- Design
- Population-based new-user cohort, propensity-score weighted
- Population
- 411,188 US adults with type 2 diabetes on a GLP-1 receptor agonist starting an opioid
- Primary outcome
- 30-day severe constipation, bowel obstruction or gastroparesis
- Effect
- Oxycodone vs hydrocodone RR 1.48 (95% CI 1.30–1.69); vs tramadol RR 1.55 (1.33–1.79)
Investigators used US insurance claims from 2016 to 2025 to follow 411,188 adults with type 2 diabetes who were taking a GLP-1 receptor agonist and started oxycodone, hydrocodone or tramadol. Groups were balanced with propensity-score weights, and the outcome was a composite of severe constipation, bowel obstruction or gastroparesis within 30 days.
The weighted 30-day risk was 0.51% with oxycodone, 0.35% with hydrocodone and 0.33% with tramadol. Oxycodone carried about one and a half times the risk of either comparator, an extra 17 to 18 events per 10,000 patients. Hydrocodone and tramadol were indistinguishable, and gastroparesis did not differ between any of the three.
Both drug classes slow gut transit, so the combination is a predictable worry; what this adds is a ranking among the opioids most often prescribed. It is observational, so residual confounding by indication (more severe pain, more surgery) could explain part of the gap.
In practice: when a patient on a GLP-1 agonist needs a short opioid course, the choice of agent is worth a thought, and a bowel plan matters whichever is chosen. Tramadol brings its own problems, including hypoglycaemia and serotonergic interactions, so this does not make it a default.
- Ask every patient needing an opioid whether they take a GLP-1 receptor agonist, including weight-loss prescriptions from elsewhere.
- Where oxycodone is not specifically needed, consider whether another agent would serve for a short course.
- Prescribe a laxative and a stop-if-no-stool rule alongside any opioid in a GLP-1 agonist user.
- Tell patients that vomiting, distension or no bowel movement for several days needs review, not another dose.
- Absolute risk stays below 1% in a month, so this is a reason for planning, not for withholding analgesia.
Why it matters
Opioid choice is usually made without regard to the GLP-1 agonist on the drug chart, and that may be the wrong default.
Don't overread it
This was a claims-based association; it cannot show that switching opioid prevents obstruction.
The statistics, in plain English
A risk ratio of 1.48 means about half as much risk again, but the starting risk is small: 0.51% versus 0.35%, a difference of about 1.7 per 1,000 patients. The confidence intervals (1.30 to 1.69) sit well clear of 1, so chance is an unlikely explanation, but propensity weighting cannot correct for factors the claims data do not record.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for diabetes & endocrinology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free