The drugs that quiet appetite do not appear to get people moving. A new review pooled seven studies covering 924 adults with overweight or obesity and found that GLP-1 receptor agonists, the class that includes semaglutide, did not raise objectively measured physical activity. In the one trial that reached statistical significance on its own, the treated group logged about 1,144 fewer steps a day (p = 0.01) than controls.
The review, published July 3 in the International Journal of Obesity ↗, matters because most of what we think we know about exercise on these drugs comes from what people say they do, not from what a device on their wrist or hip recorded. Following PRISMA review standards, the team searched the PubMed and Embase databases from inception to March 10, 2026, and restricted itself to objective measures: step counts, sedentary time, and intensity-specific activity captured by accelerometers, not questionnaires.
What they found is a quiet null with a downward lean. Across all seven studies most differences were not statistically significant. But five of the seven (71.4 percent) showed numerically lower free-living activity in the GLP-1 group, the kind of consistency that is hard to wave away as noise. The main meta-analysis of three randomized trials came out flat (Hedges' g = -0.11, a standardized effect size where zero means no difference, 95 percent CI -0.30 to 0.09). A sensitivity analysis that reweighted the studies nudged into significance for a small reduction (g = -0.24, p = 0.035).
The distinction the paper draws is the useful part. Structured exercise, meaning prescribed sessions, adherence to an exercise program, and minutes of planned workout, held steady between groups. The dip, to the extent there is one, sits in free-living activity: the incidental, non-obligatory movement of a normal day. Walking to the far coffee shop instead of the near one. Taking the stairs. Fidgeting.
Plain version: the shots do not stop people from exercising when they set out to exercise. They may slightly lower the background movement people do without thinking about it.
That fits a mechanism nobody has proven but many suspect. GLP-1 drugs work by cutting energy intake hard. The body, sensing a large calorie deficit, has a well-documented habit of quietly trimming energy output to compensate, and spontaneous activity is the easiest lever to pull. If that is what is happening here, the movement dip is not a side effect of the drug so much as the body doing arithmetic.
There is a real clinical hook in this for anyone building a weight program around these drugs. The pounds come off through appetite, but muscle and long-term weight maintenance lean on movement. If the drug is subtly pulling background activity down, the case for pairing it with a deliberate exercise prescription gets stronger, not weaker, because the structured side is exactly the part that held up here.
The evidence is thin and the authors say so. Seven studies, mixed designs, different accelerometers, different definitions of what counts as a step. The significant sensitivity result depends on how the studies are weighted. This is a signal to test properly, not a settled number. What it is not is the story the marketing implies, in which the weight comes off and an active life follows on its own.
Semaglutide, the most-studied drug in the class, is hosted on peptidemodel ↗ as a GLP-1 receptor ↗ agonist. The receptor pharmacology explains the appetite effect cleanly. It says nothing about whether a person takes the stairs, which is the gap this review is pointing at.