University of Minnesota researchers conducted one of the world’s largest observational studies of COVID-19 patients and found that diabetes and obesity increased death risks, but that a common diabetes drug protects women.

Metformin reduced COVID-19 death risks by 21 to 24% in women with COVID-19 who were already taking it to manage their blood-sugar levels and their diabetes, according to the U study results, which were posted online this weekend in advance of publication in a medical journal.

The amount of protective benefit was surprising, even though four smaller studies had found some evidence that metformin helped patients recover from COVID-19, said Dr. Christopher Tignanelli, a lead author of the study and a critical care physician.

The gender difference in results is a biological clue to understanding the corona­virus and how to disrupt it, he added. “If we can understand the difference between men and women that really drove this ... that gives us a key treatment avenue that we can really go after.”

While nobody should seek metformin for COVID-19 based on these findings alone, Tignanelli said, the results provide hope and new directions for research against an infectious disease for which there is no vaccine. He said he is eager to set up a comparative clinical trial to determine if metformin works.

Few proven treatments exist, though a U.S. clinical trial this spring found that an antiviral drug, remdesivir, helped people hospitalized with COVID-19 before the onset of the most severe respiratory symptoms. Minnesota has received multiple shipments of the drug from the federal government, which is managing its distribution.

Mayo Clinic research continues with the use of donated plasma from people who recovered from COVID-19 as a therapy for others, and Tignanelli is in the middle of a 10-site trial to determine if a blood pressure drug, losartan, offers any benefit.

Studies at the U and elsewhere have diminished hopes of an anti-malaria drug, hydroxychloroquine, working against COVID-19, though.

COVID-19 is caused by a highly infectious and novel coronavirus that has now been found through testing in 33,227 people in Minnesota and has caused 1,384 deaths. That includes 1,095 deaths of residents of long-term care and assisted-living facilities who are at heightened risk due to their ages and poorer health.

A state analysis of the first 2,428 patients hospitalized with COVID-19 in Minnesota found that 40% of them had diabetes and other metabolic diseases. Hypertension and obesity also were commonly found in COVID-19 patients in hospital care.

Tignanelli’s study is based on a UnitedHealthcare database of 73,000 patients with COVID-19, but it focused on 6,000 who had obesity or diabetes. Among them, 2,000 patients had taken metformin.

Diabetes is a disease of elevated blood sugar, which might serve as a breeding ground for the virus. It’s also possible that the condition interferes with the immune system’s response to infection.

Tignanelli said he suspects that metformin works by reducing inflammation and an immune system response to COVID-19 that can be overly aggressive in some patients and result in deaths.

The observational study also found potential protective benefits from albuterol among patients in this group who had asthma.

The strongest benefit came from a lesser-used and more costly class of drugs called TNF alpha inhibitors. However, Tignanelli said there were only 38 patients in the group taking those drugs, so the findings mostly generated a hypothesis for further research.

The study also found higher mortality risks related to certain drugs, and no evidence of benefits from an anti-parasitic drug known as ivermectin, or from steroids.

The latter finding might sound at odds with a major British research study last week showing that one steroid, dexamethasone, was associated with decreased mortality, but Tignanelli said they are consistent. The U study examined drugs people were already taking when they suffered COVID-19, and there is evidence that steroids play a harmful role in the initial stage of infection by hastening the spread of the virus. Even the British study showed no benefits from steroids early on in the course of COVID-19, yet they worked later on in severe cases by preventing an overreaction by the immune system, Tignanelli said.

Both the British results and the latest U metformin study were released online in advance of publication in a peer-reviewed journal.

Tignanelli said his research group decided to expedite the release of this information given the need for solutions to the pandemic. The research has been submitted for formal publication in medical journals.