Metformin: What It Actually Does

Metformin is probably the most prescribed diabetes medicine in the world, derived from a compound found in a plant called French lilac and in continuous medical use since the 1950s. It is cheap, unpatented, and unglamorous, and yet it has outlasted dozens of flashier drugs. If you have just been prescribed it, or you have taken it for years without anyone really explaining it, here is what it actually does, what it does to you along the way, and what to make of the recent hype about it as an anti-aging pill.

The problem metformin is built for#

Type 2 diabetes is not a shortage of insulin at the start. It is a resistance to insulin's message. Insulin's job is to tell the liver to stop releasing glucose and to tell muscle and fat to take glucose in. When cells stop listening, two things go wrong at once: the liver keeps pouring glucose into the blood even when levels are already high, especially overnight, and muscles are slow to clear glucose after meals. The pancreas shouts louder with more insulin, keeps up for years, and gradually tires.

That overnight liver release is why many people with type 2 diabetes wake to their highest reading of the day, having eaten nothing since dinner. It puzzles almost everyone, and it is precisely the fault metformin targets.

What metformin actually does#

It quiets the liver. Metformin's main action is to dampen the liver's glucose production. Inside liver cells it subtly shifts the energy state, activating a fuel-sensing enzyme system called AMPK, which acts like a factory manager noticing the power bill and switching off nonessential lines. Glucose output falls, and fasting readings, especially morning ones, come down.

It helps muscle listen to insulin again. A secondary effect is improved insulin sensitivity, so the same amount of insulin moves more glucose into muscle.

It works on the gut more than we once thought. Metformin concentrates heavily in the intestinal wall. There it increases glucose use by the gut itself, boosts hormones such as GLP-1 that improve insulin responses and blunt appetite, and measurably alters the gut bacterial population. This gut action helps explain both a slice of the benefit and most of the side effects.

Notice what is missing: metformin does not force the pancreas to release more insulin. That distinction matters enormously in daily life. Drugs that push insulin out can drop blood sugar too far. Metformin used alone very rarely causes hypoglycemia, does not demand snacking to keep sugar up, and is weight neutral or mildly weight reducing. It lowers HbA1c by roughly 1 percentage point on average, for example from 8.0 to about 7.0 percent (64 to 53 mmol/mol), though individual responses vary.

Why it is first choice#

When guidelines around the world put metformin first for type 2 diabetes, the reasoning is a stack of practical virtues rather than one dramatic one.

It is effective at lowering glucose. It does not cause weight gain, unlike sulfonylureas and insulin. It almost never causes low blood sugar by itself, which keeps drivers, machine operators, and older people safer. In the landmark UK Prospective Diabetes Study, overweight patients treated with metformin had fewer heart attacks and better survival than those treated with other strategies, a finding that anchored its reputation. It has more than half a century of accumulated safety experience. And it costs very little, which matters both to individuals and to health systems.

Newer drug families, the SGLT2 inhibitors and GLP-1 receptor agonists, have earned their own first-rank positions for people with heart disease, heart failure, kidney disease, or significant weight concerns, and modern guidelines increasingly choose by companion condition rather than by habit. But for a large share of people with newly diagnosed type 2 diabetes, metformin remains the sensible opening move, often alongside one of the newer agents rather than instead of them.

The stomach, and the slow-release option#

The honest downside of metformin is the gut. Up to a quarter of people starting it notice something: nausea, cramping, bloating, loose stools or urgency, sometimes a metallic taste. A small minority, perhaps one in twenty, cannot tolerate it at all.

Three facts take most of the sting out of this.

First, it is usually temporary. The gut adapts over days to weeks. Symptoms that are fading by week two generally disappear by week six.

Second, how you take it matters. With food rather than on an empty stomach, and starting low with slow increases, side effects drop sharply. Doctors deliberately build the dose up gradually for exactly this reason, and there is no prize for rushing.

Third, there is an extended-release version. Standard metformin dumps into the upper gut quickly; the extended-release form is engineered to release slowly as it travels, producing lower local concentrations. For many people whose stomachs objected to the standard form, the slow-release version is comfortable, and it is usually taken once daily. If your gut is the only obstacle, this switch is one of the most useful questions you can ask your doctor.

One further gut note: metformin's effects on bile handling mean that in some long-term users, persistent diarrhea is genuinely drug related even years in. It is worth mentioning rather than assuming it is just you.

The B12 story: a slow leak over years#

This is the side effect almost nobody is warned about, and it unfolds on a timescale of years rather than weeks.

Metformin interferes with the absorption of vitamin B12 in the last part of the small intestine, probably by disrupting a calcium-dependent step where B12, bound to a carrier protein, docks with its receptor. The effect is small per day and cumulative per decade. In long-term users, studies suggest somewhere between one in twenty and one in five develop low B12, with higher doses and longer use raising the odds.

Why care? Because B12 deficiency causes anemia and, more insidiously, nerve damage: numbness, tingling, burning feet, and balance problems. Here is the trap worth underlining: those nerve symptoms look exactly like diabetic neuropathy. A person with diabetes who develops tingling feet will usually be assumed to have diabetic nerve damage, when a simple blood test might reveal a correctable vitamin deficiency instead, or as well.

The practical takeaways are simple. If you have taken metformin for several years, periodic B12 testing is reasonable and increasingly recommended, and it should be insisted on if you have anemia, tingling, numbness, or balance change. Correction is straightforward once found. This is a monitoring issue, not a reason to fear the drug.

The kidney caveat#

Metformin is cleared from the body by the kidneys, essentially unchanged. If kidney function declines, the drug lingers and accumulates, and accumulation raises the risk of the one rare but genuinely dangerous complication: lactic acidosis, a buildup of lactic acid in the blood. It is very uncommon, on the order of a handful of cases per 100,000 users per year, and it almost always occurs when something else has gone badly wrong: severe dehydration, sepsis, major kidney injury.

The safety rules follow directly. Kidney function, reported as eGFR, is checked before starting and at least yearly after. Between an eGFR of 30 and 45 mL/min/1.73m², doctors use reduced doses and closer monitoring; below 30, metformin is generally stopped. It is often paused temporarily around iodinated contrast scans, major surgery, or any illness with serious vomiting, diarrhea, or dehydration, then restarted once the kidneys are demonstrably fine. Many patients are given sick day guidance to this effect; if you have never received it, ask.

The longevity hype, assessed honestly#

Metformin has acquired a second life on podcasts and in longevity clinics as a supposed anti-aging drug, taken by people with no diabetes at all. It is worth separating what is known from what is wished.

The excitement began with laboratory findings, metformin activating the same AMPK energy-sensing pathways implicated in lifespan extension in worms and mice, and with a 2014 observational study suggesting that people with diabetes on metformin lived longer than nondiabetic people not taking it. That finding was striking, widely quoted, and, on reanalysis, substantially undermined: the comparison groups were not tracked equivalently, and people who stopped metformin as their diabetes progressed left the analysis in ways that flattered the drug. Later, more careful studies have not confirmed a survival advantage over comparable people without diabetes.

Trials in people without diabetes have been sobering in another way: in healthy exercisers, metformin appears to blunt some of the beneficial adaptations to exercise, which is an odd property for a longevity pill. The proposed definitive study, the TAME trial, designed to test whether metformin delays age-related disease as a package, has been discussed for the better part of a decade and has struggled to fund and launch; as of now it has not reported.

The fair summary: metformin is a genuinely good diabetes drug whose broader anti-aging benefit in healthy people is unproven, biologically interesting, and currently not a reason to take it. If longevity is the goal, the interventions with actual human evidence remain unglamorous: activity, not smoking, blood pressure control, sleep, and keeping LDL and glucose in healthy ranges.

What I actually see in clinic#

Two patterns repeat endlessly. The first is the person who stopped metformin in week two because of diarrhea and never told anyone. They surface a year later with a much higher HbA1c, convinced the drug did not suit them, when the truth is that nobody walked them through the with-food, go-slow, extended-release ladder. Almost all of them end up tolerating it once we climb that ladder properly.

The second is quieter and bothers me more: the fifteen-year metformin veteran with burning feet, filed under diabetic neuropathy, whose B12 has never once been measured. Sometimes the level comes back frankly low, and treating it changes their symptoms meaningfully. I now treat long-term metformin plus new numbness as a B12 test until proven otherwise, and I would encourage you to ask for the same.

Questions worth asking your doctor#

• What is my current eGFR, and how often will it be rechecked while I am on metformin?
• Should I pause metformin during dehydrating illness or before a contrast scan, and what exactly is my sick day plan?
• After several years of use, when will my vitamin B12 be checked?
• If my stomach objects, can we try the extended-release form before giving up?
• Given my heart, kidney, and weight picture, should metformin be alone or paired with one of the newer classes?

The bottom line#

Metformin lowers blood sugar mainly by telling an overactive liver to stop releasing glucose, with helpful side actions on muscle and gut. It earned first-line status by being effective, safe, weight friendly, and cheap, and it very rarely causes low blood sugar on its own. Its real costs are a gut adjustment period that patience and the extended-release form usually solve, a slow B12 leak worth testing for after years of use, and a firm kidney rule that requires monitoring. The longevity claims remain unproven; the diabetes benefits are not. Taken knowledgeably, it is one of the best value medicines in existence.