Twelve adults who smoke cigarettes got nine weeks of semaglutide. Twelve got placebo. The trial's two primary endpoints did not move. Cigarette craving and body weight did.
The trial is a parallel-arm phase 2a randomized clinical trial run at one academic medical center between October 2022 and April 2024, published May 1 in JAMA Network Open ↗. Forty-five adults were enrolled, twenty-four randomized one to one to semaglutide ↗ or placebo, twenty-one completed the primary outcome assessment. Participants were non-treatment-seeking, meaning they were not trying to quit smoking when they enrolled. They smoked at least five cigarettes a day, mean fifteen point four, and the sample was eighty-three percent female with a mean body mass index of thirty-three point five.
Semaglutide was dosed subcutaneously on a standard ramp. A quarter milligram for four weeks, half a milligram for four weeks, then one milligram for the last week. Total exposure nine weeks. That is shorter than the cessation timeline most quit-smoking trials use, and well short of the dose escalation a patient on Wegovy or Ozempic would have completed before plateauing.
The two co-primary outcomes both came from a human laboratory session run before treatment and again after. The first measured smoking resistance, which is a structured test of how long a participant can hold off on smoking when cigarettes are placed in front of them. The second measured number of cigarettes consumed during the session. Both were modeled as treatment-by-time interactions.
Neither moved.
Smoking resistance changed by 0.16 standard units in the semaglutide arm relative to placebo (95% CI -0.07 to 0.40, p=0.16). Cigarette self-administration changed by -0.08 (95% CI -0.25 to 0.08, p=0.30). The confidence intervals straddle zero in both cases. With twelve participants per arm, the trial was underpowered for anything but a large effect, and a large effect is not there.
A prespecified supplementary analysis using change scores rather than treatment-by-time models did show a significantly greater reduction in lab smoking in the semaglutide arm (β=-0.69, 95% CI -1.26 to -0.13, p=0.02, effect size d=0.67). The authors flag this as a directional signal worth chasing, not as a positive result on the primary endpoint. The phase 2a design is a feasibility test, not a registration trial.
Two of the secondary outcomes did clear conventional significance thresholds. Cigarette craving fell more in the semaglutide arm over the nine weeks (β=-0.11, 95% CI -0.20 to -0.03, p=0.01). Body weight fell about four hundredths of a kilogram per week in the semaglutide arm relative to placebo (β=-0.04, 95% CI -0.05 to -0.03, p<0.001), which over nine weeks is consistent with the modest weight loss expected at this dose ramp. The trial also reports exploratory effect sizes suggesting an impact on nicotine withdrawal, though the small sample makes those effect sizes hard to interpret.
This is the first published randomized trial of semaglutide for cigarette smoking. The preclinical case for testing GLP-1 receptor agonists in smoking has been building on rat and mouse work showing reductions in nicotine self-administration. Case reports and retrospective database analyses from late 2024 and 2025 suggested some humans on semaglutide for diabetes or weight cut back on cigarettes. Off-label prescriptions for cessation were already happening before this trial reported.
The trial answers the narrowest version of the question. With twelve people per arm and a non-treatment-seeking sample, semaglutide did not reduce the number of cigarettes they smoked or how well they could resist smoking when cigarettes were placed in front of them. It did reduce how much they wanted them, by a modest amount.
The bigger question, whether semaglutide helps treatment-seeking smokers actually quit when paired with established cessation support, and how that interacts with the well-documented postcessation weight gain that drives so many relapses, is outside this trial's scope. The authors call explicitly for larger trials. The trial is registered as NCT05530577.
For the semaglutide card ↗, this is the first prospective controlled signal in the addiction column. The cardiometabolic indication anchored the molecule. The reward-system indications, alcohol use disorder, opioid use disorder, cigarette smoking, are now accumulating early-phase data, without a positive primary endpoint on any of them yet.