Acarbose for longevity
Acarbose (alpha-glucosidase inhibitor). Aging itself. Reviewed Jul 2026.
One of the most reproducible lifespan extenders in mice, mainly in males, but there is no human longevity trial.
What it is
An approved diabetes drug that slows carbohydrate digestion — and, it turns out, may extend lifespan through more than just blood-sugar control.
The evidence
ITP: ~20% median lifespan extension in male mice (~4–5% in females). New work suggests an NAD–SIRT1–PKM2 axis beyond glucose absorption. No human longevity RCT yet.
What you can do today
This is available today. Whether it suits you depends on your health, your other medicines and your risks, so discuss it with a clinician who knows your history before starting or changing anything.
Sources
- Coverage, not a sourcex.com
No primary paper or registry record is linked for this entry yet. The evidence summary names the studies it relies on.
Change history
No ladder moves or corrections logged for this entry yet. Changes are dated on the change log.
Related
More in Aging itself
- Rapamycin for longevityExtends lifespan across many animal species, but there is no human outcome data for longevity and side effects such as poor wound healing and infection are real.
- VO₂max and strength for longevityIn large observational cohorts the top VO₂max quartile has about 4 to 5 times lower all-cause mortality than the bottom, and nothing pharmacological on this list comes close.
- RentosertibSix aging clocks moved in 42 lung-fibrosis patients over 12 weeks, a design that cannot separate slowed aging from a treated lung, and no healthy-volunteer or outcome data exist.
- Senolytics (dasatinib, quercetin, fisetin)Small human studies show the drugs can be given, with off-target side effects, but no durable benefit in people has been shown.
- Partial epigenetic reprogramming (ER-100)A first-in-human safety trial in the eye has started, but every efficacy result so far comes from animals.
- Therapeutic plasma exchange for agingA small trial lowered biological-age scores by about 1.3 years, a surrogate marker, and there is no outcome data.