Does AMH Decrease With PCOS Treatment? What the Trials Actually Show

Does Weight Loss Lower AMH in PCOS?

Two of the better-designed trials on this exact question found no significant change in AMH, despite real weight loss in both. A 2009 trial followed 52 overweight and obese women with PCOS and reproductive impairment through a 20-week weight-loss programme; the 26 women who went on to improve their menstrual cycles and ovulation lost more weight on average (11.7 kg versus 6.4 kg, p = 0.001) and started with lower AMH than those who didn’t respond (23.5 versus 32.5 pmol/L, p = 0.03) — but AMH itself did not change with weight loss in either group. A larger 2022 cohort followed 246 women with severe obesity, with and without PCOS, through a one-year very-low-energy diet programme and reached the same conclusion: significant weight loss, no significant change in circulating AMH, in either group.

Read together, these two trials are the honest answer to a common assumption: losing weight does not appear to reliably lower AMH in PCOS, even over a full year. That does not mean weight loss did nothing — in the 2009 trial, the responders’ menstrual cycles and ovulation genuinely improved. It means AMH was the wrong number to watch for whether that improvement was happening.

Does Metformin Lower AMH in PCOS?

Metformin is the one intervention here with replicated evidence of actually lowering AMH. A 2023 meta-analysis pooling 14 studies and 257 women with PCOS found AMH fell significantly after metformin treatment (SMD −0.70, 95% CI −1.13 to −0.28, p = 0.001), with the effect concentrated in women younger than 28 (SMD −1.24) and in those who started with AMH above 4.7 ng/mL (SMD −0.66) — in women starting below that threshold, the effect did not reach significance. A separate 2024 meta-analysis went further and pooled only the randomized trials that compared metformin against a placebo or control group, rather than measuring people before and after with no comparison arm, and still found a significant drop (WMD −3.47 ng/mL, 95% CI −7.14 to −0.19, p = 0.047) — a real effect, though that wide confidence interval means the true size of it is genuinely uncertain. The same analysis’s pooled before-and-after data landed on a similar, tighter estimate (WMD −3.06 ng/mL, 95% CI −4.03 to −2.10, p < 0.001).

None of this is a reason to start or continue metformin specifically to move an AMH number — metformin is a prescription drug with its own dosing, monitoring and side-effect profile, covered in full in what metformin actually does for PCOS, and a lower AMH has not been shown to be a goal worth prescribing for on its own.

Does the Birth Control Pill Lower AMH in PCOS?

The evidence on the combined pill actually conflicts, and this is one of the few places in this review where two directly relevant trials point different ways. A 2016 study following 58 women with PCOS — 11 on a combined pill, 47 on the anti-androgen cyproterone acetate — found AMH fell by 22% ± 27% on the pill and 28% ± 20% on cyproterone acetate after three months, with no significant difference between the two drugs. The authors’ own reading is notable: because an anti-androgen (cyproterone) suppressed AMH no more than the pill did, the mechanism is probably shared gonadotropin (LH) suppression, not a direct anti-androgen effect on the ovary. Against that, an earlier 2007 study of 30 women with PCOS and 15 with normal cycles found that after six full cycles on a combined pill, ovarian volume and follicle count had dropped significantly on ultrasound in both groups — but AMH itself had not moved at all.

A three-month trial found a drop; a six-cycle trial in a different, larger PCOS sample found none. Both trials are small — 11 and 30 women on the pill, respectively — and neither is large enough to settle it. The honest summary is that the pill’s effect on AMH in PCOS is unresolved, which is exactly the kind of uncertainty most pages on this topic skip past rather than state.

Combined hormonal contraceptives are not the right choice for everyone regardless of their effect on AMH — they carry standard cautions around smoking, migraine with aura, and a personal or family history of blood clots, and that conversation belongs with whoever is prescribing, not with an AMH result.

Does Inositol Lower AMH in PCOS?

In the one trial that has actually tested this, myo-inositol outperformed the comparator. A 2016 randomized trial matched 106 women with PCOS into a myo-inositol-plus-folic-acid arm (52 women) and a combined-pill arm (54 women) and measured AMH before and after three months. AMH fell significantly in both groups — from 11.51 to 9.07 ng/mL on myo-inositol (a drop of 2.44 ng/mL, p = 0.002) and from 9.39 to 8.51 ng/mL on the pill (a drop of 0.88 ng/mL, p < 0.001) — and the difference between the two drops was itself significant (p = 0.048), meaning myo-inositol produced roughly three times the reduction. Ovarian volume and antral follicle counts fell in both arms too.

That is one trial, and it compared two active treatments against each other rather than against a placebo, so it cannot say how much of either drop would have happened anyway. It is also the only inositol trial in this review that measured AMH as an outcome at all — treat it as a first data point, not a settled answer.

Why AMH Falls With Treatment — and Why That Is Not the Same as a Declining Reserve

AMH is made only by granulosa cells lining small, actively growing antral follicles, so any drop in the blood level always means fewer of those specific follicles are active at that moment. But that can happen for two entirely different reasons, and conflating them is the single most common way this topic gets misunderstood. In PCOS, AMH runs roughly 2 to 3 times higher than average because that many more small antral follicles sit arrested early in development rather than being recruited and cleared normally. When a treatment reduces the androgen, insulin or LH signalling that keeps those follicles stuck — which is the plausible mechanism behind the metformin and hormonal-suppression findings above — fewer of them are being actively recruited at once, so AMH production drops. That is a change in follicle behaviour, not a change in how many eggs are left.

Separately, and in everyone, AMH also declines with age as the finite pool of primordial follicles a person is born with genuinely depletes, becoming undetectable roughly five years before menopause. That is a real reduction in the size of the remaining reserve. Both processes produce the identical result on a lab report — a lower number than last time — and nothing about the number itself tells you which one happened.

The 2023 international guideline accepts an elevated AMH as a substitute for an ultrasound when diagnosing polycystic ovarian morphology in adults — a diagnostic use, and nothing in it recommends tracking AMH as a marker of whether a PCOS treatment is working. Diagnosis and treatment-monitoring are different jobs for the same number, and the trials above were testing the second one, largely without prior validation that it is a job AMH can do.

You may see this condition written as polyendocrine metabolic ovarian syndrome (PMOS), after a 2026 global consensus of more than 50 organisations renamed it. Nothing about AMH or any trial above changed with the name — this article uses PCOS because that is still what most readers search.

What a Falling — or Unchanged — AMH Does Not Mean

None of the trials above answer whether a lower AMH means better fertility, and that question already has its own separate evidence: AMH is not validated as a predictor of natural conception odds in anyone, PCOS or not, so a number moving in either direction should not be read as a fertility forecast changing. It also says nothing about egg quality — PCOS affects egg quantity far more than egg quality, and AMH tracks the follicle-recruitment story covered above, not the health of the eggs inside those follicles. For the mechanics of what AMH and antral follicle count are actually counting, the antral follicle count page covers the ultrasound side of the same biology.

An unchanged AMH after a genuine intervention is not evidence the intervention failed, either — the 2009 weight-loss trial above is the clearest example: reproductive function improved in the women who lost more weight, with no corresponding AMH change at all. And none of these trials ran past about a year, so nobody has evidence for whether any of these changes persist, reverse once treatment stops, or matter for anything beyond the number itself. Nothing here has been tested in adolescents either — AMH is not used diagnostically under 20 at all, and none of these treatment trials enrolled that age group.

If a different fertility question brought you here, the fertility section covers ovulation, treatment options and pregnancy risk in PCOS from the ground up.