Metformin found to reduce Long Covid in clinical trial.
The inexpensive diabetes drug performed well compared to placebo.
A study posted on a preprint server operated by the medical journal The Lancet reports that Long Covid diagnoses were found to be lower in patients who received metformin (an inexpensive diabetes drug) compared to those who received placebo. The study has not yet been peer reviewed. Still, the findings are promising.
Around 10.6% (1 in 9) of patients in the placebo group were diagnosed with Long Covid compared with 6.3% (1 in 16) of patients who took metformin for two weeks following their initial SARS-CoV-2 diagnoses. This 42% reduction in Long Covid means that 23 Covid-19 patients would need to be treated in order to prevent one Long Covid diagnosis. That’s an important reduction.
I read through the study and spoke to the senior author of the paper this evening to have some questions answered.
Inside Medicine is a reader-supported publication. To support this work, consider becoming a paid subscriber.
With the caveat that I still need to read the appendix (which has not yet been made public), here are some key points from my reading and our conversation:
In addition to a reduction in Long Covid among participants who received metformin, vaccinated people had lower rates of Long Covid. Boosted people had even lower rates (although the study took place at a time when a small proportion of people had been boosted, so it’s unclear how “real” that signal may be).
This study enrolled participants who met criteria for being overweight or having obesity. We do not know whether this would work for people with a body mass index below 25.
Metformin has been shown to have some antiviral properties. Still, that does not mean that it worked in that way, necessarily. It’s possible that the drug works by helping they body’s metabolic system. If the drug were also found to help people without obesity or overweight status, that might suggest that the drug is working as an antiviral. We just do not know.
Hospitalized patients were far more likely to have Long Covid diagnosed. Around 33% of hospitalized patients eventually were given Long Covid diagnoses (per the author; I don’t have those data and he was going on recall, so I’ll update this if needed) compared to around 7-8% in non-hospitalized patients.
Metformin costs around $1 per month. Nobody is going to get rich off of this, unless they gouge the prices. It’ll be interesting to see if online sellers of Long Covid treatment have any interest in a drug that is as close to free as anything out there.
A small but noticeable proportion of Long Covid diagnoses were made in the first 60 (and even 30) days after infection. That suggests that “medium”-term symptoms may be driving at least some of the high Long Covid rates we’ve seen elsewhere in the literature. Still, this study showed that Long Covid is not rare by any means. (Note: Long Covid was simply defined as people who were given that diagnosis by a medical provider. There were not any specific diagnostic criteria in use).
The combination of being vaccinated and receiving metformin might reduce Long Covid additively or synergistically. That means that it’s possible (we do not know, though), that the combination might eventually be show to decrease Long Covid rates even further.
A little nugget I like about this trial is that researchers tested 3 drugs at once, and shared a single placebo group. This is an efficient way to test drugs during a pandemic! That meant that the investigators could look at metformin versus placebo and ivermectin as well as fluvoxamine and draw from the same placebo group for all 3 sub-studies. (Ivermectin and fluvoxamine both did not work against Long Covid; while unfortunate, those negative results add further credibility to the metformin results).
Bottom line: If a person who meets criteria for obesity or overweight status were to ask me if they should take metformin (for two weeks) starting as soon as they learn they have Covid-19, I would say yes in many if not most cases, based on this new data. Metformin may decrease acute Covid-19 hospitalizations and now has been shown in a randomized placebo-controlled trial to decrease Long Covid diagnoses. It’s a safe medication and it’s inexpensive. This is starting to look like a real win.
This is really promising. As a primary care doc with great familiarity and comfort prescribing metformin all day to people with diabetes, the learning curve should not be as steep compared with Paxlovid.
With a good number of my patients in their 40's sustaining long Covid to varying degrees, including disability, I plan on acting aggressively with both Paxlovid and metformin in the future. I will at least offer it, and waiting for studies on non-overweight patients does not seem prudent, given the pleiotropic affects of metformin (anti-inflammatory, anti-thrombotic, and even antiviral in vitro).
Below is my angle from primary care, but feel free to pick at this if you think it's off in any way. I also busted out a little home made graphic so don't laugh!
Thanks for a timely heads up.
Would you limit the metformin to overweight people? I believe the dosage was 500 day 1, then twice a day, then 500 am/ 1000 mg pm. It will be an off label prescription, but it’s so cheap and commonly prescribed. Thanks for the review.