Tirzepatide and semaglutide, the two strongest weight-loss and diabetes drugs on the market, carried the same risk of depression, anxiety, and suicidal thoughts over two years. Neither newer drug looked worse than the other, and both classes looked better than the injections that came before them.
That is the reading from a retrospective study published online July 8 in Diabetes, Obesity and Metabolism ↗. Researchers pulled from the TriNetX Global Federated Network, a pool of more than 192 million patient records. They asked a narrow question that most GLP-1 safety papers skip. Not whether these drugs affect mental health at all, but whether the risk changes as the drug class gets stronger.
What they compared
The team used a new-user design, meaning they only counted patients starting a drug fresh, with a 12-month clean window beforehand to avoid confusing a first prescription with a refill. Adults with type 2 diabetes, obesity, or both who started tirzepatide, semaglutide, or an older GLP-1 receptor agonist between July 2022 and June 2025 went into the pool. Propensity score matching then paired patients with similar age, weight, and clinical history so the comparison was not just healthier people against sicker ones.
Two head-to-head matchups came out of that. The first pitted tirzepatide ↗, the dual GIP and GLP-1 drug sold as Mounjaro and Zepbound, against semaglutide ↗, sold as Ozempic and Wegovy. That comparison had 85,546 matched pairs. The second put semaglutide against the earlier GLP-1 drugs as a group, including liraglutide ↗ and dulaglutide ↗, with 80,115 pairs.
The numbers
Tirzepatide and semaglutide came out even. For the combined measure of depression, anxiety, and suicidal ideation, the hazard ratio sat right on top of no difference in both the first year (0.984) and the second (1.002). A hazard ratio near 1.0 means neither drug carried a higher rate than the other. The one wrinkle was a slightly higher anxiety rate on tirzepatide in year two, a 5 percent bump (hazard ratio 1.052). The authors flag it as fragile. They ran many comparisons, and some will drift above the line by chance.
The older-versus-newer comparison was the more striking one. Against the earlier GLP-1 drugs, first-year semaglutide use was linked to about 19 percent fewer new depression diagnoses (hazard ratio 0.811) and roughly 9 percent less anxiety (0.915). The largest gap was in recorded suicidal ideation, which ran about half as often (0.488). The composite measure landed about 13 percent lower (0.866).
Why it matters
The neuropsychiatric question has followed these drugs for years. European regulators reviewed a possible link between GLP-1 drugs and suicidal thoughts in 2023 and 2024 and found no causal signal. The worry never fully cleared. It still shapes how cautiously doctors prescribe to patients with a psychiatric history. A study that holds the drug class constant and varies only the generation answers a different worry. The fear is that the more potent the appetite suppression, the heavier the mental-health toll. In this data, that trade did not appear. The strongest agent tested, tirzepatide, was no worse than semaglutide.
The honest caveats are the usual ones for records-based work. These are diagnosis codes, not structured psychiatric assessments, so an event only counts if a clinician wrote it down. Channeling matters too. The kind of patient a doctor steers toward the newest drug may differ from the one left on an older injection. Matching cannot fully erase that gap, and it can quietly flatter the newer agent.
This runs alongside our earlier coverage of a study where weight-loss semaglutide users were diagnosed with more depression than people not on the drug at all ↗. The two findings are not in conflict. One asks whether starting a GLP-1 drug raises risk against no drug; this one asks whether the risk moves as the drugs get stronger. On the GLP-1 receptor ↗ shelf, at least on this measure, moving up a generation did not cost patients anything.