Preliminary data from a randomized clinical trial may point to a potential way to prevent long covid. The trial found that covid-19 patients who took metformin, a commonly prescribed diabetes drug, were noticeably less likely to be diagnosed with long covid up to 10 months later than people in the control group. More research will be needed to confirm the implications of this study, however.

In late 2020, a large group of researchers at various universities across the U.S. began the COVID-OUT trial. Its original intent was to test whether three easily available and affordable oral medications could prevent covid-19 cases from worsening if taken early in a person’s illness. These medications—metformin, ivermectin, and fluvoxamine—had shown some potential antiviral and/or anti-inflammatory effect in previous studies, though mostly in the lab or in animals.

Outpatients were randomized into six groups, with all of the groups receiving two pills each. Half of these groups received metformin, either with a placebo or in combination with the other two drugs, and half received a placebo that looks like metformin, either with a second placebo or the other two drugs. This study design meant that many treatment comparisons could be made at once, including against a completely placebo-controlled group.

The study enrolled just over 1,300 patients over the age of 30 considered to be at higher risk for severe covid-19. Unfortunately, it found that none of the drugs met the pre-established goal set by the researchers, meaning a clearly significantly reduced risk of a serious covid-related event, such as hypoxemia (low blood oxygen). Some data did suggest that metformin might reduce the risk of a covid-related emergency department visit, hospitalization, or death, though. And even before the study ended, the researchers decided to keep studying their patients after the initial trial period, in hopes of seeing whether any of these drugs could have a preventive effect on long covid.

They kept track of 1,125 patients for 300 days, or about 10 months. The team asked patients via monthly survey if they had received a diagnosis of long covid from a medical provider during that time. Overall, 8.4% of patients (94/1,125) said yes, but those who took metformin were noticeably less likely to do so. 6.3% in the metformin group said they had been diagnosed, compared to 10.6% in the control—an over 40% reduction in relative risk. Those who took ivermectin or fluvoxamine were not less likely to report a long covid diagnosis compared to controls, however.

“Among adults with Covid-19, outpatient treatment with metformin decreased the development of long covid by 42% in a phase 3 randomized trial whose sample was mostly vaccinated and included enrollment during the Omicron wave,” the authors wrote in their paper, released as a preprint over the weekend on medRxiv.

It’s important to note the caveats of this research. For one, the data has yet to be formally peer-reviewed. Secondly, while these results do come from a randomized controlled trial—often considered the gold standard of clinical research—they’re technically not part of the study’s original design. Another limitation, common to long covid studies, is that the definition of this chronic condition remains difficult to pin down. And of course, no single study should be seen as definitive proof of a treatment’s effectiveness.

At the same time, there’s been very little clinical research conducted so far that’s even tried to look for possible preventive treatments of long covid. Much of the data showing that vaccines can reduce the risk of long covid, for instance, comes from retrospective observational studies, which are valuable but typically come with even more limitations than data from a randomized controlled trial. The researchers also openly made the decision to study long covid very early on into the trial, which is known as establishing a pre-specified outcome. This helps avoid the possibility of someone cherry picking nice-sounding data after the fact. And while there’s likely to be several reasons why long covid happens, the antiviral and anti-inflammatory effects of metformin might provide a plausible mechanism for its possible benefits.

There are clinical trials underway now and in the future that will study potential treatments and preventatives for long covid. So these findings, at the very least, suggest that metformin is worthy of a closer look as one of these options to be studied further, the authors say, including in combination with other drugs such as the antiviral Paxlovid.

“Further clinical trials are warranted to assess whether there is synergy with other treatments,” they wrote.