Early Metformin Linked to 53% Lower Risk of Long COVID or Death

Adults prescribed metformin within six days of a documented SARS-CoV-2 infection had a 53% lower risk of long COVID or death, according to the authors of a target trial emulation.¹ The study, published in the Journal of Medical Virology, used electronic health record (EHR) data from the National COVID Cohort Collaborative (N3C).¹

The analysis was designed to emulate COVID-OUT and ACTIV-6, two randomized, placebo-controlled trials of metformin for COVID-19.¹ Lead author Carolyn T. Bramante, MD, MPH, of the University of Minnesota Medical School, is the principal investigator of COVID-OUT.¹

The investigators conducted a new-user, active-comparator analysis of adults with documented SARS-CoV-2 infection between March 1, 2020, and September 30, 2023.¹ To emulate a placebo control, they used prescriptions for fluvoxamine, fluticasone, ivermectin, or montelukast. These medications have been used for acute COVID-19 but lack consistent trial evidence for preventing long COVID.¹

Eligible patients had a clinic visit both 0 to 6 months and 6 to 12 months before infection, and no metformin or comparator use in the prior 12 months.¹ Adults with indications for metformin, including diabetes, were excluded.¹ The primary outcome was a composite of long COVID or death within 180 days, with death included to account for competing risk.¹

Long COVID was defined by diagnosis code or computable phenotype.¹ Overall, 248 adults received a new metformin prescription and 9412 received a comparator prescription on days 0 to 6 of infection.¹ After entropy balancing, the weighted groups had a mean (SD) age of 53 (16) years, and 16% of patients were Black and 16% were Hispanic.¹

Long COVID or death occurred in 4.0% of the metformin group vs 8.5% of the control group (adjusted relative risk [aRR], 0.47; 95% CI, 0.25-0.89).¹ The authors noted their data are highly compatible with risk reductions anywhere from 11% to 75%.¹

How the Findings Fit With Prior Metformin and Long COVID Evidence

As Contagion reported in February, two randomized trials and multiple target trial emulations have shown early metformin reduces long COVID risk by approximately 40% to 50%.² In COVID-OUT, metformin lowered long COVID risk by about 41% over 10 months compared with placebo. The reduction was about 63% when metformin was started within three days of symptom onset.²

ACTIV-6 used the same dosing in a broader population and again found reduced long COVID risk, with no significant safety concerns.² Gastrointestinal adverse effects were mild, and hypoglycemia was rare and similar to placebo.²

The metformin data stand apart from a series of negative long COVID trials Contagion covered over the past year. In the RECOVER-VITAL trial of 959 adults at 69 US sites, extended-duration nirmatrelvir and ritonavir (Paxlovid; Pfizer) did not improve cognitive, autonomic, or exercise-related symptoms compared with placebo.³

Similarly, in the STIMULATE-ICP platform trial, colchicine and famotidine-loratadine produced only small reductions in fatigue at 12 weeks.⁴ Neither showed sustained benefit over usual care by 24 weeks, and rivaroxaban showed no significant benefit.⁴ Unlike those trials, which tested therapies in established long COVID, metformin was started during acute infection to prevent it.

Several limitations apply to the new analysis. As an observational study, it remains subject to residual confounding despite weighting, and long COVID ascertainment relied on EHR codes and phenotypes. The metformin group was small, which contributed to a wide confidence interval.¹

References

1. Bramante CT, Stürmer T, Huling JD, et al. Metformin at the time of COVID-19 infection and risk of long COVID: a target trial emulation study. J Med Virol. 2026;98(11):e71190. doi:10.1002/jmv.71190

2. Contagion Editorial Team. Metformin’s prospective role in preventing long COVID. Contagion. February 4, 2026. https://www.contagionlive.com/view/metformin-s-prospective-role-in-preventing-long-covid

3. Paxlovid shows no benefit for long COVID. Contagion. September 2, 2026. https://www.contagionlive.com/view/paxlovid-shows-no-benefit-for-long-covid

4. Bender K. Researching medications repurposed for long COVID. Contagion. July 21, 2026. https://www.contagionlive.com/view/researching-repurposed-medications-for-long-covid