Many Drinkers Cut Back After Starting Obesity Meds, Study Finds

Nearly half of participants in a weight-loss program who consumed alcohol at baseline began drinking less after adding an obesity drug into their weight-management regimen, a cohort study indicated.

Among the 7,491 participants who reported alcohol use at the start of the study, 45.3% said their use decreased after starting the obesity medication, 52.4% reported no change, and 2.3% reported an increase.

Decreased alcohol use was significantly more likely among women, in those with increasing obesity class, and among the heavier drinkers at baseline, reported researchers led by Lisa R. Miller-Matero, PhD, of Henry Ford Health in Detroit, in JAMA Network Open.

The vast majority (86%) of participants initiated a second-generation GLP-1 receptor agonist, tirzepatide (Mounjaro, Zepbound) or semaglutide (Ozempic, Wegovy), as their obesity medication. Other drugs included first-generation GLP-1 drugs — liraglutide (Victoza, Saxenda), and dulaglutide (Trulicity) — as well as bupropion/naltrexone (Contrave) and metformin.

Of note, dulaglutide and metformin are approved for diabetes but not chronic weight management. And naltrexone on its own is of course FDA-approved for treating alcohol dependence, under the brand name Vivitrol.

“There may be properties of anti-obesity medications that lead to reduced [alcohol] use,” the study authors explained. “For example, naltrexone decreases cravings for alcohol and GLP-1 [receptor agonists] may attenuate the rewarding effects of alcohol, similar to food.”

Even people using metformin reported a decrease in alcohol use, which Miller-Matero’s group said was unexpected.

“This may have occurred because of engagement in a weight-management program, as behavioral strategies may suggest limiting alcohol consumption, given its caloric content and disinhibitory effects on cognitive restraint,” the authors suggested. “Future research would benefit from a randomized trial comparing anti-obesity medications with a placebo-controlled or nonpharmacological weight-management group.”

Coming as less of a surprise was the decrease in alcohol use with GLP-1 receptor agonists, as a few recent studies have suggested they have a potential benefit for alcohol use disorder.

In the current study, Miller-Matero’s group looked at 14,053 adults enrolled in the WeightWatchers (WW) Clinic telehealth medical weight-management program. All had started on an obesity drug between January 2022 and August 2023 and refilled that same medication between October 2023 and November 2023.

Average age of participants was 43 years, 86% were women, and 60% were white. At baseline, average BMI was 35.97. From the time participants started an obesity medication through follow-up (average of 225 days), they experienced a mean 12.68% total weight loss.

Beyond the second-generation GLP-1 agonists tirzepatide and semaglutide, 5% of the participants started a first-generation GLP-1 drug as their obesity medication, 5% started bupropion/naltrexone, and 4% took metformin.

Differences emerged as to which participants were more likely to cut back on drinking after starting medication. For example, men were 26% less likely to report a decrease in drinking compared with women (adjusted odds ratio [aOR] 0.74, 95% CI 0.64-0.85).

Compared with those who were overweight, people with higher classes of obesity were significantly more likely to cut down on alcohol intake after starting an obesity medication:

• Class I (BMI 30 to
• Class II (BMI 35 to
• Class III (BMI ≥40): aOR 1.63 (95% CI 1.36-1.96)

Similarly, heavier drinkers at baseline had the greatest likelihood of cutting back compared with so-called category 1 drinkers (1-3 drinks per week for women and 1-6 drinks per week for men):

• Category 2 (4-6 drinks or 7-14 drinks per week, respectively): aOR 5.97 (95% CI 5.17-6.91)
• Category 3 (≥7 drinks or ≥15 drinks per week, respectively): aOR 19.18 (95% CI 13.25-28.86)

When looking at medications, and using metformin as a reference, only bupropion/naltrexone was associated with a higher likelihood of reduced drinking (aOR 1.42, 95% CI 1.01-1.99), but that finding was no longer significant after accounting for participants’ weight-loss. Other factors evaluated, age and race/ethnicity, didn’t appear to play a role.