# Tirzepatide Versus Semaglutide on Mood

By Nora Bissett, Pricing Editor (https://glpwatchdog.org/authors/nora-bissett) | Published: 2026-08-19 | Updated: 2026-10-27

Canonical page: https://glpwatchdog.org/research/tirzepatide-versus-semaglutide-on-mood

> Across more than 85,000 matched pairs over two years, psychiatric outcomes were the same for both drugs. Against older GLP-1s, semaglutide was associated with about half the hazard of suicidal ideation.

## Key points

- Composite psychiatric outcome: 0.984 (year one), 1.002 (year two). No difference.
- Against older GLP-1s, semaglutide: suicidal ideation HR 0.488, depression 0.811.
- The 'tirzepatide anxiety' figure has a lower bound of 1.001, and the authors say to interpret it cautiously.
- Two cohorts and a randomized trial point one way; spontaneous FDA adverse event reports point the other.

People choosing between these two drugs often want to know whether one is harder on mood. A study of more than 85,000 matched pairs followed for two years gives an unusually clean answer: **no difference**. The composite psychiatric outcome came out at 0.984 in year one and 1.002 in year two.[1]

## The head-to-head

Hazard ratios that land on 0.98 and 1.00 across two years, in a sample that size, are about as close to a true null as observational data produces. Whatever separates these two drugs, psychiatric risk over two years does not appear to be it.[1]

> **There is one wobble, and it is going to be misreported.** Tirzepatide showed a nominally higher anxiety hazard in year two: 1.052, with a 95% confidence interval of 1.001 to 1.106. The lower bound clears no effect by one thousandth. The authors say explicitly that it should be interpreted cautiously given multiple comparisons. If you see “tirzepatide raises anxiety risk” anywhere, that is this number, stripped of its own paper’s caveat. The trial and label record on anxiety is in [GLP-1s and anxiety](https://glpwatchdog.org/research/glp1-and-anxiety).

## The comparison that did show something

Against *earlier-generation* GLP-1 drugs rather than against tirzepatide, semaglutide had a lower hazard for every psychiatric outcome measured in year one.

*Semaglutide compared with other GLP-1 receptor agonists, year one.*

| Outcome | Hazard ratio |
| --- | --- |
| Depression | 0.811 (95% CI 0.770–0.855) |
| Anxiety | 0.915 (95% CI 0.871–0.961) |
| **Suicidal ideation** | **0.488 (95% CI 0.339–0.702)** |
| Composite psychiatric outcome | 0.866 (95% CI 0.832–0.901) |

The suicidal ideation figure matters most, given how much public argument this drug class has attracted on exactly that point. The hazard roughly halves, with an interval nowhere near one.

## Where this sits with everything else

Three independent bodies of evidence now point the same way. A Swedish national cohort found semaglutide associated with lower risk of worsening mental illness and the class associated with less self-harm, which we covered in [worsening mental illness](https://glpwatchdog.org/research/glp1s-and-worsening-mental-illness). This study finds lower psychiatric event rates against older drugs in the class. And a randomized trial in major depressive disorder found no increase in suicidal ideation, covered [here](https://glpwatchdog.org/research/semaglutide-depression-cognition-trial).

Pointing the other way is a spontaneous-reporting signal from FDA adverse event data, discussed in [the FAERS analysis](https://glpwatchdog.org/research/compounded-glp1-adverse-event-reports). That is the weakest design of the four, and the most susceptible to amplification by the news coverage this question generated.

> Two cohorts and a randomized trial point one way; spontaneous FDA adverse event reports point the other.

The same caution applies here as everywhere in this class: semaglutide reaching better outcomes than older drugs may reflect who receives the newest and most effective option rather than the molecule. That confound does not touch the tirzepatide comparison, though. Both drugs are current, both are prescribed to similar people, and the answer there was no difference.

## Frequently asked questions

### Is tirzepatide or semaglutide harder on mood?

Neither, on this evidence. Across more than 85,000 matched pairs followed two years, the composite psychiatric outcome was 0.984 in year one and 1.002 in year two, as close to no difference as observational data produces.

### Doesn't tirzepatide raise anxiety risk?

That comes from a single year-two figure of 1.052 with a confidence interval whose lower bound is 1.001, which clears no effect by one thousandth. The authors themselves say it should be interpreted cautiously given multiple comparisons.

### What about suicidal ideation?

Compared with earlier-generation GLP-1 drugs, semaglutide was associated with roughly half the rate of suicidal ideation in year one, at a hazard ratio of 0.488 with an interval nowhere near one.

### How does this fit with reports of psychiatric harm?

The reports come from spontaneous FDA adverse event data, which is the weakest design available and the most affected by news coverage. Two large cohorts and a randomized trial now point the other way.

### Could this just be who gets prescribed the newer drug?

For the comparison against older drugs, yes. That confound is live, and semaglutide reaches different patients than exenatide or dulaglutide. It does not apply to the tirzepatide comparison, where both drugs are current and go to similar people, and the answer there was no difference.

## References

1. Chen SC, Wang TJ, Chou CK, et al. Neuropsychiatric Outcomes With Tirzepatide, Semaglutide, and Other GLP-1 Receptor Agonists. Diabetes, Obesity and Metabolism. 2026. PMID: 42420795. https://pubmed.ncbi.nlm.nih.gov/42420795/

---

Cite as: GLP Watchdog, Nora Bissett, "Tirzepatide Versus Semaglutide on Mood", published 2026-08-19, updated 2026-10-27. https://glpwatchdog.org/research/tirzepatide-versus-semaglutide-on-mood

Canonical HTML page: https://glpwatchdog.org/research/tirzepatide-versus-semaglutide-on-mood

Medical disclaimer: Nothing here is medical advice. We report on this subject; we do not treat anyone. Whether to begin a medication, supplement, or treatment, whether to adjust it, and whether to stop are questions for a clinician licensed to answer them, who knows your history.
