Metformin immediate-release or extended-release
It cut new diabetes by about a third over three years and the effect lasted 15 years, but lifestyle change did better, and long-term use lowers vitamin B12.
- Research evidence score
- 8/10 · EstablishedGood human evidence that it works for this.
- Judged for
- Preventing type 2 diabetes in people with prediabetes
Measure this first
What decides whether this treatment is worth anything to you specifically.
Prediabetes is defined by it, and it is the number any change is measured against.
Long-term metformin lowers B12 — in the prevention trials one in five on metformin had a low or borderline level. A baseline makes a later result readable.
Metformin is cleared by the kidneys, and the dose depends on how well they work.
Claim by claim
Each claim carries what kind of study is behind it, what that study found, and how much it settles. Those are three different questions, and collapsing them is how a thin case is made to look strong.
“Metformin prevents type 2 diabetes in people at high risk.”
The Diabetes Prevention Program randomised 3,234 people with raised blood sugar, 68% of them women, to placebo, metformin 850 mg twice daily, or an intensive lifestyle programme. Over an average of 2.8 years, new diabetes occurred at 11.0, 7.8 and 4.8 cases per 100 person-years: metformin cut it by 31% (95% CI 17 to 43) and lifestyle by 58%. About 14 people had to take metformin for three years to prevent one case.
“Metformin's protection against diabetes lasts over the long term.”
Following 2,776 of the same people for an average of 15 years, diabetes was still 18% less common in the metformin group than the placebo group (hazard ratio 0.82, 95% CI 0.72 to 0.93), and 27% less common in the lifestyle group, though the gaps narrowed over time. By year 15, 56% of the metformin group had developed diabetes against 62% on placebo. Small-vessel complications — kidney, eye and nerve damage — did not differ between the groups overall; in women, the lifestyle group had fewer, and the metformin group did not.
Harms and limits, checked
These do not move the score above, which is only about whether it works. They are about what it costs and where it stops — a “no effect” on a harm means it was looked for and not found, and a “negative” on a limit means the limit is real.
“Metformin prevents diabetes as well as lifestyle change.”
It does not. In the Diabetes Prevention Program the lifestyle programme — at least 7% weight loss and 150 minutes of activity a week — was significantly more effective than metformin: 58% against 31% fewer cases, with about 7 people needing the programme to prevent one case against about 14 on metformin.
“Long-term metformin lowers vitamin B12.”
It does. In the same programme, low or borderline-low B12 was more common with metformin at 5 years (19.1% against 9.5%) and at 13 years (20.3% against 15.6%), and each year of metformin raised the odds of deficiency by 13%. Anaemia was more common in the metformin group, though not explained by B12 status, and nerve damage more common in those on metformin with low B12. The authors advised considering routine B12 testing.
What is still missing
- For diagnosed type 2 diabetes, metformin is usually the first medicine. That is a different use from prevention and is not what this entry scores.
Our verdict
Metformin delays type 2 diabetes: about a third fewer cases over three years, still 18% fewer after fifteen. Lifestyle change did better in the same trial, so it comes first, with metformin a well-evidenced addition when that is not enough. Long-term use lowers B12, which a periodic test catches.
Do we sell this?
Yes. It is one of our medication pathways, where online enrollment opens soon.
See what a request involves →This is education, not a recommendation for you. A licensed clinician decides what is appropriate, may suggest something different, and may decide no prescription treatment is right. Medical disclaimer.