GLP-1s and Depression, Anxiety, and Mood Changes: What the Research Actually Shows

"Zepbound and depression" and "can tirzepatide cause anxiety" are both real, common searches — and both deserve a more careful answer than either "these drugs cause depression" or "there's nothing to it." Regulators formally investigated a specific, serious signal; separately, researchers have looked at lower-severity mood effects some patients report. Those are two different questions with two different answers, and this guide keeps them apart.

What the FDA specifically investigated

In July 2023, after receiving postmarketing reports of suicidal ideation and behavior (SI/B) in patients taking GLP-1 receptor agonists, the FDA opened a formal safety review. A preliminary update published in January 2024 reported that the agency's initial look at clinical trial and adverse-event data "has not found evidence that use of these medicines causes suicidal thoughts or actions," while noting the small number of events made a small risk impossible to rule out entirely, and that review continued¹.

That review has since concluded. In a January 2026 Drug Safety Communication, the FDA reported a comprehensive meta-analysis across 91 placebo-controlled GLP-1 RA trials covering 107,910 patients (60,338 on a GLP-1 RA, 47,572 on placebo), plus a separate retrospective cohort study using FDA Sentinel System claims data on 2,243,138 new users comparing GLP-1 RAs against SGLT2 inhibitors in type 2 diabetes patients. Neither analysis found an increased risk of SI/B. The agency's stated conclusion: "the totality of these studies does not support a causal relationship between the use of GLP-1 RAs and the occurrence of SI/B"¹. Based on that finding, the FDA is now requesting that the SI/B warning be removed from the labeling of Saxenda, Wegovy, and Zepbound — the three GLP-1 RAs whose labels currently carry it¹. That is a materially stronger conclusion than the 2024 preliminary statement, and it is the current, most up-to-date FDA position on this specific question.

The pharmacovigilance signal that made this worth investigating

The FDA's conclusion doesn't mean the underlying signal in adverse-event databases was imaginary — it means the agency's larger, controlled analyses didn't confirm it as causal. A 2025 study using VigiBase, the WHO's global pharmacovigilance database, analyzed over 2 million adverse-event reports and found a statistically elevated reporting rate specifically for semaglutide: an adjusted reporting odds ratio of 1.70 for depressed mood disorders and 1.45 for suicidality, alongside 1.26 for anxiety — signals not seen at the same strength for dulaglutide or liraglutide². Critically, the same study's causal forest model — a method built to estimate actual treatment effect rather than raw reporting frequency — put semaglutide's average effect on depression/suicidality reporting at 0.0046, with the elevated signal concentrated in specific regions and demographic subgroups rather than uniform across patients². Pharmacovigilance reports like these are disproportionality signals, not proof of causation — they can't control for who gets reported, and obesity itself carries an elevated baseline rate of depression independent of any medication. That gap between "a real reporting signal exists" and "a causal effect is confirmed" is exactly what the FDA's larger meta-analysis and cohort study were built to resolve.

A 2026 systematic review pulling together 43 observational studies on this same question found the picture is genuinely mixed by study design: some case reports described a temporal relationship between starting a GLP-1 RA and suicidal ideation, while cross-sectional and case-control studies found a negative association with suicide attempts, and pharmacovigilance and cohort studies were inconsistent — some showing a positive association with semaglutide or liraglutide, others showing the opposite. The review's conclusion: "evidence does not support GLP-1 RAs as a causative agent for suicidality," while recommending clinicians routinely assess mental health status before and during treatment — and explicitly noting a history of mental health conditions should not by itself rule someone out from using a GLP-1³.

The lower-severity mood effects, and what's driving them

Separate from the suicidality question, some patients report milder mood changes — feeling flatter, more irritable, or less motivated — while losing weight quickly on a GLP-1. A 2025 study combining FAERS adverse-event data with Mendelian randomization (a genetic method used to test whether an association is likely causal) offers the most direct evidence on mechanism. It found only a mild suicide-related signal, and only within the obesity-treatment subgroup specifically⁴. More notably, its genetic analysis found GLP-1 receptor activation was associated with *reduced*, not increased, risk of anxiety, depression, emotional lability, bipolar disorder, and suicide⁴ — the opposite direction from the "these drugs cause depression" framing. The same study's mediation analysis found that weight loss itself statistically accounts for a meaningful share of the drug's effect on mood: about 18% of the effect on depression and about 8% of the effect on emotional lability was explained by weight loss as a pathway, rather than a direct pharmacological action on mood⁴. That fits the broader pattern our tirzepatide side effects and semaglutide side effects guides document elsewhere: rapid GI discomfort, appetite suppression, and a fast-changing relationship with food and eating are disruptive enough on their own to plausibly show up as irritability or low mood in some patients, independent of any direct drug effect on the brain.

What this means before you buy — or if you're already on one

The current, most complete evidence — an FDA meta-analysis of over 100,000 trial patients, a 2.2-million-person Sentinel cohort study, and genetic evidence pointing the opposite direction from causation — does not support GLP-1 RAs as a cause of suicidal thoughts or actions, which is why the FDA is moving to remove that warning from the label. A narrower, semaglutide-specific signal in pharmacovigilance databases is real and is the reason mental health screening before and during treatment remains good practice regardless of what the label says. If you notice new or worsening depression, anxiety, or unusual changes in mood after starting a GLP-1, tell your prescriber — don't assume it's unrelated, and don't stop the medication abruptly without talking to them first, since the FDA specifically warns that stopping can worsen the condition being treated. If you or someone you know is having thoughts of suicide, call or text 988, or visit [988lifeline.org](https://988lifeline.org/) — free, confidential support is available 24/7. A provider that actually screens mental health at intake, rather than treating it as a checkbox, is exactly the kind of clinical oversight scored in our methodology. The full source list behind every figure above is in our research library. Not medical advice — see our medical disclaimer and talk to a clinician or mental health professional about your own history.