An antidiabetic drug that the University of Minnesota targeted early in the pandemic as a treatment for COVID-19 also reduced risks of long COVID in a new study.

Only 6.3% of patients experienced long COVID within 300 days of their initial illnesses if they took metformin, according to U study results published Thursday. That compared to a long COVID rate of 10.4% among study participants who didn't take the drug, which is commonly used to lower blood sugar in people with Type 2 diabetes.

The finding comes amid a remarkable ebb in serious COVID-19 cases in Minnesota, where the number of patients admitted to hospital beds with the infectious disease declined from 402 on March 1 to 67 on Tuesday. The pandemic record was 1,864 COVID-19 hospitalizations on Nov. 29, 2020.

Even so, hundreds of Minnesotans are testing positive every week for infections with the coronavirus that causes COVID-19, said Dr. Carolyn Bramante, lead researcher of the U-led trial. "Anyone who is diagnosed with COVID now will be worried about the risk of long COVID. So, I think these results are very relevant."

The federal definition of long COVID is symptoms that linger for four or more weeks. Cases vary in their duration, intensity and symptoms, but they often involve fatigue, headaches, dizziness and cognitive problems (or brain fog) — especially after exertion or exercise.

The study results are part of the U's COVID-OUT clinical trial, which focused on metformin and fluvoxamine because the widely used drugs target the same proteins as the coronavirus. Ivermectin was added to the trial because of its anti-inflammatory benefits and because many people were championing it without evidence that it worked.

More than 1,300 patients across the country signed up for the trial shortly after they tested positive for COVID-19, and they received either combinations of the study drugs or nonmedicating placebo pills. Only metformin produced a statistically significant effect in the first phase of the trial, reducing COVID-related deaths and hospital visits.

U researchers expanded their goals beyond the short-term effects of the drugs when long COVID became an established problem. More than 1,100 patients extended their participation in the study and agreed to long-term monitoring.

Metformin was associated with a 41% reduction in long COVID in this group, according to the results published in the Lancet Infectious Diseases, a medical journal. The protective effect was stronger in people who were obese, female, younger or unvaccinated — although the small size of these subgroups limited the statistical significance of the findings.

Long COVID risks remained lower for participants who took metformin, even as the dominant variant of the coronavirus shifted from alpha to delta to omicron.

About 12% of Minnesota adults have experienced long COVID since the start of the pandemic in early 2020, and almost 5% were dealing with the condition at the time of a federal survey in early May. About 1 in 5 Minnesotans with active cases reported in earlier surveys that long COVID was significantly limiting their daily activities.

Bramante said the U study results only show a preventive benefit to taking metformin immediately after infection. Speed mattered. One of the groups with the lowest incidence of long COVID was study participants who took metformin within three days of their first COVID symptoms.

Whether the drug could help people currently struggling with the post-COVID condition is unclear.

"We didn't look at use of metformin after acute COVID," she said. "This was only for metformin taken right at the time when people got infected. So, it's preventing severe COVID, preventing long COVID, but we don't know if it would treat long COVID."

Follow-up studies are examining how metformin works. Bramante said the drug might affect the microbiome, or the collection of bacteria and other organisms in the body. One study will look at changes in microbiome levels in participants who submitted specimens before and after COVID.

Metformin appears to inhibit viral replication. Bramante and colleagues just posted a study on a pre-print server — ahead of publication in a peer-reviewed journal — showing lower viral loads in the noses of study participants who took the drug.

"We're all trying to learn more about long COVID," she said, "but one theory that makes sense is just having more virus in the body for longer might make you more likely to have long-term effects from the virus."