How Much Weight Can You Lose on Metformin: What Trials Show

Most clinical trials show metformin produces a modest weight loss, typically in the range of 2 to 6 kilograms over six months to a year, depending on the population studied and the dose used. That is far less dramatic than newer injectable weight-loss drugs, but the effect is real, durable across years of continued use, and supported by decades of trial data in multiple patient groups. The story gets more interesting when you look at who responds best, how the drug actually suppresses appetite, and what happens when you stop taking it.

What the Diabetes Prevention Trials Found

The most influential data on metformin and weight come from the Diabetes Prevention Program (DPP), a large U.S. trial that enrolled over 3,000 people at high risk of developing type 2 diabetes. Participants were assigned to either metformin, an intensive lifestyle program, or placebo. The lifestyle group lost more weight initially, but they partly regained it over time. The metformin group lost less up front but held on to their loss more consistently.

In the long-term follow-up study (DPPOS), which tracked participants for a decade after randomization, weight loss remained significantly greater in the metformin group than in the placebo group: about 2% of body weight versus 0.2%. That difference was closely tied to how consistently people kept taking their metformin. Those who stuck with it lost more; those who dropped off did not.¹ The lifestyle group, by contrast, lost more initially but drifted upward as behavior changes faded.²

A separate analysis of that same trial identified what predicted whether someone achieved lasting weight loss with metformin. Greater weight loss in the first year was the strongest signal, along with older age and continued use of the drug. If you lost a meaningful amount in the first year, the odds of keeping it off long-term were considerably better.³

Weight Loss in People Without Diabetes

Metformin is prescribed off-label for weight management in people who do not have diabetes, and the trial evidence here is mixed but leans positive. A 2024 systematic review and meta-analysis of randomized controlled trials in people with obesity but no diabetes found that metformin reduced BMI by a modest amount compared to placebo, at doses ranging from 500 to 2,550 mg per day. The effect did not differ by baseline weight, dose, duration, or whether the person also had fatty liver disease.⁴

Individual studies sometimes show larger effects. One trial in people with obesity and no diabetes reported an average weight loss of about 5.8 kg over the treatment period, while untreated controls actually gained a small amount. People with more severe insulin resistance lost significantly more weight than those who were insulin-sensitive, but sex, age, and starting BMI did not change the outcome.⁵ A network meta-analysis also found that metformin lowered BMI in overweight and obese people without diabetes, though the effect was smaller than what lifestyle changes alone could achieve.⁶

The honest summary: if you are taking metformin for weight loss without having diabetes, expect a real but modest effect. The drug is not going to produce the kind of dramatic transformation that makes before-and-after posts go viral. What it can do is nudge the scale a few kilograms in the right direction and help keep it there.

Metformin and Polycystic Ovary Syndrome

Metformin has been used in PCOS management for decades, partly because insulin resistance is central to the condition. A 2023 systematic review and meta-analysis of randomized trials in women with PCOS found that metformin reduced BMI compared to placebo, with the effect concentrated in women whose BMI was 25 or higher. In that overweight-to-obese subgroup, the BMI reduction was roughly twice as large as the overall pooled estimate. For normal-weight women with PCOS, metformin did not significantly reduce BMI but did improve other metabolic markers.⁷

Dose appears to matter more in PCOS than in some other populations. A trial comparing 1,500 mg and 2,550 mg daily in obese women with PCOS found that the higher dose produced greater weight loss in the obese subgroup, roughly 3.6 kg versus 1.5 kg. In morbidly obese women, both doses produced similar losses of about 3.8 to 3.9 kg.⁸

Beyond weight, metformin in PCOS trials has been shown to improve menstrual regularity, lower testosterone levels, improve cholesterol, and reduce markers of inflammation. For many women with PCOS, the weight loss is useful but the broader metabolic improvement is the more compelling reason to take it.

Preventing Weight Gain from Antipsychotic Medications

Antipsychotic drugs, particularly clozapine and olanzapine, commonly cause substantial weight gain. Metformin is one of the few medications with evidence for blunting this effect. A Cochrane review and subsequent meta-analyses have found that when metformin is started at the same time as an antipsychotic, patients gain roughly 3 to 4 kg less than those on placebo.⁹ In patients already taking antipsychotics who had gained weight, metformin produced meaningful reductions averaging around 3.2 kg compared to placebo.¹⁰

A separate meta-analysis found that the effect was strongest in people who were new to antipsychotics and weakest in those who had been on chronic treatment for years, which makes intuitive sense: preventing weight gain is easier than reversing it.¹¹ An updated meta-analysis found that metformin’s effects could be maintained from 12 to 24 weeks of treatment.¹²

Children and Adolescents

Pediatric obesity is an area where metformin has attracted growing interest, though the evidence base is smaller and more cautious. A randomized trial in obese children with normal blood sugar found that those given metformin lost about 4.9 kg over the study period, compared to 1.7 kg in the diet-and-exercise-only group.¹³ A large retrospective study within an integrated health system found that at 6 months, children and adolescents on metformin had lost roughly 3.5 kg more than controls, and at 12 months the gap widened to about 5.3 kg. Systolic blood pressure and cholesterol also improved.¹⁴

A separate trial in obese, insulin-resistant children using body composition measurement found that metformin reduced fat mass more than placebo, not just total body weight. The kids on metformin lost more actual fat tissue, which is an important distinction from simply losing water weight.¹⁵

How Metformin Actually Causes Weight Loss

For years, the assumption was that metformin’s weight effects were just a side effect of the nausea and gastrointestinal upset that many people experience when starting the drug. That explanation never fully held up, because weight loss persists long after GI side effects resolve. Emerging research points to several distinct mechanisms working together.

The most significant recent discovery involves a hormone called GDF15. Research in both mice and humans has shown that metformin stimulates the liver to produce and secrete this protein, which acts on the brain to suppress appetite. In mice engineered to lack GDF15, metformin’s effects on food intake and body weight were eliminated entirely. In humans with type 2 diabetes, higher GDF15 levels were associated with greater weight loss on metformin.¹⁶ This finding was a genuine surprise to the field, because GDF15 had previously been studied mainly as a stress signal unrelated to diabetes treatment.¹⁷

Metformin also boosts the release of GLP-1, the same gut hormone targeted by semaglutide and other newer weight-loss injections. Studies using intestinal tissue have confirmed that metformin directly stimulates the cells that produce GLP-1 in the gut, and blocking GLP-1’s effects partly cancels out metformin’s glucose-lowering action.¹⁸ Metformin additionally increases secretion of peptide YY, another appetite-suppressing hormone, and alters bile acid handling in ways that further influence hunger signals.¹⁹

A third pathway involves the gut microbiome. A systematic review of 30 studies found that metformin consistently reshapes the bacterial communities living in the intestines, increasing some species and decreasing others.²⁰ Metformin use has been linked to increases in Akkermansia muciniphila and E. coli, along with decreases in other species, and these shifts are themselves associated with metabolic changes.²¹ The precise chain from altered gut bacteria to weight loss is still being worked out, but the pattern is consistent enough that researchers consider it a genuine contributing mechanism rather than a coincidence.

Dose, Duration, and When Weight Loss Plateaus

One counterintuitive finding across trials is that, outside of PCOS, higher doses of metformin do not reliably produce more weight loss. An analysis of 21 randomized trials found no significant dose-response relationship within the standard prescribing range. What did matter was time: there was a clear relationship between duration of treatment and weight change, with effects building gradually over weeks and months.²²

The exception is PCOS, where at least one head-to-head dose comparison found a meaningful difference. As noted earlier, obese women with PCOS lost roughly twice as much weight on 2,550 mg daily as on 1,500 mg.⁸ Why this dose-response relationship shows up in PCOS but not consistently in other populations is not well understood, though the high rates of insulin resistance in PCOS may play a role.

Most people reach a weight-loss plateau within six months to a year. The DPP data suggest that from that point forward, metformin functions more as a weight-maintenance drug than a weight-loss drug. That is actually one of its strengths compared to calorie-restriction diets, where regain is the norm. But if you are expecting the scale to keep dropping indefinitely, the data do not support that expectation.

How Metformin Compares to Newer Weight-Loss Drugs

The arrival of GLP-1 receptor agonists like semaglutide and tirzepatide has dramatically changed the conversation about pharmacological weight loss. Metformin, even in its best showing, produces weight loss in the single-digit kilogram range. Semaglutide at higher doses can produce 10 to 15% body weight loss in many patients, roughly two to three times what metformin achieves.

A small pilot study directly comparing the two in patients with antipsychotic-induced weight gain found that after 16 weeks, the semaglutide group lost an average of 4.5 kg (about 4% of body weight) while the metformin group lost 2.9 kg (about 2.5%).²³ This was a small, non-randomized study, so the numbers should be taken as suggestive rather than definitive. But the general pattern, that GLP-1 agonists produce substantially more weight loss than metformin, is consistent across larger trials in other populations.

That said, metformin has some practical advantages. It costs a fraction of what injectable GLP-1 drugs cost. It comes as a pill rather than a weekly injection. Its side-effect profile is well-characterized after decades of use. And for people whose primary goal is not dramatic weight loss but rather prevention of diabetes or management of metabolic risk factors, metformin’s total package of benefits may be more appropriate than a drug designed primarily for weight reduction.

Combining Metformin with Other Medications

In clinical practice, metformin is often paired with other drugs that have their own weight effects. A meta-analysis of combination therapy with SGLT2 inhibitors found that adding an SGLT2 inhibitor to metformin produced an additional 2 kg of weight loss beyond metformin alone.²⁴ A trial of empagliflozin added to metformin showed weight reductions of roughly 2.3 to 2.9 kg with the combination versus 1.2 kg with metformin plus placebo.²⁵

A real-world study of people with both diabetes and obesity found that those using combinations of metformin with SGLT2 inhibitors, GLP-1 agonists, and orlistat achieved the greatest weight loss, particularly younger patients with higher starting BMIs and the greatest reductions in blood sugar.²⁶ Metformin in these combinations serves as a metabolic foundation drug rather than the primary driver of weight loss.

Where the Weight Comes From

Not all weight loss is equal in metabolic terms. Losing visceral fat, the deep abdominal fat surrounding your organs, has far greater health implications than losing subcutaneous fat or lean tissue. Metformin appears to preferentially target visceral fat. Animal research has shown that metformin significantly reduces visceral fat mass while upregulating enzymes involved in fat burning in the liver and muscle.²⁷

Human trials reinforce this. A study comparing metformin to liraglutide in patients with both type 2 diabetes and fatty liver disease found that metformin reduced total fat mass as well as trunk and android fat specifically.²⁸ In insulin-resistant children, metformin reduced fat mass more than placebo, confirming that the weight loss involved actual fat loss rather than just shifts in fluid or muscle.¹⁵ This is a point worth emphasizing, because many interventions that move the scale do so partly through water loss or lean-mass loss, which are not metabolically meaningful in the same way.

Why Some People Respond and Others Do Not

If you have taken metformin and noticed no weight change, you are not imagining things. There is genuine biological variability in how people respond, and some of it has a genetic basis. The protein OCT1 (organic cation transporter 1) is responsible for shuttling metformin into liver cells, where much of the drug’s action takes place. Roughly one in five white Americans carries a variant of the OCT1 gene that reduces metformin transport. In clinical studies, people carrying these reduced-function variants had significantly weaker responses to metformin on glucose tolerance tests, and in mice engineered to lack OCT1 entirely, metformin’s blood-sugar-lowering effects were abolished.²⁹

Beyond genetics, insulin resistance appears to be a consistent predictor of who benefits most. Across multiple populations, people with greater insulin resistance tend to lose more weight on metformin. This makes mechanistic sense: metformin’s effects on glucose handling and appetite-related hormones are more impactful when those systems are more disrupted to begin with. It also partly explains why metformin seems to work better for weight loss in people with PCOS, prediabetes, or metabolic syndrome than in metabolically healthy individuals who simply want to lose a few pounds.

Immediate-Release vs. Extended-Release Formulations

Metformin comes in two main formulations: immediate-release (IR), typically taken two or three times daily, and extended-release (XR), usually taken once daily. A systematic review and meta-analysis comparing the two found a small, non-significant tendency toward greater weight loss with the extended-release form, about 1 kg more on average. More meaningfully, delayed-release and extended-release formulations significantly reduced gastrointestinal side effects compared to immediate-release, cutting the odds of GI problems roughly in half.³⁰

A head-to-head trial found that both formulations produced similar reductions in body weight and BMI after six months, but the extended-release version delivered better improvements in insulin resistance and cholesterol markers.³¹ For many people, the practical difference comes down to tolerability. If GI side effects are making you skip doses or abandon the drug entirely, switching to extended-release may help you stay on it long enough for the weight effects to accumulate, since adherence is one of the strongest predictors of success.

What Happens When You Stop

One legitimate concern with any weight-loss medication is what happens when you discontinue it. The data on post-metformin weight regain are surprisingly sparse given how long the drug has been around. A study tracking women with PCOS after stopping metformin (following both short-term and long-term treatment) found that insulin resistance markers did not rebound over six months of follow-up. None of the participants developed impaired glucose tolerance or diabetes in that window.³² That study focused on metabolic markers rather than weight directly, so the weight-regain question remains somewhat open. The DPP data offer indirect evidence: participants who stopped taking metformin during the follow-up period did tend to regain weight, which is consistent with the finding that continued adherence predicted sustained weight loss.¹

This mirrors the broader pattern seen with virtually every weight-management drug studied long-term: the benefits persist as long as you take it, and tend to erode once you stop. Metformin’s advantage is that it is considered safe for long-term continuous use, with a tolerability profile well-established over more than 60 years of clinical experience. For many patients, “staying on it indefinitely” is a realistic and medically reasonable plan, which sidesteps the discontinuation question altogether.