Metformin is cheap, old and — crucially — it lowers blood sugar by activating AMPK in the liver rather than by boosting insulin. Read on for a clear, evidence‑based explanation of the cellular mechanism, the slow B12 depletion risk most long‑term users aren’t told about, and what the ongoing TAME trial actually aims to answer about metformin and ageing.

Key Takeaways

  • Metformin lowers blood glucose by activating a protein called AMPK, which tells the liver to stop manufacturing sugar — it doesn’t rely on extra insulin.
  • Long-term users have a significantly higher risk of vitamin B12 deficiency, which can cause nerve symptoms that look identical to diabetic neuropathy; a simple annual blood test can catch it early.
  • A major NIH-backed trial (TAME) is currently testing whether metformin can delay heart disease, cancer, cognitive decline, and physical frailty in people who don’t have diabetes — results are still pending.

What Metformin Actually Does

Metformin isn’t insulin, and it doesn’t push your pancreas to release more of it. That distinction matters, because it shapes who can take it and how it behaves in the body.

The drug works in the liver, where it activates a protein called AMPK — AMP-activated protein kinase. Think of AMPK as the cell’s built-in fuel gauge. When it switches on, the cell reads the situation as low energy and starts conserving resources: in liver cells, that means dialling back gluconeogenesis, the process by which the liver manufactures new glucose from protein and fat and releases it into the bloodstream.

The result is lower blood sugar — not because there’s more insulin circulating, but because the liver has stopped flooding the system in the first place.

The gut side effects many people experience in the first weeks — nausea, loose stools, a distinctive metallic taste — are partly because metformin also alters gut bacteria and the gut lining. For most people, that settles once the body adjusts. Taking it with food helps significantly.

The B12 Deficiency Risk Most Long-Term Users Don’t Know About

Here’s the part that rarely makes it into the standard prescription handout.

Metformin interferes with calcium-dependent receptors in the small intestine, and one consequence is reduced absorption of vitamin B12. It’s not a sudden drop — it’s a slow, quiet drain that builds over years.

Studies have found B12 deficiency in anywhere from roughly 10 per cent to over 30 per cent of long-term metformin users. A large 2025 study from the All of Us Research Program — looking at nearly 15,000 adults — found that long-term users had a 67 per cent higher likelihood of B12 deficiency compared with non-users. The Diabetes Prevention Program Outcomes Study has also documented this link going back to 2016.

Here’s the distinction that matters: low B12 can cause peripheral neuropathy — tingling, numbness, and weakness in the hands and feet.

That’s also one of the most familiar complications of diabetes itself.

So some people have been managing those nerve symptoms for years under the label of diabetic neuropathy, without anyone checking whether metformin was a contributing factor. This doesn’t mean the drug caused all of it — but in some cases it may have been pulling in the same direction as the disease.

The fix is simple. A blood test, once a year, can catch a falling B12 level before it causes symptoms. That check isn’t yet universally built into clinical guidelines everywhere, which is why it’s worth raising with your doctor directly if you’ve been on metformin for years.

The Longevity Question Nobody Saw Coming

In 2014, a large UK study made an unexpected observation: people with type 2 diabetes who were taking metformin were surviving longer, in shorter-term analyses, than comparable diabetics who weren’t on the drug. In some of the data, they were even outperforming matched healthy adults who didn’t have diabetes.

A drug being used to treat a serious metabolic condition appeared — at least in the short term — to be beating not having the condition.

Subsequent analyses with longer follow-up made the picture more complicated, as observational data tends to. But the original observation was striking enough, and the cellular mechanism specific enough, to seed a serious scientific idea.

That mechanism is, again, AMPK activation. When you flip the cell into fuel-saving mode, you also switch on autophagy — the body’s cellular housekeeping process, where damaged proteins and worn-out components get broken down and recycled. Autophagy tends to slow as we age. The hypothesis is that metformin might be prompting cells to keep running that maintenance programme for longer.

This is what the TAME trial — Targeting Aging with Metformin — is designed to test. It’s an NIH-backed randomised controlled trial running across 14 sites in the United States, enrolling 3,000 adults aged 65 to 79 who don’t have diabetes. Rather than tracking one disease at a time, TAME is measuring whether metformin can delay the onset of a cluster of age-related conditions together: heart disease, cancer, cognitive decline, and physical frailty.

Results aren’t in yet. This is live science. The fact that this question attracted serious NIH funding and a rigorous trial design tells you something about how credible the hypothesis has become among researchers. But “AMPK activation promotes cellular housekeeping in the lab” is not the same claim as “metformin extends healthy human lifespan” — and the trial exists precisely because that gap hasn’t closed yet.

Where Metformin Fits in a Changing Treatment Landscape

Metformin doesn’t sit in isolation. For people managing type 2 diabetes, the drug landscape has changed considerably in recent years.

SGLT2 inhibitors — drugs like empagliflozin — work through a completely different mechanism: they prompt the kidneys to filter excess glucose directly into urine. Beyond blood sugar control, they carry separately proven benefits for the heart and kidneys. GLP-1 receptor agonists — semaglutide, tirzepatide — lower blood sugar, produce significant weight loss in many people, and have shown meaningful cardiovascular benefits in trial data of their own.

And for some people, medication isn’t the most powerful option available. Sustained dietary change, regular exercise, and meaningful weight loss can put type 2 diabetes into clinical remission in a significant proportion of people — remission meaning the disease is no longer detectable in the labs, not just managed. That’s now solid enough to appear in international guidelines.

Metformin holds its first-line position for good reasons: it’s inexpensive, it has nearly seven decades of safety data behind it, and it works. The landscape it sits in, though, is broader and more active than it once was.

The Bottom Line on Metformin

Metformin is neither a forgotten relic nor a miracle molecule — it’s a well-understood drug with a clear mechanism, a strong safety record, and an enduring place in treatment. The B12 story is real and worth knowing: if you’ve been on it for years, an annual B12 check is a straightforward, low-cost step that can catch a deficiency before it causes problems.

The longevity research is genuinely interesting, and the TAME trial is a serious scientific effort. But drawing conclusions before those results arrive — in either direction — overstates where the evidence currently sits. Watching those findings come in is worthwhile. Acting as if they’ve already arrived isn’t.

Frequently Asked Questions

Can I take vitamin B12 supplements to offset the deficiency risk from metformin? Supplementing B12 is a reasonable conversation to have with your doctor, particularly if a blood test reveals your levels are low or falling. Oral B12 supplements or dietary adjustments can correct a deficiency in many people. The key step is getting your level checked first — knowing your actual B12 status gives you and your doctor something concrete to act on.

Does metformin cause weight gain? No — metformin is generally weight-neutral, and some people experience modest weight loss. This is one feature that distinguishes it from older diabetes medications. It doesn’t produce the significant weight changes associated with GLP-1 receptor agonists like semaglutide, but it also doesn’t work against you in that respect.

Is the TAME trial giving metformin to healthy people? Yes, in the sense that the 3,000 participants don’t currently have diabetes. The trial isn’t testing whether metformin treats an existing disease — it’s testing whether it can delay the onset of several age-related diseases in people who are currently disease-free. That’s a genuinely novel question for a drug this old, and the trial design reflects how seriously the scientific community is taking it.

References

Advertisement
Advertisement