Does metformin lower your risk of cancer or dementia?

Metformin has been the most widely prescribed drug for type 2 diabetes for decades. Observational research in diabetic patients shows that users develop cancer less often and die from cancer less often than non-users. That pattern is intriguing, but the precise mechanism is not yet well understood. Moreover, the evidence comes almost exclusively from observational research, which is susceptible to bias: people who take metformin differ from people who do not in a wide variety of ways.

The picture varies by cancer type. For breast cancer the evidence is frankly contradictory: some studies found a lower risk, but a summary of 15 studies found no statistically demonstrable difference in women with type 2 diabetes (the risk was around 0.82, but the margin of uncertainty ran from 0.60 to 1.12, meaning the effect could be zero). For prostate cancer the picture is even clearer: a summary of more than 52,000 cases showed no association whatsoever. For lung cancer the evidence is too thin to draw any conclusion.

On the subject of dementia, the available studies give us nothing to go on. No studies were provided that examined that association, so presenting metformin as protection against dementia is an overclaim.

What is well supported: metformin is safe, well tolerated, and extremely inexpensive. That makes it interesting as a potential agent outside of diabetes, but demonstrable clinical benefit in healthy people or as a preventive agent against cancer has not yet been established. Anyone without diabetes who is considering using metformin for life extension or cancer prevention is doing so without evidential support. Always discuss this with your doctor first.