Thread by @ZdenekVrozina

Metformin cuts long COVID risk.
A massive UK study (n = 624k) just confirmed.
Metformin, started within 3 months of COVID-19, may significantly reduce the risk of post-COVID condition.
Let’s unpack 🧵

https://twitter.com/atranscendedman/status/1962594416842912103

The study tracked over 624,000 people with overweight or obesity who had tested positive for COVID-19 between 2020 and 2023.
Only 3,000 of them started metformin within 90 days after infection.
Everyone else served as a control group.

Outcome: long COVID (PCC)
Defined as:
a PCC diagnosis or
at least 1 symptom from the WHO list (eg brain fog, fatigue, cough, breathlessness) starting 90–365 days post-COVID.
Symptoms had to be new (not present in the 180 days before infection).

The results?
Metformin group had a hazard ratio (HR) of 0.36
64% relative risk reduction for long COVID
But the absolute risk reduction was only 12.6%
That’s a crucial distinction.

Let’s translate that into real-world terms.
Among people with overweight/obesity -
If they didn’t take metformin, about 1 in 3 developed long COVID.
If they did take metformin, it was 1 in 5.
That means 12 out of every 100 were spared long COVID thanks to metformin.

This effect held regardless of:
age
sex
diabetes status
BMI class
or SARS-CoV-2 variant (Alpha, Delta, Omicron)
And it was especially strong in people under 45.

The study tested when metformin was started (0, 30, or 60 days post-infection).
Results:
HR ranged from 0.34 to 0.39
Absolute risk drop: 11–13%
So, effect was robust even with delayed start - but early is probably better.

Mechanistically, metformin might work through -
lowering inflammation (IL-6, TNFα, etc)
activating AMPK
modulating host-virus interactions
reducing viral load and protein synthesis via mTOR inhibition
Still under investigation.

Robust findings.
Adjusted for confounding using target trial emulation
Sensitivity analyses confirmed the effect
Negative control (cancer) showed no effect, ruling out general healthy user bias
Solid observational science.

Most people didn’t get long COVID, even without metformin.
And 1 in 5 still did, despite taking it.
So, metformin is not a magic bullet - but a real, low-risk way to reduce odds.

Importantly, this study only looked at people with BMI ≥25.
We don’t yet know if the same benefit applies to those with normal weight.
Further RCTs needed - but the signal is consistent.

Chaichana at al., Effect of Metformin on the Risk of Post-coronavirus Disease 2019 Condition Among Individuals With Overweight or Obese: A Population-based Retrospective Cohort Study. academic.oup.com/cid/advance-ar…