The medicine in the diabetes drug Mounjaro helped people with obesity or who are overweight lose at least a quarter of their body weight, or about 60 pounds on average, when combined with intensive diet and exercise, a new study shows.

By comparison, a group of people who also dieted and exercised, but then received dummy shots, lost weight initially but then regained some, researchers reported Sunday in the journal Nature Medicine.

“This study says that if you lose weight before you start the drug, you can then add a lot more weight loss after,” said Dr. Thomas Wadden, a University of Pennsylvania obesity researcher and psychology professor who led the study.

The results, which were also presented Sunday at a medical conference, confirm that the drug made by Eli Lilly & Co. has the potential to be one of the most powerful medical treatments for obesity to date, outside experts said.

“Any way you slice it, it’s a quarter of your total body weight,” said Dr. Caroline Apovian, who treats obesity at Brigham and Women’s Hospital and wasn’t involved in the study.

The injected drug, tirzepatide, was approved in the U.S. in May 2022 to treat diabetes. Sold as Mounjaro, it has been used “off-label” to treat obesity, joining a frenzy of demand for diabetes and weight-loss medications including Ozempic and Wegovy, made by Novo Nordisk.

All the drugs, which carry retail price tags of $900 a month or more, have been in shortage for months.

Tirzepatide targets two hormones that kick in after people eat to regulate appetite and the feeling of fullness communicated between the gut and the brain. Semaglutide, the drug used in Ozempic and Wegovy, targets one of those hormones.

The new study, which was funded by Eli Lilly, enrolled about 800 people who had obesity or were overweight with a weight-related health complication — but not diabetes. On average, study participants weighed about 241 pounds (109.5 kilograms) to start and had a body-mass index — a common measure of obesity — of about 38.

After three months of intensive diet and exercise, more than 200 participants left the trial, either because they failed to lose enough weight or for other reasons. The remaining nearly 600 people were randomized to receive tirzepatide or a placebo via weekly injections for about 16 months. Nearly 500 people completed the study.

Participants in both groups lost about 7% of their body weight, or almost 17 pounds (8 kilograms), during the diet-and-exercise phase. Those who received the drug went on to lose an additional 18.4% of initial body weight, or about 44 pounds (20 kilograms) more, on average. Those who received the dummy shots regained about 2.5% of their initial weight, or 6 pounds (2.7 kilograms).

Overall, about 88% of those taking tirzepatide lost 5% or more of their body weight during the trial, compared with almost 17% of those taking placebo. Nearly 29% of those taking the drug lost at least a quarter of their body weight, compared with just over 1% of those taking placebo.

That’s higher than the results for semaglutide and similar to the results seen with bariatric surgery, said Apovian.

“We’re doing a medical gastric bypass,” she said.

Side effects including nausea, diarrhea and constipation were reported more frequently in people taking the drug than those taking the placebo. They were mostly mild to moderate and occurred primarily as the dose of the drug was escalated, the study found. More than 10% of those taking the drug discontinued the study because of side effects, compared with about 2% of those on placebo.

Lilly is expected to publish the results soon of another study that the firm says shows similar high rates of weight loss. The U.S. Food and Drug Administration has granted the company a fast-track review of the drug to treat obesity, which Eli Lilly may sell under a different brand name. A decision is expected by the end of the year.

The Associated Press Health and Science Department receives support from the Howard Hughes Medical Institute’s Science and Educational Media Group. The AP is solely responsible for all content.