What the research found

A large randomized trial (MAST) tested whether metformin could slow cancer progression in 408 nondiabetic men under active surveillance for low-risk prostate cancer. Participants received either metformin (850 mg twice daily) or placebo and were monitored over three years using scheduled biopsies and PSA testing. About 35% of men in each group experienced cancer progression—70 on metformin versus 74 on placebo—showing no meaningful difference between treatments.

This result challenges a long-standing hypothesis rooted in animal studies and observational data suggesting metformin might suppress cancer cell behaviour. The trial was specifically designed to test whether the drug could act as a "secondary prevention" agent in men who already had confirmed cancer but weren't yet requiring active treatment—a setting that seemed theoretically ideal for testing metformin's anti-cancer properties.

Why it matters for you

If you're considering metformin as part of a longevity or health-optimisation protocol, this trial reinforces a broader pattern: the gap between what metformin does in cells and animals versus what it actually accomplishes in humans is often wider than popular accounts suggest. The hype around metformin as a geroprotector has largely rested on observational studies prone to confounding—people taking metformin tend to be healthier or more health-conscious in ways that also improve outcomes independent of the drug itself.

For men tracking biomarkers related to prostate health (PSA, DHT, androgen sensitivity), this doesn't necessarily mean avoiding metformin if you have other reasons for taking it (metabolic health, insulin sensitivity). But it does mean you shouldn't expect it to meaningfully alter cancer risk or progression on its own. If you're interested in secondary prevention of low-risk prostate cancer, the evidence here points toward established options like dutasteride or enzalutamide—both with known mechanisms and documented (though modest) efficacy—rather than betting on metformin's off-label potential.

Caveats

• Limited outcome window: Three years of follow-up may be too short to detect slow changes in cancer biology, particularly in low-risk disease
• Healthy volunteer selection bias: The trial enrolled nondiabetic men willing to participate in a placebo-controlled study, which may not reflect real-world metformin users or broader populations
• Observational epidemiology remains conflicted: The source acknowledges the metformin literature is riddled with selection bias; this trial doesn't resolve why observational studies seemed to show benefit
• Active surveillance cohort specificity: Results apply only to men with low-risk, localized prostate cancer; effects in other cancer types or disease stages remain untested