Metformin Cuts Clinician-Diagnosed Long COVID Risk by 50% in ACTIV-6 Trial
Key Insights
Metformin did not meet the prespecified efficacy threshold for preventing patient-reported COVID-19 (search) symptoms at 6 months in the ACTIV-6 trial of nearly 3,000 outpatient adults.
Participants receiving metformin experienced a 50% relative reduction in clinician-diagnosed long COVID (search) compared with placebo (0.56% vs. 1.17% at 180 days).
The trial enrolled participants between September 2023 and May 2024, with over 83% having prior immunity from vaccination or infection, making results relevant to the current pandemic phase.
A large randomized clinical trial has found that metformin, a widely available and inexpensive diabetes medication, reduced the risk of clinician-diagnosed long COVID (search) by approximately 50% when initiated during acute SARS-CoV-2 infection, even as the study's primary endpoint fell short of statistical significance.
The ACTIV-6 trial, published in Clinical Infectious Diseases and co-led by a University of Minnesota Medical School research team, evaluated 2,983 outpatient adults with mild-to-moderate COVID-19 (search) across 90 U.S. sites. Participants were randomized to receive either immediate-release metformin or placebo within seven days of symptom onset and were followed for six months.
Primary Endpoint Not Met, but Clinical Signal Emerges
The trial's prespecified primary endpoint—participant-reported COVID-19 (search) symptoms at day 180—did not meet the efficacy threshold. At six months, 79 participants (2.6%) reported symptoms they attributed to COVID-19, with 33 in the metformin group and 46 in the placebo group. The adjusted risk difference was 0.8 percentage points favoring metformin, with a 95% credible interval of −2.2 to 0.6. The posterior probability of efficacy was 0.83, below the 0.975 threshold required to declare efficacy.
Symptom rates remained low at earlier time points as well, with 3.2% reporting COVID-19 (search) symptoms at day 90 and 3.4% at day 120. The low event rate may be clinically important when interpreting results, particularly given that many participants had prior vaccination, prior infection, or both.
Clinician-Diagnosed Long COVID (search): A 50% Relative Reduction
A stronger signal emerged for clinician-diagnosed long COVID (search). By day 180, 8 participants (0.56%) in the metformin group and 18 participants (1.17%) in the placebo group reported receiving a clinician diagnosis of long COVID. Compared with placebo, metformin lowered the adjusted risk by 0.7 percentage points, with a posterior probability of efficacy of 0.96 and a risk ratio of 0.50.
"This trial provides additional evidence that treating acute infection with this intervention that acts on metabolic health can reduce the likelihood of developing long COVID (search)," said Carolyn Bramante, MD, MPH, assistant professor at the University of Minnesota Medical School and lead author of the study. "The finding is particularly important because metformin is inexpensive, globally available and has decades of clinical use supporting its safety."
Study Population Reflects Contemporary Immunity Landscape
The quadruple-blinded, randomized, placebo-controlled, decentralized platform trial enrolled participants between September 2023 and May 2024. The median age was 47 years, 63% were female, 47% identified as Hispanic or Latino, and 83% reported at least one prior COVID-19 (search) infection or at least two SARS-CoV-2 vaccine doses—making the results highly relevant to the current phase of the pandemic.
Participants had confirmed SARS-CoV-2 infection and at least two COVID-19 (search) symptoms within seven days of symptom onset. They received treatment for 14 days.
Safety and Replication
Safety findings were reassuring, with no episodes of lactic acidosis reported. Six participant-reported hypoglycemia events occurred—two with metformin and four with placebo.
The findings replicate results from the earlier University of Minnesota-led COVID-OUT randomized trial, which reported a similar reduction in long COVID (search) among participants treated with metformin during acute infection.
"Reproducing research is very important, and both trials have also been replicated in analyses of electronic health record data," said David Boulware, MD, MPH, professor at the University of Minnesota Medical School and steering committee co-chair of the trial. "Together, these independent studies support that in low- to high-risk adults, metformin is an effective strategy to reduce the risk of long COVID (search)."
Exploratory analyses suggested that participants without known prior infection may have been more likely to benefit, although the authors caution that these findings require further investigation. Next steps include examining biospecimens taken during acute infections and exploring whether similar actions exist during other infections. The research was funded by the National Center for Advancing Translational Sciences (search) (NCATS).