This is an educational summary of adverse events reported in published mazdutide trials. Mazdutide is approved in China and is not FDA-approved; nothing here substitutes for a clinician's assessment or for approved product information in the jurisdictions where it is licensed.
Overview
The dominant side effects of mazdutide in trials are gastrointestinal, they are dose-related, and they are common enough that they should be treated as an expected feature rather than a rare complication. Most were described as mild to moderate, and discontinuation rates stayed low — but "mild to moderate" and "affecting half the participants" are both true at once, and only one of those usually makes it into marketing copy.
What the trials reported
The clearest adverse-event breakdown comes from GLORY-2, the 60-week trial of 9 mg weekly in 461 Chinese adults with obesity (JAMA 2026; PMID 42251595):
| Event | Mazdutide 9 mg | Placebo |
|---|---|---|
| Vomiting | 53.1% | 1.3% |
| Nausea | 46.9% | 3.2% |
| Diarrhea | 39.4% | 6.5% |
| Discontinued for an adverse event | 2.9% | 0% |
At the lower doses used in GLORY-1 — 4 mg and 6 mg weekly over 48 weeks in 610 adults (NEJM 2025; PMID 40421736) — the pattern was the same in kind but milder in degree: gastrointestinal events were again the most frequently reported category, described as mostly mild to moderate, with discontinuation for adverse events in 1.5% (4 mg) and 0.5% (6 mg) against 1.0% on placebo.
Reading those two trials together is the useful exercise. Tripling the dose from 4 mg to 9 mg buys additional weight loss and costs a markedly higher burden of vomiting and nausea. That trade-off, not a single tolerability verdict, is what the data actually describe.
Why the GI effects happen
GLP-1 receptor activation slows gastric emptying and acts on brainstem and hypothalamic circuits that regulate appetite and nausea. The same mechanism that produces early satiety produces queasiness, which is why the two are hard to separate pharmacologically across this entire drug class.
This is also the reason trials escalate the dose stepwise over weeks rather than starting at target. Adverse events cluster during escalation and tend to ease at a stable dose, though that pattern describes a population average, not an individual guarantee.
What the trials do not cover
- Duration. The longest published trial ran 60 weeks. Rare events and cumulative effects surface over years, not months.
- Population. Both pivotal trials enrolled Chinese adults exclusively; adverse-event frequencies observed in one population do not automatically transfer.
- Class-level questions. Regulators evaluating incretin-based agents routinely weigh gallbladder events, pancreatitis signals, and rodent thyroid C-cell findings. Those are class considerations that any new agent on these pathways inherits as questions, whether or not a given trial detected them.
- Unregulated material. Vials sold as "mazdutide" outside a licensed supply chain carry uncertainty about identity, purity, and concentration that has nothing to do with the molecule's own safety profile — and none of the trial data above describes them.