Association of GLP-1 receptor agonists with alcohol use disorder-related and substance use disorder-related hospital admissions during treatment and after discontinuation: a Swedish register-based within-individual observational study.
Level 3 - non-randomized controlled study
Non-randomized, within-individual register-based cohort study.
PubMed 42456704 · doi:10.1016/S2215-0366(26)00172-0
What was done
A Swedish register-based, within-individual observational cohort study evaluated 167,026 individuals with type 2 diabetes in Stockholm County (diagnosed 2005–2024; active follow-up 2015–2024). Fixed-effects Poisson regression models compared rates of alcohol use disorder (AUD)-related and substance use disorder (SUD)-related hospitalisations during periods of GLP-1 receptor agonist exposure, two post-discontinuation windows (days 1–182 and 183–364), and unexposed periods within the same individuals. DPP-4 inhibitors were analysed as a comparator.
What was found
Among 167,026 individuals, 41,109 received GLP-1 receptor agonists and 23,666 received DPP-4 inhibitors. GLP-1 receptor agonist exposure was associated with lower rates of AUD-related hospitalisations (rate ratio [RR] 0.55, 95% CI 0.43–0.70) and SUD-related hospitalisations (RR 0.61, 95% CI 0.50–0.74) compared with unexposed periods in the same individuals. For AUD, the reduction persisted during days 1–182 post-discontinuation (RR 0.70, 95% CI 0.50–0.98) but was no longer evident during days 183–364 (RR 1.07, 95% CI 0.71–1.63). For SUD, reductions were not statistically significant after discontinuation (days 1–182: RR 0.79, 95% CI 0.61–1.02; days 183–364: RR 0.89, 95% CI 0.63–1.26). DPP-4 inhibitors showed no consistent rate reductions.
Why it matters
This study provides real-world evidence that GLP-1 receptor agonist use is associated with substantial reductions in severe AUD and SUD hospitalisations, but indicates that protective effects on AUD diminish within 6 to 12 months after stopping treatment.
Limits
The observational design cannot establish causality. Outcomes were limited to inpatient hospital admissions, capturing only severe presentations and missing outpatient or subclinical substance use. The sample was restricted to individuals with type 2 diabetes in Stockholm County, limiting generalizability to non-diabetic populations. Ethnicity data were unavailable, and time-varying confounding could still influence treatment initiation or discontinuation.
Cited by
- supports A study showed that in individuals with substance use disorders, GLP-1 receptor agonist use resulted in a ~39% improvement that persisted even after discontinuing the drug.