose, timing and whether inositol is a reasonable addition at all belong to the clinician managing the pregnancy.

Anxiety, OCD and panic: very high doses, very small studies, very long ago

Inositol’s psychiatric reputation rests on a short run of crossover trials from the 1990s out of a single Israeli research group. In panic disorder, 21 patients completed a double-blind, placebo-controlled, four-week crossover trial of 12 g/day, with significantly greater declines in panic frequency, panic severity and agoraphobia on inositol than on placebo.^[15] In obsessive-compulsive disorder, 13 patients completed a controlled crossover of 18 g/day for six weeks each and scored significantly lower on the Yale-Brown scale on inositol.^[16] A later crossover in 20 patientscompared inositol up to 18 g/day against fluvoxamine up to 150 mg/day and found similar improvement on anxiety, agoraphobia and global impression scores, with less nausea and tiredness on inositol.^[17]

Note the doses. 12 to 18 g/day is three to nine times what PCOS protocols use — this is not the same intervention as a 2 g twice-daily myo-inositol capsule, and it is not what “inositol for anxiety” products deliver.

And the signal did not survive pooling. A 2014 meta-analysis assembled 7 double-blind randomised trials in depression (n = 242) and 4 in anxiety disorders (n = 70) and found no statistically significant effect of inositol on depressive, anxiety or obsessive-compulsive symptoms, nor on discontinuation. What remained were non-significant trends: marginally more responders in depression (p = 0.06), a trend toward benefit in premenstrual dysphoric disorder (p = 0.07), and marginally more gastrointestinal upset than placebo (p = 0.06). The authors flagged the tiny number of included studies as the central limitation.^[18] The fair grade for the psychiatric use is therefore an interesting 30-year-old signal in a few dozen patients that has never been replicated at scale — a category away from the PCOS evidence, not a peer of it.

Dosing, the GI ceiling, and what you are actually buying

The doses used in trials cluster tightly by indication. PCOS studies overwhelmingly use 2 g myo-inositol twice daily (4 g/day), often as a 40:1 blend that adds roughly 50 mg of D-chiro-inositol, and usually with folic acid. Gestational-diabetes prevention trials used2 g myo-inositol plus 200 micrograms folic acid twice daily, started in the first trimester.^[14] The psychiatric trials used 12–18 g/day.^[15][16]

That 4 g/day figure is also the number most product labels quietly fail, since a 500 mg capsule means eight of them a day. How to check the isomer split, the real dose and the cost per gram before buying is set out in our criteria for judging an inositol label.

Tolerance is what caps the dose. Inositol is unusually benign at PCOS-range doses — the guideline evidence review found fewer gastrointestinal adverse events on myo-inositol than on metformin^[10] — but the ceiling is real: in the pooled psychiatric trials, where doses were far higher, gastrointestinal upset was marginally more common on inositol than placebo.^[18] Practically, nausea, gas, bloating and loose stools are the dose-limiting symptoms, and they are why nobody takes 18 g/day casually. Splitting the dose and taking it with food is the usual mitigation.

The last thing to be clear about: inositol is a food-derived compound sold as a dietary supplement. It is not FDA-approved for PCOS, for preventing gestational diabetes, for anxiety or for anything else, and the 2023 PCOS guideline explicitly notes that the regulatory status, quality control and actual dosing of inositol supplements may differ from pharmaceutical products, varying product to product.^[8] What is on the label is a claim, not a guarantee.

It is not a weight-loss drug

Inositol shows up in weight-loss marketing, usually adjacent to PCOS. The measured effect does not support it: in the most favourable pooled analysis, the BMI difference versus placebo was −0.45 kg/m²^[9] — on the order of a couple of pounds for an average-height adult, over months, in trials designed around cycles and androgens rather than weight. That is not in the same category as an incretin drug, and inositol is in no sense a substitute for one: for what the medications actually do in this population see semaglutide for PCOS. If the pattern feels familiar, it is the same shape as the “nature’s Ozempic” problem we take apart in berberine: a supplement with genuine, modest metabolic data getting drafted into a weight-loss story it cannot carry.

The honest bottom line

Inositol earns a genuine, narrow place. For PCOS it plausibly improves cycle regularity and nudges androgens and glucose measures in the right direction, is better tolerated than metformin, and is cheap — while the international guideline rates that whole body of evidence very low certainty, prefers metformin for hirsutism and central adiposity, and declines to endorse any particular type, dose or ratio.^[8][10] Live birth is unproven.^[11] The gestational-diabetes prevention data are the most striking numbers in the field and also the most geographically narrow.^[13] The psychiatric case rests on 1990s crossovers at 12–18 g/day that did not survive meta-analysis.^[18] And the 40:1 ratio is a reasonable mechanistic bet whose superiority to plain myo-inositol has not been demonstrated in humans. If you are weighing it against a pharmaceutical insulin sensitiser, the comparison worth reading is berberine versus metformin, which walks through the same supplement-versus-drug trade-off with a different compound.

This article is research information, not medical advice. Inositol is sold as a dietary supplement and is not FDA-approved to treat or prevent polycystic ovary syndrome, gestational diabetes, infertility, anxiety, or any other condition. PCOS, subfertility and pregnancy care require a licensed clinician; do not start, stop or dose inositol in pregnancy, alongside metformin or other prescribed medication, or in place of prescribed treatment without discussing it with your own clinician.