MINNEAPOLIS — A cheap, widely used diabetes drug taken soon after a COVID-19 infection was associated with a substantially lower risk of developing Long COVID, according to a large new study of United States health records — a finding the researchers say adds to growing evidence that metformin could help prevent the lingering condition, though it is not yet approved for that use.
Illustration: The study found the clearest benefit when metformin was started within about a week of a documented COVID-19 infection.
The study, published in the Journal of Medical Virology, found that people who were newly prescribed metformin within six days of a documented SARS-CoV-2 infection had a 53% lower relative risk of developing Long COVID or dying over the following six months, compared with people given other medicines that had been tested against COVID-19. The analysis drew on the National COVID Cohort Collaborative (N3C), a database of de-identified electronic health records covering more than 21 million people in every US state.
"Metformin prescription within the week following documented SARS-CoV-2 infection was associated with a 53% reduction in LC/D [Long COVID or death] over 6 months in this primarily Omicron-infected sample," wrote the authors, led by researchers from the University of Minnesota Medical School.
What Happened
The research team, led by Dr Carolyn Bramante of the University of Minnesota, used a method called target trial emulation — a statistical approach that designs an observational study to mimic a randomised trial as closely as possible. They analysed adults whose first documented SARS-CoV-2 infection occurred between March 2020 and September 2023, and compared two groups who received a new prescription in the days after their infection.
One group received metformin, the first-line treatment for type 2 diabetes. The other received one of four comparator medicines — fluvoxamine, fluticasone, ivermectin or montelukast — drugs that had also been studied as possible COVID-19 treatments. Crucially, the researchers excluded anyone who was already taking metformin, or who had a condition such as diabetes or polycystic ovary syndrome that would explain a metformin prescription, so the comparison focused on people taking the drug for the first time because of their infection.
After statistical adjustments to make the two groups comparable, the researchers followed 248 people in each group for 180 days. In the metformin group, 10 of 248 people (4.0%) developed Long COVID or died. In the comparison group, 21 of 248 (8.5%) did. That translates to an adjusted risk ratio of 0.47 — a 53% lower relative risk — with a 95% confidence interval of 0.25 to 0.89. The absolute difference between the groups was 4.5 percentage points.
The timing of the prescription appeared to matter. When metformin was started on the day of infection or the day after, the adjusted risk ratio was 0.39, although the confidence interval was wide (0.12 to 1.24). When the window was widened to prescriptions within 14 days, the risk ratio was 0.75 (0.52 to 1.08). The authors concluded that the benefit was clearest when the drug was started within about a week of infection.
Illustration: The analysis used de-identified health records from more than 21 million people in the US National COVID Cohort Collaborative.
Background
Long COVID — the persistence of symptoms for weeks, months or even years after a SARS-CoV-2 infection that cannot be explained by another diagnosis — remains one of the pandemic's most stubborn problems. Symptoms can include fatigue, brain fog and cognitive difficulties, breathlessness, sleep disturbances and problems with everyday activities. Some people improve over time; others live with prolonged illness that limits their ability to work or lead a routine life. Research has also suggested that brain inflammation may persist even after the virus is no longer detectable.
Metformin has been prescribed for decades, mainly to control blood sugar in type 2 diabetes, and researchers have been exploring whether its antiviral and anti-inflammatory effects could help during the early stages of a COVID-19 infection. Laboratory studies have shown that metformin can inhibit SARS-CoV-2 in cell cultures, and earlier clinical evidence pointed in the same direction as the new study.
In the COVID-OUT randomised trial of more than 1,000 outpatients, Long COVID occurred in 6.2% of people given metformin compared with 10.3% of those given a placebo by day 300 — a 41% lower risk — and the benefit rose to 63% lower among participants who started the drug within three days of symptom onset. A separate sequential target trial analysis in the United Kingdom, which assessed metformin started within 90 days of infection, reported a hazard ratio of 0.36. The new study's authors noted that their results were consistent with both findings.
The new paper, "Metformin at the Time of COVID-19 Infection and Risk of Long COVID: A Target Trial Emulation Study," was published in the Journal of Medical Virology in November 2026. Its authors include researchers from the University of Minnesota, the University of North Carolina, Vanderbilt University and the Scripps Research Translational Institute, including the cardiologist and geneticist Dr Eric Topol.
Why It Matters
If the association holds up in further research, the implications could be large. Metformin is inexpensive, widely available around the world, and has a safety profile that doctors have understood for decades. A cheap, familiar pill that cuts the risk of a disabling post-viral condition by half would be a rare piece of good news in Long COVID research, where proven preventive options remain scarce.
The study also matters because of how it was done. Randomised trials are the gold standard, but they enrol only people who agree to take part, which can limit how well the results apply to the general population. By analysing real-world health records from more than 21 million people, the researchers tested the question in a far broader sample — and reached a similar answer to the randomised COVID-OUT trial, strengthening confidence in the overall signal.
At the same time, the authors and independent experts urge caution. The study is observational: it shows an association, not proof that metformin caused the lower risk. The matched comparison involved only 248 people per group, a small number for drawing firm conclusions, and Long COVID was identified from clinician diagnoses recorded in health records, which may miss cases or misclassify others. The authors themselves wrote that "the replication of results in clinical trials is important for informing treatment guidelines."
Importantly, metformin is not approved anywhere as a treatment or preventive for Long COVID, and it is not a drug to take without medical supervision. It can cause side effects, including gastrointestinal problems, and in rare cases a serious complication called lactic acidosis, particularly in people with kidney problems. Anyone who tests positive for COVID-19 should discuss treatment options with their own doctor rather than self-medicating.
Illustration: Long COVID symptoms can linger for months or years — the study tracked whether early metformin could lower that risk.
What Happens Next
The next step is more evidence. The authors' call for replication points to further randomised trials testing metformin specifically as a Long COVID preventive, including in larger and more diverse populations and with longer follow-up. The ACTIV-6 trial, one of the studies the new analysis was designed to emulate, is among the randomised trials examining metformin during acute infection.
Regulators and guideline committees will also be watching. No health authority has yet recommended metformin for Long COVID prevention, and any change would require consistent results across trials, not a single observational study — however large the underlying database.
For now, the study adds a substantial piece to a growing puzzle: a familiar drug, taken at the right moment, may help the body avoid one of COVID-19's longest shadows. Whether that promise becomes a prescription will depend on the trials still to come.
Sponsored: explore today's featured sponsored offers
FAQ
What did the new metformin study find?
People newly prescribed metformin within six days of a documented COVID-19 infection had a 53% lower relative risk of developing Long COVID or dying over the next six months, compared with people prescribed other medicines. In the matched groups, 4.0% of the metformin group and 8.5% of the comparison group developed Long COVID or died.
How was the study conducted?
Researchers from the University of Minnesota and colleagues used target trial emulation on the US National COVID Cohort Collaborative, a database of de-identified health records from more than 21 million people. They compared new metformin users with people newly prescribed fluvoxamine, fluticasone, ivermectin or montelukast, following 248 people in each group for 180 days.
Does this prove metformin prevents Long COVID?
No. The study is observational, so it shows an association rather than proof of cause and effect. The matched sample was small, and the authors say replication in randomised trials is needed before treatment guidelines could change. The findings are, however, consistent with the COVID-OUT randomised trial and a UK analysis.
Is metformin approved for Long COVID?
No. Metformin is approved as a diabetes treatment, not for preventing or treating Long COVID. It should not be taken for this purpose without a doctor's advice, as it can cause side effects and, rarely, a serious complication called lactic acidosis.
What is Long COVID?
Long COVID refers to symptoms that persist after a SARS-CoV-2 infection and cannot be explained by another diagnosis. They can include fatigue, brain fog, breathlessness, sleep problems and difficulty with everyday activities, and may last for months or years.
Sources
- Bramante CT et al., "Metformin at the Time of COVID-19 Infection and Risk of Long COVID: A Target Trial Emulation Study," Journal of Medical Virology, November 2026, doi: 10.1002/jmv.71190
- Health and Me, "Common Diabetes Drug Metformin May Slash Long COVID Risk By More Than 50%," November 5, 2026
- Wiley Online Library, Journal of Medical Virology, Bramante et al., 2026
- PubMed, PMID 42954425
Comments
Post a Comment