Why Am I Gaining Weight on Metformin? The Real Reasons

Why Would an Insulin-Sensitising Drug Cause Weight Gain?

Metformin’s own trial record argues against the idea that it drives weight gain: a systematic review of randomised trials found metformin produced an average BMI reduction of 0.53 kg/m² compared with placebo (95% CI −0.95 to −0.12) — small, but on the losing side of zero, not the gaining side. That average is exactly the kind of number that hides what happened to any one person. A confidence interval running from a 0.95-point drop to a 0.12-point drop means some trial participants lost far more than half a point and others barely moved, and a pooled result sitting that close to zero at one end of its own range does not rule out individual participants moving the other way entirely.

The same pattern shows up even more starkly in metformin’s other headline PCOS effect. Pooled data found metformin raised the odds of ovulating 2.64-fold against placebo, but the 13 trials feeding that number disagreed with each other substantially — a heterogeneity statistic (I² = 61%) that means trials of the same drug, in the same condition, landed on meaningfully different results. Trials studying the same intervention disagree with each other that much; one person’s result differing from the pooled average is not an exception to the evidence, it is what the evidence already predicts.

None of that explains your specific number, and it is not meant to. It does rule out one common assumption — that metformin is “supposed to” produce weight loss for everyone, so a gain means something was done wrong. It does not. The mechanisms below are the more useful place to look; for metformin’s full trial record, see what metformin actually does for insulin resistance.

You may see PCOS referred to as polyendocrine metabolic ovarian syndrome (PMOS) after a 2026 global consensus of more than 50 organisations renamed it. Nothing about the mechanisms below changes with the name — this article uses PCOS because that is still what people search.

Could Restored Ovulation Be Adding Fluid Weight?

Metformin’s most consistent PCOS effect sits closer to the ovaries than to appetite, and restoring ovulation restores a hormone pattern that carries its own cyclical weight. The 2023 international guideline positions metformin as an adjunct for ovulation induction rather than the first-line drug — letrozole holds that place — but metformin alone still measurably raises the odds of ovulating, and for someone who was not cycling regularly before starting it, that change can be the biggest thing that happens in the following months.

A study that weighed 42 women twice weekly across a full cycle found body weight ran about 0.45 kg higher during menstruation than in the first week of the cycle, and traced almost all of that difference to a 0.474 kg rise in extracellular water — not fat. That study was not conducted in women with PCOS, and it is rising progesterone after ovulation, not metformin itself, driving the fluid shift, so this applies to you only to the extent metformin actually restores a cycle for you. Someone whose periods were already regular before starting the drug will not see this effect newly appear.

But someone who was not ovulating consistently, or at all, and starts doing so for the first time in months or years picks up a fluid fluctuation of roughly half a kilo that was not there to notice before. It recurs monthly, is most visible in the days before or during a period, and is gone again by the following week. Weighed at the wrong point in that new cycle, that reads exactly like “the metformin made me gain weight” — when it is closer to the metformin making a cycle happen again.

Are Appetite and Eating Patterns Shifting as Side Effects Fade?

Reduced appetite, not increased appetite, is part of how metformin produces its average weight effect in the first place. In the trial that first documented this, metformin reduced calorie intake in a dose-dependent way — women with obesity and type 2 diabetes ate measurably less and rated their hunger lower after a metformin dose than after placebo, and the metformin group lost more weight over 24 weeks than a placebo group on the same calorie-controlled diet. That trial was not conducted in PCOS, but the mechanism it describes is a direct effect on hunger and food intake, not on ovaries, and there is no obvious reason it would behave differently in PCOS than in type 2 diabetes.

Gastrointestinal side effects on metformin are well documented to be heaviest in the first weeks of treatment and to ease as the gut adjusts — how metformin’s side effects change over time covers that timeline in full. It is a reasonable, mechanistically consistent explanation, though not a separately proven trial finding, that as nausea and appetite suppression fade, eating drifts back toward whatever pattern existed before starting the drug. Nobody has run a trial that isolates this exact transition and measures the resulting weight change, so this is reasoning from an adjacent finding, not a number this article can hand you. What it does explain is why “it worked for the first month and then stopped” is a plausible sequence of events, rather than the drug quietly reversing itself.

Is Vitamin B12 Depletion Part of the Picture?

Long-term metformin measurably lowers vitamin B12, and the risk grows the longer someone stays on it. In the Diabetes Prevention Program Outcomes Study, people taking metformin 850 mg twice daily had low or borderline-low B12 in 19.1% of cases at five years, versus 9.5% on placebo, and each additional year of use raised the odds of deficiency by 13%. That trial recruited adults at risk of type 2 diabetes, not PCOS specifically, but the mechanism — reduced B12 absorption in the gut — is a property of the drug, not of the condition it happens to be treating.

B12 depletion does not put weight on a scale by itself, and this article is not claiming that it does. What it does is drain the energy available for movement, and it does so with a specific, checkable pattern: persistent fatigue that feels different from ordinary tiredness, tingling or numbness in the hands or feet, and sometimes brain fog. Every one of those overlaps with PCOS itself, which is exactly why a treatable deficiency gets waved off as “just PCOS” instead of tested. B12 is a blood test, not a guess — if it has been more than a year since yours was checked, that is worth naming at your next appointment regardless of what the scale is doing.

Could Something Else Be Happening at the Same Time?

Starting metformin rarely happens in isolation. A new prescription often lands alongside a new diagnosis, a change in sleep or stress, another medication change, or simply a season of life that would have moved the scale regardless. That does not make “I gained weight after starting metformin” an inaccurate report — it is not about willpower or discipline, weight here is a metabolic variable responding to several physiological inputs at once. But it is a different claim from “metformin caused this,” and the small average effect in Table 1 was only measurable because those trials isolated metformin from everything else in a person’s life. Real life does not run that experiment for you.

That is a reason to bring the actual timeline to your prescriber rather than assume the drug alone explains it: when the gain started, what else changed in that window, and whether your cycles, energy, or digestion shifted alongside it. The full PCOS weight-loss guide lays out diet, exercise, and other medications side by side with the same standard of evidence, and how insulin resistance actually drives PCOS covers the mechanism metformin is targeting in the first place, in more depth than fits here.

What Weight-Gain Pattern Needs More Than This Explanation?

Everything above describes gradual change over weeks to months that tracks loosely with a cycle, appetite, or energy shift. That is a different picture from weight that rises by several kilos over a few days, or that comes with swelling in the legs or face, shortness of breath, or a noticeable drop in how much urine you are passing. Those point toward fluid retention from a cause other than metformin’s ordinary effects and warrant same-day medical attention rather than a wait-and-see approach.

Nobody has run a trial that specifically isolates “why did this individual gain weight on metformin,” so everything above is reasoning from adjacent, verified findings rather than a single study answering the question directly. That is also why a pattern that does not fit any mechanism here — rapid gain, gain with no cycle or appetite change at all, or gain alongside new symptoms elsewhere in the body — is worth naming to whoever prescribed the metformin rather than filing away as unexplained.

Common Questions

This paradox is one piece of a larger picture. The weight-loss section of this site covers diet, exercise, and every other PCOS medication people ask about alongside it, with the same standard of evidence used here.