GLP-1 Drugs Linked to Lower Alcohol Hospitalization Risk

Key Takeaways

• Among adults with alcohol use disorder plus type 2 diabetes or obesity, initiation of semaglutide or tirzepatide was associated with lower observed risk of alcohol-related hospitalization than active comparator therapies.
• Compared with other diabetes medicines in the ADM trial, semaglutide or tirzepatide initiation was associated with a 26% lower risk of alcohol-related hospital admission. Ccompared with other obesity medicines in the AOM trial, it was associated with a 32% lower risk.
• Compared with alcohol use disorder medications, semaglutide or tirzepatide initiation was associated with a 63% lower risk of alcohol-related hospital admission among adults with type 2 diabetes and a 65% lower risk among adults with obesity.

Adults with alcohol use disorder who initiated semaglutide or tirzepatide had markedly lower observed risk of alcohol-related hospitalization, including a 63% lower risk versus alcohol use disorder medications among those with type 2 diabetes, in a BMJ Open multi-target trial emulation study.

The cohort included 40,703 adults with alcohol use disorder and type 2 diabetes or obesity who initiated semaglutide, tirzepatide, or comparator therapy between Jan. 1, 2018, and Dec. 31, 2024. The investigators emulated four active-comparator trials—ADM, AOM, MAUD-T2D, and MAUD-obesity—and reported alcohol-related hospitalization as the outcome captured from treatment initiation to discontinuation. Across those comparisons, the direction of effect was consistently lower for GLP-1 receptor agonist initiators than for other diabetes medicines, other obesity medicines, and alcohol use disorder medications.

The authors noted several cautions around interpretation, including likely undercapture of both alcohol use disorder and alcohol-related outcomes, along with ascertainment based on diagnosis codes and laboratory testing for alcohol exposure. They also cited potential confounding related to socioeconomic status, clinical stability, alcohol use disorder severity, and healthcare engagement, and reported that residual confounding and high discontinuation made the alcohol use disorder medication comparisons particularly caution-bound.

Their interpretation was that the pattern may suggest a potential role for GLP-1 receptor agonists in the context of alcohol use disorder, but these findings do not establish causation or direct effects on alcohol consumption or craving.

Clinician Questions

How were semaglutide or tirzepatide associated with alcohol-related hospitalization risk in adults with alcohol use disorder and type 2 diabetes?

Among adults with alcohol use disorder and type 2 diabetes, initiation of semaglutide or tirzepatide was associated with a 63% lower risk of alcohol-related hospital admission versus alcohol use disorder medications.

What comparators were used against semaglutide or tirzepatide in the alcohol use disorder hospitalization analysis?

The four emulated trials used active comparators: other diabetes medicines in the ADM trial, other obesity medicines in the AOM trial, and alcohol use disorder medications in the MAUD trials, with named alcohol use disorder comparators including acamprosate, disulfiram, and naltrexone.

What population and study window were reported in the BMJ Open analysis of GLP-1 receptor agonists and alcohol-related hospitalizations?

The reported cohort included 40,703 adults with alcohol use disorder and type 2 diabetes or obesity who started semaglutide, tirzepatide, or a comparator drug between Jan. 1, 2018, and Dec. 31, 2024, and alcohol-related hospitalization was the reported outcome.

What limitations were reported for the GLP-1 receptor agonist analysis in alcohol use disorder?

Reported limitations included possible undercapture of alcohol use disorder and alcohol-related outcomes, outcome definitions based on diagnosis codes and laboratory testing for alcohol exposure, residual confounding that was a particular concern in the alcohol use disorder medication trials, and high treatment discontinuation that increased the potential for bias; the findings do not establish causation.