Yes, and most of the evidence for Ozempic was gathered in people doing exactly that. In SUSTAIN 7, all 1,201 adults were already on metformin when they were randomized, and those on semaglutide 1.0 mg lowered their HbA1c by 1.8 percentage points over 40 weeks [3]. A separate interaction study gave healthy volunteers metformin before and alongside semaglutide 1.0 mg, and found no change in metformin exposure large enough to call for a different dose [1].
The question usually arrives at the pharmacy counter, when the bag holds a bottle of metformin that has been in the kitchen cabinet for years and a new pen that needs a shelf in the refrigerator. The worry is reasonable: two drugs, both aimed at blood sugar, both with a reputation for upsetting the stomach. What follows is what the trials and the label actually recorded about taking them side by side, and where the record thins out.
Most of the trials started with metformin already in the cabinet
For people with type 2 diabetes, semaglutide was mostly tested as the second drug rather than the first. The Ozempic label lists trials of the drug alone and in combination with metformin, with metformin and a sulfonylurea, with metformin and/or a thiazolidinedione, and with basal insulin [6]. Three of the head-to-head trials went further and required metformin as the entry ticket, so every person randomized was taking both drugs from the first week.
SUSTAIN China randomized 868 adults whose HbA1c sat between 7.0% and 10.5% on metformin, and both semaglutide doses beat sitagliptin on HbA1c and body weight after 30 weeks [2]. SUSTAIN 8 randomized 788 people on stable daily metformin to semaglutide 1.0 mg or canagliflozin, and 739 of them completed the year [4]. In SUSTAIN 7 the weight fell by 6.5 kg on semaglutide 1.0 mg against 3.0 kg on the higher dulaglutide dose, again with metformin running underneath the whole time [3]. Whether a weekly shot should replace metformin instead of joining it is a different question, and whether a GLP-1 is better than metformin explains why no trial has settled it.
What the interaction study measured
The direct test of the pairing was a pharmacokinetic study in healthy subjects, funded by the manufacturer, with most of its authors on the company’s staff [1]. Volunteers took metformin 500 mg twice daily for three and a half days, once before semaglutide and again once semaglutide 1.0 mg had reached steady state. The rule set in advance was that the ratio of metformin exposure with and without semaglutide had to fall between 0.80 and 1.25, and it did.
That result is narrower than it sounds, and the scale matters. It shows that semaglutide does not push metformin levels up or down in the blood, which is the worry behind most interaction questions. It does not show how the two drugs feel together over months, it was not run at the 2 mg Ozempic dose, and it enrolled people without diabetes. The Wegovy label repeats the conclusion, listing metformin among drugs whose pharmacokinetics showed no clinically significant difference alongside semaglutide [7].
The stomach is where the two overlap
The practical concern is nausea, loose stools and the first bad week after a dose step, and here the metformin-background trials give a useful number without answering the whole question. In SUSTAIN 7, gastrointestinal disorders were the most common adverse event, reported by 43% of people on semaglutide 0.5 mg and 44% on 1.0 mg [3]. In SUSTAIN 8 the figure was 47%, or 184 of 392 people, and 10% stopped semaglutide because of an adverse event against 5% on canagliflozin [4].
Those percentages describe people taking both drugs, not the extra burden that combining them creates, because no one in those trials took semaglutide without metformin. The dulaglutide arms, on the same metformin background, landed at 33% and 48%, so the gap between arms followed the GLP-1 drug and its dose rather than the metformin every arm shared. How long the queasy stretch tends to last, dose step by dose step, is laid out in how long nausea lasts on a GLP-1, and the mechanism behind it in why GLP-1 drugs cause nausea.
Low blood sugar comes from a different partner
The label’s interaction section warns about hypoglycemia when Ozempic is combined with an insulin secretagogue such as a sulfonylurea, or with insulin, and suggests lowering those doses when Ozempic starts [6]. Metformin is not named in that warning. When Ozempic was taken with a sulfonylurea, severe hypoglycemia occurred in 0.8% on 0.5 mg and 1.2% on 1 mg, which is the combination the label singles out.
SUSTAIN 4 shows the same pattern from the other side. Its 1,089 participants were on metformin, some with a sulfonylurea as well, and severe or glucose-confirmed hypoglycemia was reported by 4% on semaglutide 0.5 mg and 6% on 1.0 mg, against 11% on insulin glargine [5]. For someone whose regimen includes insulin, the dose changes that follow a new GLP-1 are described in how each semaglutide dose step took insulin away.
When there is no diabetes in the picture
Many people asking about the pair do not have diabetes at all, and take metformin for insulin resistance or polycystic ovary syndrome. The Wegovy weight-loss program did include a trial of 807 people with type 2 diabetes, who could be taking up to three oral diabetes drugs with metformin among them [7]. The small trial literature on GLP-1 drugs in polycystic ovary syndrome, where metformin is often the comparison, is gathered in insulin resistance in PCOS on GLP-1 drugs. What stays unresolved is the everyday version of the question: nobody has published how the pair behaves over a year in someone without diabetes, from the first pen to the maintenance dose.