Metformin: the oldest drug in longevity research

From a hedgerow plant to the most prescribed diabetes drug on earth

Metformin descends from Galega officinalis, goat's rue, used in European folk medicine for centuries. The active principle, guanidine, lowered blood sugar but was too toxic. Metformin was synthesised in the 1920s, overlooked while insulin dominated, revived by Jean Sterne in France in the 1950s, and only licensed in the United States in 1995.

It is now first‑line for type 2 diabetes in essentially every major guideline, and has a longer real‑world safety record than almost anything else in the field.

What it does, as far as anyone can say

The mechanism is not fully settled, which is a striking thing to say about a drug this widely used.

The main established action is on the liver: it reduces hepatic glucose production. Most explanations invoke AMPK activation, though evidence for AMPK‑independent effects is substantial, and work on the gut — altered bile acid handling, incretin signalling, changes to the intestinal microbiome — has grown in prominence.

What it does not do is force the pancreas to release insulin, which is why it does not usually cause low blood sugar on its own.

Why longevity research keeps returning to it

Several observational studies reported that people with diabetes taking metformin had mortality that compared favourably with expectations, sometimes with non‑diabetic comparison groups. That generated the hypothesis that it might affect ageing rather than only glucose.

Two cautions, both important:

• Observational comparisons of this kind are prone to systematic bias. People prescribed metformin differ from people prescribed other agents — in kidney function, in duration of disease, in how sick they are. Several later analyses accounting for those differences found the advantage shrinking.
• The trial designed to test it has not reported. TAME, intended to test whether metformin delays age‑related disease in people without diabetes, has faced prolonged funding difficulty.

So the honest position is: an interesting hypothesis, biologically plausible, not demonstrated in humans. Anyone selling it as an anti‑ageing treatment is ahead of the evidence, and the evidence is not close behind.

The exercise question

Worth knowing because it cuts against the longevity framing. Several studies have reported that metformin blunts some adaptations to exercise training — smaller gains in fitness, and in some work attenuated muscle hypertrophy, compared with exercise alone.

The findings are not uniform and the mechanism is debated. But for a healthy person considering it speculatively, the possibility that it interferes with the intervention that has the strongest evidence for healthspan is a real consideration.

What is well established

• In type 2 diabetes it lowers HbA1c effectively, is weight‑neutral or modestly weight‑reducing, and is inexpensive.
• In prevention, the Diabetes Prevention Program showed it reduced progression to diabetes in people at high risk — though intensive lifestyle change performed better.
• In polycystic ovary syndrome it has an established role.

The practical cautions

• Gastrointestinal effects are common at the start — nausea, diarrhoea, metallic taste. Slow titration and the extended‑release form both help.
• Kidney function governs whether it can be used, and at what dose. This is why creatinine and eGFR sit in our baseline panel.
• Vitamin B12 absorption falls with dose and duration, and guidance recommends periodic monitoring. B12 deficiency can cause nerve damage that does not fully reverse, and its symptoms are easily attributed to the diabetes.
• Lactic acidosis is very rare, serious, and the reason for the kidney rules and for temporarily stopping around contrast imaging or acute illness.

Where this clinic stands

Metformin is registered, cheap, well understood and appropriate for the indications it is approved for. Whether it has a place in someone without diabetes is a question for the prescribing physician, who should say plainly that the longevity case is a hypothesis rather than a finding.

Our baseline includes the two tests that govern its safe use — creatinine for the kidneys and the full blood count — and B12 sits on the addition list precisely for people taking it long term. That is the sequence: test, prescribe, monitor.

The boundary

Nothing here recommends metformin to anyone. It is a prescription medicine with real contraindications, and whether it suits you depends on your kidney function, your other medicines and what you are treating.