Glucagon-like peptide-1 analogues reduce alcohol intake

Alcohol use disorder (AUD) is a chronic relapsing condition that poses global health challenges. In 2019, alcohol use accounted for approximately 2.6 million deaths worldwide, representing 4.7% of all global deaths.¹ Addiction pathogenesis involves neurobiological and psychosocial factors altering the brain's reward system within the mesolimbic dopamine pathway. Alcohol increases dopamine release, reinforcing substance-seeking behaviour and transitioning from voluntary use to compulsive use with impaired executive function.²

Treatments for AUD include behavioural, motivational, and pharmacological interventions. Medications approved for AUD include disulfiram, naltrexone, and acamprosate. However, the relapse rate is 70% within the first year of treatment.³

In preclinical animal studies, glucagon-like peptide-1 receptor agonists (GLP-1 RAs), such as liraglutide and semaglutide, reduced alcohol intake and attenuated alcohol-induced behaviours.⁴ These agents modulate the mesolimbic reward circuitry by decreasing dopamine release in response to alcohol thereby reducing the motivation to consume alcohol. They also affect stress regulation and cognitive functions which are critical factors influencing relapse propensity.⁵ We aimed to evaluate changes in alcohol use when patients were treated for obesity with GLP-1 RAs in a real-world setting.

As part of routine clinical care, data were collected prospectively over 15 months from January 2023 to March 2024. Electronic health records with manual validation were used to extract demographics, weight and body mass index (BMI), GLP-1 RAs therapy (type, dose, frequency, and duration), and self-reported alcohol intake at baseline and follow-up visits 3–6 months after therapy initiation. We included 262 adult patients with a BMI ≥27 kg/m² who were initiated on liraglutide or semaglutide. Patients were excluded if insufficient data were obtained. Alcohol intake was categorized into non-drinkers, rare drinkers, and regular drinkers. The number of weekly units was recorded for those able to quantify their intake.

The primary outcome was the change in alcohol intake from baseline to follow-up. The secondary outcome was the change in body weight. Continuous variables were expressed as mean ± standard error of the mean (SEM); categorical variables were presented as counts and percentages. The Mann–Whitney U test assessed changes in alcohol intake due to non-normal data distribution. A p-value <0.05 was considered statistically significant. Correlation analyses evaluated the relationship between changes in alcohol intake and weight loss using Pearson correlation coefficients. The study was approved by St Vincent's Healthcare Group (reference 2024/4161); all patients provided information voluntarily as part of routine clinical care.

Of the 262 patients (79% female, mean age 46 years) treated with GLP-1 RAs, 31 (11.8%) were non-drinkers before the intervention, 52 (19.8%) consumed alcohol rarely, and 179 patients (68.3%) consumed alcohol regularly. Of the patients who drank alcohol regularly before the intervention (n = 179), 143 patients (54.6%) had quantifiable alcohol intake recorded, while 36 patients (13.7%) had non-quantifiable alcohol intake recorded.

After initiating GLP-1 RA in 262 patients, 188 (71.8%) had at least two visits during the study, with a mean interval to the second visit of 112 days. Seventy-four patients (28.2%) were lost to follow-up (Figure 1).

Post-intervention, alcohol intake (quantifiable and non-quantifiable) was documented for 117 patients (44.7%). Thirty-one patients (11.8%) had non-quantifiable intake, while 86 patients (32.8%) had quantifiable alcohol intake. No patient reported an increase in alcohol intake. Alcohol intake reduced from a pre-intervention mean ± SEM of 11.8 ± 1.0 units/week (female: n = 67, 10.3 ± 1.0; male: n = 19, 13.4 ± 2.5) to a post-intervention mean of 4.3 ± 0.5 units/week (female: 3.9 ± 0.4; male: 5.8 ± 1.4), (p < 0.001). For patients consuming ≥11 units/week (high consumers), intake decreased from 23.2 ± 1.8 units/week (female: n = 21, 20 ± 1.8; male: n = 9, 22.1 ± 3.1) to 7.8 ± 0.9 units/week (female: 7.0 ± 0.8; male: 9.7 ± 2.3).

For those pre-intervention consuming <11 units/week (low consumers), intake decreased from 5.5 ± 0.3 units/week (female: n = 46, 5.9 ± 0.5; male n = 10, 5.5 ± 1) to 2.5 ± 0.3 units/week (female: 2.5 ± 0.4, male: 2.4 ± 0.6) (Figure 2). Mann–Whitney U tests confirmed the reduction for both high consumers (U = 78, Z = 6.7, p < 0.001) and low consumers (U = 1118, Z = 5.76, p < 0.001). Mann–Whitney U test confirmed that the difference between males and females was not significant in percentage reduction (U = 545, Z = 0.9, p = 0.35) or in units/week (U = 526.5, Z = −1.13, p = 0.25).

At the second visit, weight loss was calculated for 181 participants (69.1%), (female: n = 148, 92.6 ± 1.2 kg, male: n = 33, 114.4 ± 3.3 kg) with a mean loss of 7.7 ± 0.3 kg (female: 7.5 ± 0.3 kg; male: 8.8 ± 0.7 kg) after approximately 4 months. Mann–Whitney U test confirmed that the difference between males and females was not significant in percentage reduction (U = 2247, Z = 0.7, p = 0.47) A overall weak positive correlation was observed between alcohol reduction and weight loss in patients with data for both alcohol intake and weight at both visits (r = 0.24, n = 72; male: n = 16, r = −0.16; female: n = 56, r = 0.29).

There were 51 patients with weight data that reported not reducing their alcohol intake, of these there was a reduction of 7.6 ± 0.6 kg (7.4 ± 0.6 kg females, 9.4 ± 1.7 kg males).

Our results demonstrate a significant reduction in alcohol intake (mean decrease of 7.1 units/week) among patients treated with GLP-1 RAs, with a weak positive correlation between alcohol reduction and weight loss. These results align with existing evidence that GLP-1 RAs may influence alcohol use. GLP-1 receptors are expressed in the brain and are associated with reward and addiction. Activation of these receptors has been linked to decreased alcohol intake and attenuated alcohol-induced dopamine release by influencing dopamine signalling in reward circuits.^{6, 7} Clinical trials have indicated that patients treated with GLP-1 RAs experience a reduction in alcohol cravings and intake.⁸ The mechanisms may involve the modulation of neurochemical pathways associated with reward and satiety.

The reduction in alcohol intake for the high consumers was 68%. This compared with the 61% reduction reported in clinical trials for nalmefene—a medication approved by the European Medicines Agency for reduction of alcohol consumption.

The significant weight loss observed in our cohort is consistent with other clinical trials involving GLP-1 RAs. The weak positive correlation suggests that decreased alcohol intake may partly contribute to weight loss, given alcohol's high caloric content.⁹ This relationship underscores the multifaceted benefits of GLP-1 RA therapy in managing metabolic health.^{10, 11}

Our findings may have important clinical implications. The ability of GLP-1 RAs to reduce both alcohol intake and body weight suggests a potential therapeutic role for these agents in patients with co-occurring obesity and AUD. Clinical trials have begun to explore this avenue, with preliminary results indicating that GLP-1 RAs may reduce alcohol cravings and intake in humans.⁴ The dual efficacy of GLP-1 RAs could lead to integrated treatment strategies that address both metabolic and addictive disorders, improving patient outcomes.¹²

However, several limitations must be acknowledged. The total number of patients in the study is relatively small and limits the importance of the results. The reliance on self-reported alcohol intake may have introduced recall bias and underreporting. Nearly one-third of patients were lost to follow-up. This may introduce selection bias and limit generalizability. The absence of a control group and the observational nature of the study restrict our ability to establish causality between GLP-1 RA therapy and reductions in alcohol intake. The short follow-up duration also prevents the assessment of long-term effects. Furthermore, potential confounding variables include the effect of physician counselling regarding alcohol intake and decisions patients may have made about changes in their lifestyle.

Despite these limitations, our prospective data collection in a real-world setting enhances the relevance of our findings. Observer bias was minimized because data were collected as part of routine care without preconceived biases.

The use of GLP-1 RAs in patients with obesity was associated with a significant reduction in alcohol intake and a weak positive correlation with weight loss. These findings suggest a potential therapeutic role for GLP-1 RAs in managing co-occurring obesity and alcohol use. Future randomized controlled trials with larger and more diverse populations are necessary to validate these results, explore underlying mechanisms, and assess long-term outcomes for integrated treatment strategies.

The authors report that no funding was received for this work.

F.A. reports no competing interests. M.O'.F provides obesity clinical care in the Medication Weight Loss Clinic and is the owner of this clinic. M.O'.F has served on advisory boards for Novo Nordisk and Johnson & Johnson, both positions unremunerated. C.l.R reports grants from the EU Innovative Medicine Initiative, Irish Research Council, Science Foundation Ireland, Anabio, and the Health Research Board. He serves on advisory boards and speakers panels of Novo Nordisk, Roche, Herbalife, GI Dynamics, Eli Lilly, Johnson & Johnson, Gila, Irish Life Health, Boehringer Ingelheim, Currax, Zealand Pharma, Keyron, AstraZeneca, Arrowhead Pharma, Amgen, and Rhythm Pharma. C.l.R is the Chair of the Irish Society for Nutrition and Metabolism. C.l.R provides obesity clinical care in the My Best Weight clinic and Beyond BMI clinic and is a shareholder.