Vitamin B12 for PCOS: The Metformin Connection and the Right Test

How Much Does Metformin Actually Lower B12, and How Fast?

Metformin lowers vitamin B12 in a dose- and duration-dependent way, not as a one-time drop after starting the drug. The clearest evidence comes from a Dutch randomised, placebo-controlled trial of 390 people with type 2 diabetes already on insulin, who took 850 mg of metformin three times daily, or placebo, for 4.3 years. By the end of the trial, mean B12 concentration had fallen 19% relative to placebo, and the absolute risk of B12 deficiency (below 150 pmol/L) was 7.2 percentage points higher on metformin, with a number needed to harm of about 14 people treated for 4.3 years to produce one additional case. A broader threshold — low or deficient B12 combined — was 11.2 points higher. That trial is in type 2 diabetes, not PCOS, which matters for reading the number correctly: it describes what metformin does over years in a population taking a similar dose to many PCOS prescriptions, not a PCOS-specific result.

The longest-running data comes from the Diabetes Prevention Program Outcomes Study, which followed people with prediabetes (elevated fasting glucose, impaired glucose tolerance, and overweight or obesity) assigned to metformin 850 mg twice daily or placebo for over a decade. Combined low and borderline-low B12 was 19.1% on metformin versus 9.5% on placebo at five years, and 20.3% versus 15.6% at thirteen years, and each additional year of metformin use raised the odds of biochemical B12 deficiency by 13% (odds ratio 1.13, 95% CI 1.06–1.20). Anaemia was also more common on metformin, and neuropathy was more common specifically among those with low B12 — not among metformin users generally, which points at B12 status itself as the mechanism rather than the drug acting on nerves directly.

The one study that looked at this in PCOS specifically found the same direction of effect. A cross-sectional comparison of 50 newly diagnosed, drug-naive women with PCOS against 52 women with PCOS who had taken metformin for at least six months found median serum B12 of 272.0 pg/mL in the metformin group versus 385.5 pg/mL in the drug-naive group (p < 0.001), with B12 deficiency in 15.4% versus 6% and borderline-low levels in 42.3% versus 18% of each group. It is a single, modestly sized study from one hospital, and it has not been replicated — but the gap it found lines up with the much larger diabetes and prediabetes trials above, in the same direction and a similar rough magnitude.

The 2023 international PCOS guideline does not include B12 monitoring in its metformin recommendations at all — that guidance comes from the diabetes and endocrinology literature on metformin itself, cited above, not from a PCOS-specific source. A 2021 review proposing screening criteria for metformin-treated patients put it plainly: no definite guidelines currently exist for when to check B12 in someone on long-term metformin, and the deficiency remains frequently unrecognised as a result. If you are on metformin long-term and nobody has mentioned checking your B12, that is a real gap worth raising yourself, not a sign nothing is wrong.

Why Can a Normal Serum B12 Result Still Miss a Deficiency?

A standard serum B12 test measures total circulating B12, which includes B12 bound to a carrier protein that cells cannot actually use — so a result inside the “normal” range does not guarantee your cells have enough B12 available at the level that matters. Two other blood markers move before serum B12 clearly falls, or move even when serum B12 still looks acceptable: methylmalonic acid (MMA) and homocysteine. Both rise when a B12-dependent enzyme reaction cannot proceed normally, which makes them functional markers of deficiency rather than a measure of how much B12 is simply present in the blood. A 2017 review in Nature Reviews Disease Primers describes exactly this diagnostic gap: decreased total B12 alongside elevated homocysteine and methylmalonic acid are the recognised biomarker pattern, and the cut-offs for classifying subclinical deficiency remain genuinely debated — subclinical B12 deficiency affects somewhere between 2.5% and 26% of the general population depending on which definition is used, a range wide enough on its own to show that “normal serum B12” is not a settled line.

The de Jager trial above shows this pattern inside its own data: participants who met the threshold for confirmed B12 deficiency had a mean homocysteine of 23.7 micromol/L, compared with 18.1 micromol/L in those with only low (not deficient) B12, and 14.9 micromol/L in those with a normal result — a clean stepwise relationship that tracked severity more precisely than the serum B12 category alone. A 2024 systematic review and meta-analysis of functional B12 status specifically in vegan adults found the same disconnect from a different angle: vegans had lower serum B12 than omnivores, but also significantly higher homocysteine (a functional marker of deficiency) and a borderline-significant rise in methylmalonic acid — a functional signal beyond what the serum B12 gap alone would suggest. Read together, both findings point the same direction: serum B12 alone can sit in a range that reads as reassuring while the functional markers say otherwise.

Why Do B12 Symptoms Get Blamed on PCOS or “Just Metformin”?

B12 deficiency produces fatigue, pins-and-needles or numbness in the hands and feet, unsteady balance, and slowed thinking — and every one of those overlaps with something already on a PCOS symptom list or already written off as an unavoidable trade-off of being on metformin. That overlap is not a coincidence worth shrugging off: in the DPPOS data above, neuropathy was more common specifically among metformin users with low B12, not among metformin users as a group, which is a biological reason to separate “metformin gave me this” from “my B12 gave me this,” because only one of those is a fixable lab value. The fuller list of causes behind PCOS fatigue covers the other candidates worth ruling out — thyroid function, sleep apnoea, insulin resistance itself — and B12 belongs on that list specifically for anyone on metformin long-term, not as a default explanation for everyone tired with PCOS.

The same confusion runs the other way with metformin’s gastrointestinal side effects, which are common, dose-related, and usually show up in the first weeks. The standard list of metformin side effects covers what that early pattern actually looks like — nausea, cramping, diarrhoea. B12-related fatigue and neuropathy are a different problem on a different timeline: they build slowly over years, not weeks, and food does not fix them. If new numbness, tingling, or balance trouble shows up after a long stretch on metformin, that is a reason to ask for a blood test, not to assume it is one more entry on the same early side-effect list.

Who Else Is at Extra Risk Beyond Long-Term Metformin?

Two other groups carry real, additive B12 risk that has nothing to do with metformin, and PCOS readers land in both more often than average. The first is anyone eating a vegan or vegetarian diet, since B12 occurs naturally only in animal-derived foods. The 2024 meta-analysis above, spanning 19 studies, found vegans had significantly lower serum B12 and significantly higher homocysteine than omnivores, and — encouragingly — also found that B12 supplementation among vegans meaningfully improved every biomarker measured compared with unsupplemented vegans, which is the clearest piece of good news in this whole picture: this particular risk factor responds well to a fix once it is identified. What a vegan or vegetarian diet needs to account for with PCOS covers the wider nutrient picture, of which B12 is one part.

The second group is anyone on a long-term proton pump inhibitor (PPI) or H2-receptor antagonist for reflux, since stomach acid is what releases B12 from food for absorption in the first place. A case-control study of over 200,000 people in a single US healthcare system found that two or more years of PPI use raised the odds of B12 deficiency by 65%, and two or more years of an H2-blocker raised it by 25%, with higher PPI doses carrying a larger effect than lower ones. Anyone stacking two or three of these — long-term metformin, a daily PPI, and a vegan or vegetarian diet — is stacking independent risk factors on the same nutrient, which is a reasonable thing to mention explicitly at your next blood draw rather than assuming any one of them alone explains how you feel.

Does Taking B12 Treat PCOS?

No. B12 supplementation corrects a nutrient deficiency — one that metformin, a restrictive diet, or long-term acid suppression can cause — and correcting it can resolve the fatigue and neuropathy that deficiency itself produces. It does nothing to insulin resistance, ovarian androgen production, cycle regularity, or any of the mechanisms metformin itself is prescribed for. No trial has tested B12 supplementation as a treatment for PCOS symptoms, and nothing in the evidence above suggests one would find an effect beyond correcting the deficiency itself. If B12 is low, correcting it is worth doing for its own sake — it is not a step toward treating PCOS.

Who This Does Not Help

If your B12 level is already normal, on the standard test or the functional markers, there is no evidence that supplementing further does anything beyond what ordinary nutrition already provides — more is not a hedge against a problem you do not have. And if fatigue or brain fog is your primary complaint but you are not on long-term metformin, not on a long-term PPI, and eating a B12-replete diet, B12 status is unlikely to be the explanation; the other causes on the PCOS fatigue page are the more probable place to look next.

You may see PCOS written as polyendocrine metabolic ovarian syndrome (PMOS), after a 2026 global consensus of more than 50 organisations renamed it. Nothing about metformin’s effect on B12 changed with the name — this article uses PCOS because that is still what most readers search.