Metformin, a diabetes drug that costs pennies per pill, has emerged as one of the more promising repurposed treatments for COVID-19, particularly for preventing long COVID. In the largest randomized trial to date, metformin reduced the incidence of long COVID by roughly 40% compared to placebo, and the benefit was even stronger when treatment started early. The story behind these findings touches on viral load, inflammation, vascular protection, and the question of why a decades-old blood sugar medication would have anything to do with a respiratory virus.
The COVID-OUT Trial and Long COVID Prevention
The most closely watched evidence comes from COVID-OUT, a randomized, quadruple-blind, phase 3 trial that enrolled over 1,100 overweight or obese adults who had tested positive for SARS-CoV-2. Participants were treated as outpatients and followed for 10 months. By day 300, the cumulative incidence of long COVID was about 6.3% among those who received metformin, compared to 10.4% in the placebo group. That translates to a hazard ratio of 0.59, meaning metformin cut the risk of developing long COVID by roughly 41%.1The Lancet Infectious Diseases. Outpatient treatment of COVID-19 and incidence of long COVID over 10 months (COVID-OUT)
This trial was notable for its rigor. It tested three repurposed drugs simultaneously: metformin, ivermectin, and fluvoxamine. Only metformin showed a meaningful effect on long COVID. Ivermectin had a hazard ratio of 0.99, essentially indistinguishable from placebo. Fluvoxamine actually trended in the wrong direction, with a hazard ratio of 1.36.2PubMed Central. Outpatient treatment of COVID-19 and incidence of post-COVID-19 condition over 10 months (COVID-OUT) The fact that only one of the three drugs produced a benefit in the same trial, with the same patient population and the same measurement tools, makes the metformin finding harder to dismiss as a fluke.
A separate analysis using a large electronic health record database (the National COVID Cohort Collaborative) attempted to emulate a target trial among adults without diabetes or prediabetes during the Omicron wave. Metformin prescribed within a week of infection was associated with a 53% reduction in long COVID or death, with 4.0% of the metformin group developing the outcome versus 8.5% of controls.3PubMed Central. Metformin reduces the risk of Long COVID or Death over 6 months in an Emulated Target Trial
Metformin Lowers Viral Load
One of the more surprising findings is that metformin appears to have a direct antiviral effect. In a prespecified analysis of the COVID-OUT trial, researchers measured SARS-CoV-2 viral load in nasal swabs collected at baseline, day 5, and day 10. Participants randomized to metformin had a mean viral load reduction roughly 3.6-fold greater than those on placebo. They were also less likely to have any detectable virus at follow-up, and viral rebound (a higher viral load at day 10 than day 5) occurred in about 3.3% of the metformin group compared to roughly 6% with placebo.4PubMed Central. Favorable Antiviral Effect of Metformin on SARS-CoV-2 Viral Load in a Randomized, Placebo-Controlled Clinical Trial of COVID-19
Neither ivermectin nor fluvoxamine had any effect on viral load at day 5 or day 10 in the same trial, reinforcing the idea that metformin’s benefit is not just about managing symptoms or inflammation but may involve directly suppressing viral replication.5PubMed Central. Favorable Antiviral Effect of Metformin on SARS-CoV-2 Viral Load in a Randomized, Placebo-Controlled Clinical Trial of COVID-19
Lab studies support this. A randomized phase IIb trial using metformin glycinate found that the drug inhibited SARS-CoV-2 replication in cell cultures without toxic effects at relevant doses, and the effect held across multiple variants including Alpha, Delta, and Epsilon. In the clinical arm of that study, treated patients cleared the virus faster and needed less supplementary oxygen.6PubMed Central. Treatment with metformin glycinate reduces SARS-CoV-2 viral load: An in vitro model and randomized, double-blind, Phase IIb clinical trial
Hospitalizations and Deaths
An updated systematic review and meta-analysis of randomized controlled trials found that metformin reduced the combined risk of hospitalization, emergency department visits, or death, with a risk ratio of 0.60 compared to placebo. The authors rated the quality of evidence as moderate.7Reviews in Medical Virology. Metformin for COVID-19: an updated systematic review and meta-analysis of randomized controlled trials
In the COVID-OUT trial’s prespecified secondary analyses, the effect on acute outcomes followed a similar pattern. The adjusted odds ratio for a composite of emergency department visits, hospitalization, or death was 0.58 with metformin, while ivermectin (1.39) and fluvoxamine (1.17) showed no benefit.8PubMed Central. Randomized Trial of Metformin, Ivermectin, and Fluvoxamine for Covid-19 The primary endpoint of that trial, which focused on preventing hypoxia or ER visits within 14 days, did not reach statistical significance for metformin (adjusted odds ratio 0.84, confidence interval crossing 1.0). But the secondary endpoints, which captured more severe outcomes over a longer window, consistently favored metformin.
Observational data from earlier in the pandemic told a similar story, especially among people with diabetes. A retrospective analysis of hospitalized diabetic patients in China found in-hospital mortality of 2.9% among those already taking metformin versus 12.3% among diabetics not on the drug.9PubMed Central. Metformin Treatment Was Associated with Decreased Mortality in COVID-19 Patients with Diabetes in a Retrospective Analysis Observational studies like this cannot prove causation, since people taking metformin may differ from those who are not in ways that affect outcomes. But the consistency of the signal across study designs is what made researchers take it seriously enough to run the randomized trials.
Why Timing Matters
A recurring theme in the metformin-COVID literature is that the drug seems to work best when started early. In the COVID-OUT trial, a subgroup analysis found that participants who began metformin within four days of symptom onset had a potentially stronger reduction in long COVID risk (hazard ratio 0.37) compared to those who started later (hazard ratio 0.64).10The Lancet Infectious Diseases. Outpatient treatment of COVID-19 and incidence of long COVID over 10 months (COVID-OUT) The trial authors cautioned that subgroup analyses have limited statistical power, so this specific split should be taken as suggestive rather than definitive.
The ACTIV-6 trial, a separate large platform trial, struggled with this timing issue. The median time between symptom onset and drug delivery was five days, and only about 21% of participants received metformin within three days of symptoms.11JAMA Internal Medicine. Metformin and Time to Sustained Recovery in Adults With COVID-19: The ACTIV-6 Randomized Clinical Trial This delay may partially explain why some trials show weaker results than others. If the drug’s main mechanism involves suppressing early viral replication, as the viral load data suggests, getting it into patients a week after infection might miss the window where it can do the most good.
A population-based retrospective cohort study reinforced this point, finding that starting metformin within three days of a COVID-19 diagnosis in overweight or obese individuals reduced long COVID incidence by about 63%. The authors explicitly noted that findings may not apply to people with a normal body mass index.12PubMed Central. Effect of Metformin on the Risk of Post-COVID-19 Condition Among Individuals With Overweight or Obese
Who Benefits Most
Most of the randomized trial data on metformin and COVID comes from people who are overweight or obese, because the trials were designed around that population. COVID-OUT specifically enrolled adults with a BMI of 25 or above, reasoning that these individuals faced higher risk and that metformin’s metabolic effects might be particularly relevant for them. Whether the drug offers the same benefit to people at a healthy weight remains unclear.
Sex differences have also appeared in the data. A large U.S. retrospective cohort study of adults with obesity and type 2 diabetes found that metformin use before hospitalization was associated with reduced COVID mortality in women (odds ratio 0.76 after propensity matching) but not in men.13The Lancet Healthy Longevity. Outpatient metformin use and COVID-19 outcome in adults with obesity and type 2 diabetes in the USA This sex-specific effect is not fully understood. It may relate to differences in immune response, body fat distribution, or hormonal interactions with metformin’s pathways. Or it could be a statistical artifact that would disappear in a larger randomized trial. Researchers have flagged it as something that needs further investigation.
How Metformin Might Work Against a Virus
Metformin’s best-known job is lowering blood sugar by improving insulin sensitivity and reducing glucose production in the liver. But the drug has a much broader biological footprint than most diabetes medications, and several of those broader effects are plausibly relevant to fighting a viral infection.
One key mechanism involves the AMPK pathway, which metformin activates. This activation inhibits a cellular growth-signaling pathway called mTOR, which viruses often hijack to fuel their replication machinery.14Exploration of Drug Science. Metformin in COVID-19: a magical role beyond the hyperglycemia By turning down mTOR signaling, metformin may starve the virus of the cellular resources it needs to make copies of itself. This is consistent with the viral load data showing faster clearance in treated patients.
Metformin also has anti-inflammatory and immunomodulatory properties. Severe COVID often involves a runaway inflammatory response that damages blood vessels and organs. Research has described metformin’s ability to reduce endothelial dysfunction and confer vascular protection, which could reduce the microvascular complications and blood clotting events that make COVID dangerous.15PubMed Central. Therapeutic Potential of Metformin in COVID-19: Reasoning for Its Protective Role This vascular protection may be especially relevant for long COVID, which some researchers believe involves persistent microvascular damage or viral reservoirs maintained by endothelial dysfunction.
There is also a broader aging angle. Metformin has been studied extensively as a potential anti-aging drug because it affects many of the hallmarks of biological aging, including cellular senescence, mitochondrial dysfunction, and disrupted nutrient sensing.16PubMed Central. Metformin in aging and aging-related diseases: clinical applications and relevant mechanisms COVID outcomes are strongly linked to age, and some researchers have speculated that metformin’s broader effects on cellular health may explain why its benefits extend beyond simple blood sugar control.
Safety When Used for COVID
One of metformin’s biggest advantages as a repurposed treatment is its well-understood safety profile. It has been prescribed to hundreds of millions of people worldwide for diabetes since the 1950s, so its risks are not a mystery. The specific question is whether giving it to non-diabetic people during an acute viral infection introduces new problems.
The answer from trial data is reassuring. In both the COVID-OUT and ACTIV-6 trials, metformin did not cause clinically significant gastrointestinal side effects and raised no safety concerns. Hypoglycemia, the most obvious worry when giving a glucose-lowering drug to people who do not have diabetes, was vanishingly rare. In the ACTIV-6 trial, just 2 participants in the metformin group (0.14%) reported hypoglycemia, compared to 4 in the placebo group (0.26%).17Clinical Infectious Diseases. Preventing Long COVID With Metformin
Metformin does carry a well-known risk of lactic acidosis, a rare but serious condition, primarily in people with severe kidney or liver impairment. This is why the COVID trials excluded participants with significantly reduced kidney function. For otherwise healthy adults, short-term use during an acute infection appears to carry minimal additional risk beyond the mild nausea or diarrhea some people experience when first starting the drug.
How Metformin Compares to Other Repurposed Drugs
The early pandemic saw a rush of interest in repurposing existing drugs, from hydroxychloroquine to ivermectin to fluvoxamine. Most of these have since been tested in rigorous trials and found wanting. The COVID-OUT trial is particularly useful for comparison because it tested three drugs head-to-head against placebo in the same population.
For the primary composite endpoint of preventing hypoxia, emergency visits, hospitalization, or death within 14 days, metformin showed a trend toward benefit (adjusted odds ratio 0.84) but did not cross the threshold of statistical significance. Ivermectin (1.05) and fluvoxamine (0.94) were clearly ineffective. But for the more severe secondary endpoint of emergency visits, hospitalization, or death, metformin stood apart with an adjusted odds ratio of 0.58, while ivermectin (1.39) and fluvoxamine (1.17) again showed nothing.18PubMed Central. Randomized Trial of Metformin, Ivermectin, and Fluvoxamine for Covid-19
Paxlovid (nirmatrelvir/ritonavir), the antiviral developed specifically for COVID, remains the standard outpatient treatment for people at high risk. Its mechanism is different: it directly inhibits a viral protease. Metformin’s potential value lies partly in that it targets different pathways, is far cheaper, and does not carry the drug-interaction concerns that make Paxlovid complicated to prescribe for people taking common medications like certain blood thinners and statins.
Cost and Global Access
Metformin is one of the cheapest prescription drugs in the world. A course of treatment costs a few dollars in most countries and is manufactured generically on every continent. The COVID-OUT investigators explicitly highlighted this in their interpretation, noting that metformin is “globally available, low-cost, and safe.”19The Lancet Infectious Diseases. Outpatient treatment of COVID-19 and incidence of long COVID over 10 months (COVID-OUT) – Section: Interpretation This matters because most of the world’s population cannot access expensive patented antivirals. If even a fraction of metformin’s trial benefits hold up in real-world deployment, the drug could prevent long COVID at a scale that expensive treatments simply cannot reach.
The drug also requires no cold chain, no infusion centers, and no specialized monitoring for short-term use. You can pick it up at a pharmacy and take it at home. For health systems in low- and middle-income countries, these practical advantages are arguably as important as the clinical data.
What Remains Uncertain
For all its promise, the metformin-COVID evidence has real gaps. The largest randomized trial enrolled just over 1,100 people, all with a BMI of 25 or above, during a specific phase of the pandemic. Whether the benefits apply to lean individuals, children, or immunocompromised patients has not been tested in randomized trials. The Omicron-era observational data from the N3C database showed a benefit in people without diabetes,20PubMed Central. Metformin reduces the risk of Long COVID or Death over 6 months in an Emulated Target Trial but observational analyses cannot fully account for confounders the way randomized trials can.
The timing question is also unresolved. If metformin needs to be started within three or four days of symptoms to work optimally, that narrows the practical window considerably. Many people do not seek medical care until they have been sick for a week, and some do not test positive until several days in. Designing a real-world treatment pathway that gets metformin to patients fast enough is a logistical challenge that the clinical trials did not have to solve.
There is also the evolving nature of the virus itself. The trial data spans pre-Omicron and early Omicron variants. SARS-CoV-2 continues to mutate, and whether metformin’s antiviral and anti-inflammatory effects remain consistent against newer variants is an open question. The in-vitro evidence showing activity against multiple variants is encouraging but not definitive for strains that emerged after the studies were conducted.21PubMed Central. Treatment with metformin glycinate reduces SARS-CoV-2 viral load: An in vitro model and randomized, double-blind, Phase IIb clinical trial
Finally, the mechanism story, while plausible and increasingly detailed, still has not pinpointed exactly why metformin works against this virus and not others, or whether the same pathways that help in COVID would matter in a future pandemic with a different pathogen. Researchers studying metformin’s effects on aging and cellular health suspect the drug’s benefits are broad enough to extend beyond any single virus, but that hypothesis is a long way from proven.