Semaglutide’s ability to help people lose weight is legendary, but so are the side effects. The active ingredient in medications Wegovy and Ozempic, semaglutide is renowned for causing severe nausea in some people, sometimes to the extent that they stop treatment. But new research published in the journal Diabetes Care suggests there may be a simple way to combat that effect.

Led by a team in Israel, the researchers found that giving people more time and flexibility to ramp up to the recommended 1 milligram dose of the drug seemed correlated with lower levels of nausea. People who followed the slower dosing schedule also seemed less likely to stop using the drug than those who followed a more typical regimen. What’s more, the slow-and-steady group still lost about as much weight as their peers.

Semaglutide works by mimicking GLP-1, a hormone that plays a key role in regulating our metabolism and hunger. Since U.S. regulators first approved the drug in 2017 to treat diabetes, it has become highly sought after for its appetite-stifling effects and weight-loss benefits—but it has its trade-offs. At least a third of people taking GLP-1s report feeling nauseous, while others also experience constipation and gastrointestinal distress.

As people increase their dose of the drug, the unpleasant GI effects can get worse before they get better. Typically, they fade over time once people reach their recommended dose, but at least some never make it that far and just stop taking the medication.

In the new trial, researchers split 104 people with type 2 diabetes into two groups: One set took increasing amounts of semaglutide on a typical, 8-week dosing program, while the other group ramped up their dose on a slower, more flexible schedule that took 16 weeks. People in the second group were also told to hold off on upping their dose if they experienced any GI symptoms and wait till they felt better. Both groups were tracked for 24 weeks.

Overall, both groups saw similar improvements to their weight and blood sugar control. But flexible users were less likely to report having nausea than their peers (45.1% vs 64.2%), and experienced fewer days of nausea (2.88 vs 6.3 days per month). Perhaps most notably, only 2% of flexible patients stopped taking the drug by the study’s end, compared to 19% of standard users.

“Slower, flexible titration improved adherence and reduced adverse events without compromising efficacy,” the study authors wrote.

While some doctors have anecdotally reported similarly positive experiences in their patients who switched to a slower schedule, the study’s results are limited: The sample size is quite small, for one, and more work in larger and more diverse groups will be needed to understand if different dosing schedules could help more people stay on their medication, without getting the pukes.