Does myo-inositol work as well as metformin if you have lean PCOS?

The numbers that matter

Testosterone change shown for the metformin group (3.1 to 2.0 nmol/L). The inositol group moved 2.9 to 2.0, landing at the same endpoint.

Can we actually trust this?

Yes, with real caveats. The core design is sound: participants were randomized, treated at a university hospital gynecological endocrinology department, diagnosed by Rotterdam criteria, screened for the conditions that mimic PCOS, and confirmed insulin resistant using a three-hour glucose tolerance test rather than a single fasting number, which is more rigorous than most studies bother with. The findings also line up with a large body of existing research on both drugs, which is the strongest reason to take them seriously. The caveats are about precision rather than direction. The two arms weren't treated for equal lengths of time, with metformin assessed at six months and inositol at eight, so any head-to-head comparison is uneven. There was no placebo group, which the authors explain on ethical grounds. And of the 80 women enrolled, 60 were analyzed, with eight inositol participants excluded because they became pregnant and six metformin participants excluded because side effects made them stop, which removes successes from one arm and failures from the other. The paper also carries some internal inconsistencies, including several identical p-values reported for two separate groups and a participant count that differs between the abstract and the methods. Trust the shape of this result, which is that both treatments help; hold the specific numbers loosely.

Does this apply to you?

This study is most relevant if:

• You have PCOS diagnosed by Rotterdam criteria
• Your BMI is in the normal range, roughly 19 to 25
• You have confirmed insulin resistance
• You're over 18, with an average participant age of 27
• You're not taking other regular medications or supplements

Why this matters

If you have PCOS and a normal BMI, you have almost certainly been handed advice written for someone else. The standard opening move for insulin resistance is weight loss, and when there's no excess weight to lose, that recommendation doesn't just fail to help, it quietly implies your insulin resistance isn't real. It is: insulin resistance affects an estimated three in four lean women with PCOS, and this trial's participants averaged a HOMA-IR of 3.5 to 4.0 against a healthy threshold of 2.5, at an average BMI of about 21. So a study that recruits only normal-weight women and confirms insulin resistance in all of them is filling a gap that matters, and what it found is that both treatments worked on the things that actually bother people: androgens came down, insulin came down, and over 90% of women in both groups had regular cycles by the end, up from under 7% at the start. The catch worth carrying with you is that this trial can't tell you which option is better, because the two arms weren't run for the same length of time and the dropouts were removed in opposite directions. What it does show is that in lean PCOS, insulin resistance is treatable without the weight-loss conversation, which for a lot of readers is the first time that sentence has been said out loud.

If you want to try this

1. Both are prescription-adjacent decisions, not shopping decisions.

   Metformin requires a prescription. Myo-inositol is sold as a supplement, which makes it feel casual, but it's still an active compound being used here to treat a diagnosed metabolic condition. This is a conversation with your doctor either way.

2. Know what the doses actually were.

   Metformin started at 500 mg daily with meals and increased by 500 mg every three weeks to 1500 mg daily, reached at week six. The inositol group took 2000 mg of myo-inositol with 200 mcg of folic acid, twice a day, from the start.

3. Expect months, not weeks.

   Effects were measured at six months for metformin and eight for inositol. Nothing here was assessed at four weeks, so a supplement that hasn't obviously changed anything by then hasn't failed.

4. Side effects are the real difference between these two.

   Six of forty metformin participants stopped entirely because of gastrointestinal effects, against one in the inositol group. That gap is consistent with the broader evidence, and it's a legitimate factor in the decision.

5. If pregnancy is on the table, mention it.

   Eight women in the inositol group became pregnant during the study and were removed from the analysis. That wasn't what the trial set out to measure and it can't be read as proof of anything, but it's relevant context for a conversation with your clinician, in both directions.

6. Ask to have insulin resistance measured properly.

   These women were assessed with a three-hour glucose tolerance test with insulin levels, not a single fasting glucose. If you're lean and suspect insulin resistance, the test you get matters.

FAQ

Can you have insulin resistance with PCOS at a normal weight?

Yes, and it's common. Insulin resistance is estimated to affect around 75% of lean women with PCOS, compared to about 95% of those with obesity, though it's usually less severe in lean women. Every participant in this trial had a normal BMI and confirmed insulin resistance, with average scores well above the healthy threshold.

Is myo-inositol as effective as metformin for PCOS?

In this trial the two produced similar results on insulin, androgens, and cycle regularity, with no significant difference between them on any measure. But the two arms were treated for different lengths of time, so the comparison isn't clean. Larger reviews have generally also found the two comparable, with metformin somewhat stronger on insulin resistance and inositol somewhat better tolerated.

What does myo-inositol do for PCOS?

It acts as a messenger in the insulin signalling pathway and in follicle-stimulating hormone pathways in the ovary, which is why it's studied for both metabolic and ovulatory effects in PCOS. In this trial it lowered fasting insulin, testosterone, androstenedione, DHEAS, and the free androgen index, and raised SHBG.

Why does metformin cause stomach problems?

Gastrointestinal effects including nausea, diarrhoea, bloating, and abdominal pain are the most commonly reported side effects of metformin and are well documented. In this study six participants stopped treatment because of them. Doses are often started low and increased gradually, as they were here, partly to reduce this.

Study at a glance

Study

The comparative effects of myo-inositol and metformin therapy on the clinical and biochemical parameters of women of normal weight suffering from polycystic ovary syndrome

Design

Prospective randomized controlled trial, two active arms, no placebo group. Metformin assessed at 6 months, myo-inositol at 8 months

Participants

80 women enrolled and randomized (40 per arm); 60 analyzed (30 per arm). All with PCOS by Rotterdam criteria, confirmed insulin resistance, and BMI 19 to 25. Mean age 27, mean BMI about 21.7. Phenotype A most common in both groups

Intervention

Metformin titrated from 500 mg to 1500 mg daily by week 6, versus myo-inositol 2000 mg with folic acid 200 mcg, twice daily

Key result

Both arms significantly reduced insulin area under the curve, HOMA-IR, fasting insulin, testosterone, androstenedione, DHEAS, and free androgen index, and raised SHBG. Regular cycles rose from 3.3% to 93.3% with metformin and 6.7% to 90% with inositol. No significant difference between treatments on any measure

The honest limit

Unequal treatment duration between arms, no placebo, and 20 of 80 participants excluded from analysis in opposite directions (8 inositol pregnancies, 6 metformin side-effect withdrawals). Several identical p-values reported across the two groups, a misprinted formula for the free androgen index, and a participant count that differs between the abstract and methods

Tolerability

6 metformin participants stopped due to gastrointestinal effects, versus 1 in the inositol group

Setting

Clinic for Gynecology and Obstetrics, University Clinical Center of Serbia, Belgrade, June 2017 to June 2018

Journal

Biomedicines, 2024

DOI

10.3390/biomedicines12020349