Head to head
Metformin against Acarbose
Two old, cheap oral diabetes drugs with very different reputations in longevity circles: acarbose produced one of the largest male lifespan effects in the mouse programme designed to catch false positives, and metformin's geroscience record is consistently null. In people, they have been compared directly, and only on glucose.
Column A
MetforminMetformin hydrochloride (biguanide)
The most discussed drug in longevity, with enormous human safety data and a geroscience record that is consistently null. It also blunts the muscle and fitness gains from training.
Column B
AcarboseAcarbose (alpha-glucosidase inhibitor)
Produced one of the largest male lifespan effects in the mouse programme built to catch false positives, and has almost no human geroscience data at all.
These two were tested against each other directly
A head to head trial, not two separate records compared by us.
HbA1c fell 1.17 percent on acarbose and 1.19 percent on metformin at week 24. At week 48 the reductions were 1.11 and 1.12 percent, a difference of 0.01 percentage points (95% CI 0.12 to 0.14, p = 0.8999). Serious adverse events occurred in 2 percent of each group and hypoglycaemic episodes in 1 percent of each.
What the trial was
48 week, randomised, open-label, non-inferiority trial at 11 sites in China. 788 patients newly diagnosed with type 2 diabetes and a mean HbA1c of 7.5 percent were randomised after a four week lifestyle run-in to 24 weeks of monotherapy with metformin or acarbose, followed by 24 weeks in which add-on therapy was allowed if glucose targets were missed. The non-inferiority margin was 0.3 percentage points.
What the authors concluded
The authors conclude that acarbose is similar to metformin in efficacy and is therefore a viable choice for initial therapy in Chinese patients newly diagnosed with type 2 diabetes.
Where this result is weak
This is a glucose trial, not an ageing trial, and nothing in it speaks to the mouse lifespan result that draws people to acarbose. It was open-label, ran 48 weeks, and enrolled only Chinese patients, whose carbohydrate intake plausibly favours a drug that blocks starch digestion. It was funded by Bayer, which makes acarbose. Add-on therapy after week 24 blurs what the second half measures.
MARCH (ChiCTR-TRC-08000231), Lancet Diabetes and Endocrinology, 2014
Side by side, on the facts we can check
| Attribute | Metformin | Acarbose |
|---|---|---|
| Category | Longevity | Longevity |
| FDA status | FDA approved for type 2 diabetes since 1994. Not approved for ageing or longevity. | Approved for type 2 diabetes. The originator US product was discontinued and generics are marketed; the Canadian originator listing was cancelled post-market in 2023 with a generic available. |
| Half-life | 4 to 9 hours (plasma), longer in erythrocytes | About 2 hours; acts locally in the gut with minimal systemic absorption |
| Molecular weight | 129.2 Da (165.6 Da as hydrochloride) | 645.6 Da |
| Mechanism | Inhibits mitochondrial complex I and activates AMPK, reducing hepatic glucose output and improving insulin sensitivity. The AMPK account is incomplete and the relative contribution of complex I inhibition, mitochondrial glycerophosphate dehydrogenase and gut-mediated effects is actively disputed between research groups. | Competitively inhibits intestinal alpha-glucosidases, slowing the breakdown of complex carbohydrates and blunting the postprandial glucose rise. Undigested carbohydrate reaching the colon is fermented by gut bacteria, which produces the characteristic side effects and may itself contribute to the metabolic effects. |
| Human studies cited | 3 | 2 |
| Legal status, US | Prescription only; approved for type 2 diabetes | Prescription only; approved for type 2 diabetes |
Frequently asked questions
What is the difference between Metformin and Acarbose?
Metformin: The most discussed drug in longevity, with enormous human safety data and a geroscience record that is consistently null. It also blunts the muscle and fitness gains from training. Acarbose: Produced one of the largest male lifespan effects in the mouse programme built to catch false positives, and has almost no human geroscience data at all.
Which has stronger research evidence, Metformin or Acarbose?
This database does not grade compounds. It counts what was run in people: 3 of the 4 studies cited on the Metformin profile were human work, against 2 of 4 for Acarbose. Those counts are of our own citation lists and understate any larger literature, and neither is a recommendation.
Are Metformin and Acarbose FDA-approved?
Metformin: FDA approved for type 2 diabetes since 1994. Not approved for ageing or longevity.. Acarbose: Approved for type 2 diabetes. The originator US product was discontinued and generics are marketed; the Canadian originator listing was cancelled post-market in 2023 with a generic available..
Related posts
Blog articles that cover Metformin or Acarbose, newest first.
Reviewed Sep 2026. This page sets two published records side by side. It is educational, it is not medical advice, and it contains no dosing protocols or recommendations for personal use.