Semaglutide & Smoking Cessation: What the Evidence Actually Shows

Two randomized trials. One positive pilot, one negative larger study. A 220,000-patient observational signal. Two new RCTs enrolling. Here's exactly what the smoking-cessation evidence base looks like in 2026 — and what it doesn't yet support.

Why GLP-1s ended up in a smoking-cessation conversation at all

The mechanistic case starts in animal models. GLP-1 receptors are expressed in the ventral tegmental area and nucleus accumbens — the same dopamine-driven reward circuits that nicotine, alcohol, cocaine, and amphetamines act on. Preclinical work has shown that GLP-1 receptor agonists blunt nicotine-induced dopamine release in the nucleus accumbens and reduce conditioned place preference for nicotine in rodents.

That mechanism is not unique to nicotine. The same reward-circuit dampening has been linked to reductions in alcohol intake, cue-driven food cravings, and other reinforcing behaviors in animal studies. That is why "GLP-1s for addiction" became a research question across multiple substances — not just smoking.

The exenatide pilot (Yammine 2021): positive, but small

The first randomized human trial of a GLP-1 agonist for smoking cessation was a pilot run at UTHealth Houston and published in Nicotine & Tobacco Research in 2021. Eighty-four treatment-seeking smokers with overweight or prediabetes were randomized 1:1 to extended-release exenatide 2 mg weekly or placebo, on top of nicotine replacement therapy (a 21 mg patch) and brief behavioral counseling.

Exenatide-treated participants also had slightly lower craving scores and prevented the typical post-quit weight gain — a mean weight change of −0.49 lb in the exenatide arm versus +2.96 lb in the placebo arm at end of treatment.

Yammine 2025 follow-up: who responded

A secondary analysis published in 2025 explored which patients in the original trial responded best to exenatide. The signal was strongest in heavier smokers (more than 20 cigarettes per day), and in participants without prediabetes or obesity. The authors framed this as hypothesis-generating, not definitive — but it does narrow where future trials should look.

The dulaglutide trial (SKIP, Lengsfeld 2023): no abstinence benefit

Roughly two years after the exenatide pilot, a Swiss group at University Hospital Basel published the SKIP trial in eClinicalMedicine. SKIP randomized adult smokers with at least moderate nicotine dependence to 12 weeks of dulaglutide 1.5 mg weekly or placebo, on top of standard care — which here meant varenicline 2 mg/day plus behavioral counseling, not NRT.

The primary outcome — biochemically confirmed point-prevalence abstinence at week 12 — showed no benefit from dulaglutide over placebo. Twelve-month follow-up published in 2024 confirmed the lack of long-term abstinence effect. Dulaglutide did, however, prevent post-cessation weight gain at week 12 and modestly improved HbA1c — though the weight effect did not persist at one year.

The Wang/Volkow real-world analysis (Annals of Internal Medicine, 2024)

The largest body of human data on this question is observational. A target-trial emulation published in the Annals of Internal Medicine in 2024 used U.S. electronic health records from December 2017 through March 2023 to compare semaglutide users with patients on seven other antidiabetic drugs (insulin, metformin, DPP-4 inhibitors, SGLT2 inhibitors, sulfonylureas, and other GLP-1 agonists). The cohort included patients with both type 2 diabetes and tobacco use disorder.

Across the seven head-to-head comparisons, patients newly prescribed semaglutide had lower rates of three TUD-related healthcare measures over 12 months: medical encounters with a TUD diagnosis, smoking-cessation medication prescriptions, and (in some comparisons) smoking-cessation counseling. The signal was strongest within the first 30 days of prescription.

What's currently enrolling

Two registered randomized trials are now running:

Note that the UTHealth trial's primary endpoint is weight management around quitting, not abstinence. Abstinence is exploratory. A definitive RCT of GLP-1s for smoking cessation as the primary endpoint, in a representative population, has not yet been completed.

What this means for someone considering GLP-1s today

No FDA-approved GLP-1 medication is indicated for smoking cessation. The current FDA-approved indications for semaglutide (Wegovy, Ozempic, Rybelsus) are weight management and type 2 diabetes — not nicotine dependence. Prescribing or marketing semaglutide for smoking cessation today would be off-label.

What the evidence currently supports:

• FDA-approved smoking-cessation pathways are validated and effective. Per USPSTF, varenicline carries a 6-month abstinence risk ratio of 2.24 versus placebo; bupropion 1.64; nicotine replacement 1.55. Behavioral interventions add another 1.76 RR. These should remain first-line.
• If a person is already taking a GLP-1 for an approved indication (weight or diabetes) and is also trying to quit smoking, the dulaglutide data suggest GLP-1 therapy may help blunt post-cessation weight gain — a meaningful side benefit, since weight gain is one of the top reasons quit attempts fail.
• Starting a GLP-1 specifically to help quit smoking is not currently supported by the trial data. Wait for SURMOUNT-style endpoints in a properly powered cessation trial.

If you start a GLP-1 for an approved indication

The same prescribing safeguards apply as for any GLP-1 use — including the boxed warning for thyroid C-cell tumors, contraindications in personal/family history of medullary thyroid carcinoma or MEN2 syndrome, and the standard pre-procedure precautions.

For licensed prescribers and patients enrolled in care

What we'll be watching for

Three things that would actually move the field:

1. A primary cessation endpoint in NCT06173778. Even though abstinence is currently exploratory in the UTHealth trial, the data will inform power calculations for a definitive trial.
2. Mechanistic neuroimaging. No published human fMRI study has yet examined how GLP-1 medications affect reward circuitry in response to nicotine cues specifically. Inferences are still being drawn from food and alcohol cue-reactivity work.
3. Dose-response signal in semaglutide vs exenatide. Whether the abstinence signal seen with exenatide replicates with the higher-affinity, longer half-life semaglutide will be informative — both pharmacologically and for which agent (if any) deserves a phase 3 program.

Until then: the trial data do not justify starting a GLP-1 to quit smoking. They do justify the trials. We'll cover the readouts as they come.