In a propensity-matched analysis of 15,191 ICU patients, prior GLP-1 agonist use was not associated with worse in-hospital mortality, hospital length of stay, or ICU length of stay compared to matched controls.
No difference in ICU outcomesPrior GLP-1 agonist use showed mortality OR 1.05 (p=0.88) with no significant differences in hospital or ICU length of stay
What the researchers found
Prior GLP-1 agonist use showed no association with adverse ICU outcomes: in-hospital mortality 5.1% vs 4.9% (OR 1.05, 95% CI 0.58-1.91), hospital LOS 13.7 vs 13.4 days, ICU LOS 5.9 vs 5.4 days. All differences were non-significant.
Why it matters
As millions take GLP-1 drugs and concerns grow about muscle loss side effects, this is the first study examining whether GLP-1 users fare worse during critical illness. The reassuring null finding suggests clinicians need not be concerned about GLP-1-related risks in ICU settings.
The numbers in context
15,191 eligible ICU patients. 468 (3.1%) had GLP-1 prescriptions within 12 months before admission. 452 propensity-matched pairs analyzed. Similar in-hospital mortality, hospital LOS, and ICU LOS between groups.
How the study worked
Retrospective propensity-matched cohort study using Stanford Health Care data (Jan 2015-Jul 2024). 15,191 eligible ICU patients, 468 with prior GLP-1 use, 452 matched pairs created using high-dimensional propensity score matching with lasso regression.
Who was studied
ICU-admitted adults at Stanford Health Care, comparing prior GLP-1 users to non-users
What this study cannot tell us
Retrospective observational design cannot establish causation. Single-center (Stanford) may limit generalizability. Could not assess GLP-1 dose, duration, or adherence. 3.1% exposure rate may limit power for rare outcomes. No data on functional recovery or ICU-acquired weakness specifically.
How to read the evidence
Large retrospective propensity-matched cohort study with rigorous statistical methodology (high-dimensional propensity scoring, lasso regression). Strong for observational evidence but limited by single-center design.
When this study was published
Published in 2025; the first study to examine GLP-1 agonist use and critical care outcomes.
The bigger picture
This addresses a critical safety question: as GLP-1 prescriptions surge, patients on these drugs will inevitably face hospitalization and critical illness. Finding no adverse ICU outcomes helps allay fears that GLP-1-related body composition changes compromise resilience during severe illness.
Questions still open
- Would longer-duration GLP-1 use (>12 months) with greater lean mass loss show different ICU outcomes?
- Does prior GLP-1 use affect ICU-acquired weakness and functional recovery specifically?
- Should GLP-1 drugs be continued, held, or discontinued during critical illness?
Common questions
Is it safe to be on GLP-1 drugs if you end up in the ICU?
Why were there concerns about GLP-1 drugs and critical illness?
Read the original research
Prior GLP-1 agonist use is not associated with adverse inpatient critical care outcomes: A propensity-matched analysis.
Journal of hospital medicine
Citation
Park, Albert K; Hom, Jason; Lorenzo, Javier; Rao, Vidya; Hui, Gavin; Vickers, Matthew; Ahuja, Neera. (2025). Prior GLP-1 agonist use is not associated with adverse inpatient critical care outcomes: A propensity-matched analysis.. Journal of hospital medicine. https://doi.org/10.1002/jhm.70228