The Flattened Feeling: Mood and Emotional Blunting on a GLP-1

The word women reach for is "flat"

There are two entirely different conversations about GLP-1s and mood, and they keep getting mistaken for each other.

The first is about catastrophe — depression, suicidal thinking, a regulatory investigation. That conversation has now largely resolved, and it resolved in the drug's favour, in a way almost nobody has reported.

The second is much quieter and has no evidence base at all. It is the woman who says she is fine, nothing is wrong, and also that she has not really wanted anything in about four months. Online it has acquired a name — "Ozempic personality" — and the clinical word underneath it is anhedonia: a reduced ability to take pleasure in things that used to give it.

The warning that came off the label in February

Start with the primary document, because it changed and the change went unremarked.

Both semaglutide and tirzepatide once carried a Warnings and Precautions section headed "Suicidal Behavior and Ideation", added while European and American regulators investigated a signal from spontaneous reports. Download the current prescribing information today and it is gone. The Wegovy label's Recent Major Changes table records "Suicidal Behavior and Ideation (5.10) … (Removed) 02/2026", and the word appears nowhere else in the document¹. The Zepbound label carries the identical entry, dated the same month².

The evidence behind that removal is unusually good for a safety question. A retrospective cohort study in Nature Medicine matched 240,618 patients with overweight or obesity prescribed semaglutide against patients on non-GLP-1 anti-obesity medications and found semaglutide associated with a lower risk of both first-time suicidal ideation, at a hazard ratio of 0.27, and recurrent ideation, at 0.44 with a confidence interval of 0.32 to 0.60 — replicated in 1.59 million patients with type 2 diabetes³. A 2026 meta-analysis of 86 randomised trials and 133,378 participants found no association with psychiatric disorders overall (relative risk 0.96), nor with depression, anxiety, suicide, sleep disorder or addictive disorder taken individually⁴.

What the trials measured, and what they could not

The most directly relevant analysis pooled the STEP obesity trials and looked specifically at psychiatric safety.

Participants in the STEP 1, 2 and 3 trials — 3,377 adults, 69.6% of them women, mean age 49 — completed the PHQ-9 depression questionnaire and the Columbia-Suicide Severity Rating Scale throughout. Baseline scores were 2.0 and 1.8 out of 27, which is to say no or minimal depressive symptoms. At week 68 the semaglutide group sat slightly lower than placebo, a difference of 0.56 points, and were less likely to shift into a more severe PHQ-9 category. One per cent or fewer reported suicidal ideation or behaviour, with no difference between groups⁵.

Now read the title of that paper, because it contains the limitation: in people without known major psychopathology. Anyone with significant psychiatric history was screened out of these trials before they began. And a nine-item depression scale scored between 1.8 and 2.4 is a floor — a woman describing a life that has gone slightly grey does not necessarily register on it at all.

There is one more trap worth naming. The only "Anxiety" row in either adverse-reaction table — 4% on semaglutide against 2% on placebo — sits in Wegovy's paediatric table, from a trial of 201 patients aged twelve and over¹. It is not an adult figure, and it should not be quoted as one.

The trials measured whether you got ill. The complaint women actually have is that they got quiet. Those are not the same instrument.

Why a quieter head can feel like a smaller life

The plausible mechanism for the flattening is the same one that makes the drug work.

"Food noise" — the intrusive, repetitive mental chatter about eating that these medications famously switch off — has only recently been given a formal definition and measurement agenda in the research literature⁶. In a survey of 1,659 people taking GLP-1s, more than half of whom had been trying to lose weight for over a decade, participants reported that the medication reduced food-related thoughts and improved satiety, and named that quiet as one of the things they valued most about it⁷.

But appetite is regulated through reward circuitry that is not reserved for food. Turn the volume down on wanting a thing, and it is not obvious the mechanism knows which things it was hired to quieten.

That is exactly how the people living it describe it. Christopher McGowan, a gastroenterologist and obesity medicine specialist in North Carolina, told TODAY in May: "It is not in their heads. We are absolutely hearing this from people." He offered a patient's own phrasing, which is better than any clinical term: "One patient described to me as just a feeling that the lights had dimmed, but you couldn't really explain why."⁸ Dave Knapp, who hosts a podcast about these medications, described noticing three years into Mounjaro that he had stopped following the baseball team he had followed his whole life — "the drive and the motivation that maybe once would have been there was no longer there at all."⁸ He is now on a lower dose.

The honest position on all of this is that anhedonia is not listed as a side effect of any of these drugs, was not measured in the trials, and is not known to be caused by them. It is also, plainly, being described.

Irritability is a different problem, with a duller cause

Not every mood complaint is blunting, and the second-most common one has an unglamorous explanation.

A woman eating substantially less, sleeping badly through a titration, nauseated for three days out of seven and living on a fifth of the carbohydrate she used to is going to be short-tempered. The labels list irritability and mood changes only as symptoms of low blood sugar, in the section about combining these drugs with insulin or a sulfonylurea¹ — but the general principle survives the specific context. Undereating makes people snappish. So does broken sleep, which is its own thread on this medication.

The largest study of how any of this is actually discussed found the same mixed picture. Analysing 12,136 Reddit comments alongside YouTube and TikTok content, researchers found sleep problems the most common mental-health-adjacent theme, ahead of anxiety and then depression — and found weight loss on these drugs associated with "either a marked improvement or, in some cases, a deterioration, in mood"⁹. Both directions, in the same corpus. That is a real finding, not a failure to find one.

What to do with it

Three things, in order.

First, name it as a symptom rather than a character flaw. A woman who has spent thirty years being told her relationship with food is a discipline problem is extremely well trained to interpret "I feel nothing" as ingratitude.

Second, take it to the prescriber, and be specific about function rather than mood — not "I feel low" but "I have not opened the piano since March." McGowan's own threshold is behavioural: if someone is not getting out of bed, or says they no longer feel like working out on this medicine, that is a problem worth acting on. The available responses are a dose reduction, a switch of molecule, or stopping — and none of them should be improvised alone, because abruptly stopping is its own decision with its own consequences⁸.

Third, if the low mood is heavy rather than flat — hopelessness, or any thought of harming yourself — that is not a titration question. Call your clinician, or in the United States call or text 988, today rather than at the next appointment.

The thing that is genuinely unfair here is that the well-evidenced question got answered and the unmeasured one did not. Nobody has run a trial on whether a medication that quietens the wanting of food quietens the wanting of anything else. Until somebody does, the only instrument is a woman noticing, and saying so out loud to someone who can change the dose.