After spinal fusion the gut is expected to be sluggish, and the complication surgeons watch for is ileus — bowel function failing to restart. Since these drugs slow gastric emptying and transit, as the gastroparesis evidence describes, the reasonable prediction is that patients taking one would take longer. They took less[1].
Bowel function, defined as passing flatus or having a bowel movement, returned at a mean of 1.35 days in GLP-1 users against 1.90 days in non-users, a difference of about half a day that held in regression (rate ratio 0.69, P = 0.028). Everything else was flat: no differences in length of stay, discharge destination, inpatient complications, wound problems, or patient-reported outcomes at follow-up.
If the effect is real, the explanation is not obvious. One possibility is that the relevant comparison is not the drug against nothing but the drug against the metabolic state it treats: people with poorly controlled diabetes have autonomic gut dysfunction of their own, and a treated patient may simply be in better shape. Another is that chronic exposure produces tolerance to the motility effect, so the slowing seen on starting does not persist.
The nulls are the more useful part of this paper and will get less attention. In a field where perioperative caution around these drugs has grown quickly — the fasting question in what to tell the anesthesiologist and the aspiration findings in the thrombectomy cohort — a study finding no excess complications, no longer stays and no worse recovery scores after major spinal surgery is worth having, even at this size. The pattern of small perioperative studies reaching reassuring conclusions is itself examined in bile duct procedures and every complication halved.