Metformin vs. Ozempic: a daily tablet or a weekly GLP-1 injection

They belong to different classes and work in different places. Metformin is a biguanide. Its label says it decreases glucose production by the liver, reduces absorption of glucose from the gut and improves insulin sensitivity, without increasing insulin secretion. It is a tablet, usually started at 500 mg twice a day or 850 mg once a day with meals, and raised gradually to a maximum of 2,550 mg a day. Ozempic is semaglutide, a GLP-1 receptor agonist that increases insulin release when glucose is high, lowers glucagon and slows stomach emptying. It is injected once a week, starting at 0.25 mg for four weeks and rising to 0.5, 1 or 2 mg. Both are approved to improve blood sugar control in adults with type 2 diabetes. Ozempic's label adds two outcome indications: reducing major cardiovascular events in people with established heart disease, and reducing kidney decline in people with chronic kidney disease. Neither is approved for weight loss.

No trial has randomized people to metformin alone versus Ozempic alone, so the numbers below come from separate placebo-controlled studies and cannot be compared directly. In Ozempic's 30-week monotherapy trial, reported in its label, HbA1c fell by 1.4 points on 0.5 mg and 1.6 points on 1 mg from a baseline near 8.1%, against 0.1 on placebo; body weight fell 3.8 kg and 4.7 kg, against 1.2 kg. In metformin's 29-week placebo-controlled trial, reported in its label, HbA1c fell 1.4 points from 8.4% while it rose 0.4 on placebo. The closest thing to a direct test is Ozempic's 56-week trial against sitagliptin, in which 94% of participants were already on metformin: adding Ozempic lowered HbA1c by 1.3 to 1.5 points, against 0.7 with sitagliptin, and 66% to 73% reached an HbA1c below 7%, against 40%. For weight, the Diabetes Prevention Program found metformin reduced weight by about 2.1% on average over its blinded phase, against almost none on placebo, and the loss persisted over ten years in those who kept taking it. That is modest next to Ozempic's effect in its trials, though the DPP enrolled people with prediabetes rather than diabetes, so the populations differ as well as the drugs.

Both mainly affect the gut. In metformin's placebo-controlled trial, diarrhea affected 53% against 12% on placebo, and nausea or vomiting 26% against 8%; diarrhea led 6% to stop. In Ozempic's pooled placebo-controlled trials, nausea affected 15.8% on 0.5 mg and 20.3% on 1 mg against 6.1%, with vomiting, diarrhea, abdominal pain and constipation also more common, mostly during dose increases. The serious risks differ. Metformin carries a boxed warning for lactic acidosis, a potentially fatal build-up of acid whose risk factors include poor kidney function, heavy alcohol intake and liver impairment, and it can lower vitamin B12, which fell below normal in about 7% of trial patients. Ozempic carries a boxed warning about thyroid C-cell tumors seen in rodents, and warnings for pancreatitis, gallbladder disease, worsening diabetic eye disease and kidney injury from dehydration. Both labels warn that the risk of low blood sugar rises when either is combined with insulin or a sulfonylurea.

Metformin suits adults with type 2 diabetes who have adequate kidney function and would rather take a tablet than inject. It is a generic that has been on the US market since 1995, per its label. It does not suit anyone with an eGFR below 30, its label advises against starting it at an eGFR of 30 to 45, and people with heavy alcohol use, liver disease or a history of lactic acidosis need a careful conversation first. Ozempic suits people with type 2 diabetes who have established heart disease or chronic kidney disease, because its label carries outcome indications for both, and people whose HbA1c is not at target on metformin alone. It does not suit anyone with a personal or family history of medullary thyroid carcinoma or MEN 2, or severe gastroparesis, and it needs care with a history of pancreatitis. For many people the answer is both, since Ozempic was largely studied on top of metformin. Metformin's label says it is not for type 1 diabetes, Ozempic is approved only for type 2, and neither is approved for weight loss.