Yes, and the size of it has been measured. Forty-six people taking a GLP-1 scored 28.61 on a 53-item taste test against 40.63 in matched controls, at p<0.001 and an effect size of η²=0.37 [1]. Eighty-five per cent of them scored worse than the specific control they were matched against, while smell barely moved at p=0.076. Nobody’s taste was measured before they started, so this is a large difference between two groups rather than a demonstrated cause. What changes alongside it is how much gets eaten: at a standardized test meal the gap ran 291.9, 240.2 and 269.5 kcal at weeks 20, 40 and 60 [2].
Everybody taking these drugs is asked whether their appetite has changed. Almost nobody is asked whether food still tastes like food. It turns out to be a measurable question with a large answer, and one that sits outside every question a seller thinks to answer.
What was measured
Forty-six people taking a GLP-1 and forty-six controls matched on age, sex, smoking and COVID-19 history each completed two standardized instruments: a 53-item waterless taste test covering the five basic qualities, and a 40-item smell identification test [1].
Taste scores in the drug group averaged 28.61, 95% CI 25.66 to 31.56, against 40.63, 95% CI 38.35 to 42.91, in controls. The difference was significant at p<0.001 with an effect size of η²=0.37, which is large. Eighty-five per cent of the people on the drug scored worse than the specific control they had been matched against, and all five taste subtests moved together, each at p<0.001.
Smell was a different story: it was slightly lower on average and the difference did not reach significance, at p=0.076. So this looks like a taste effect rather than a general chemosensory one, which is unusual, because most things that blunt flavor do it through smell.
The finding that points the wrong way
Here is the part that stops this being a tidy story. Taste and smell scores were higher — better — in the participants reporting nausea, diarrhea and other drug-related side effects, which the authors call remarkable and which it is.
If the taste effect were simply a marker of how strongly the drug was acting on somebody, you would expect the people with the most side effects to have the worst taste scores; they had the best. That undercuts the simplest explanation and leaves the mechanism open — the authors suggest receptors in the brainstem and afferent taste pathways, and vagal processes, all of which is hypothesis rather than finding. It is also a reminder that side effects are not a dose meter.
Why it matters practically
Because what you eat follows what you can taste. When food registers as flat, people drift toward salt, sugar, texture and convenience — the things that still deliver something. That is the opposite of the dietary pattern anybody would prescribe alongside rapid weight loss. It compounds a problem covered already on this site, that people on these drugs are eating below requirements for several nutrients and largely not being tested for it.
It also has a quieter effect on how people judge the drug. Someone who has stopped enjoying meals may read that as discipline, or as the drug working. It may just be that dinner has stopped tasting like dinner.
Where the mechanism is being looked for
The authors point at brainstem receptors, afferent taste pathways and vagal processing, and that is where the laboratory work has gone too. A 2026 circuit study mapped which brainstem neurons carry which effect of these drugs and found that neurons in the area postrema did nothing for ordinary feeding while carrying both the weight loss and the aversion [3]. Its conclusion is that the nauseating and weight-lowering effects cannot be separated at a circuit level. That is preclinical work in laboratory animals and the indexed abstract does not name the species, so it is a direction to look rather than an explanation of a human taste score.
How much to lean on it
Not far: ninety-two people, one time point, no baseline, and a title that overstates what the design can show. It is a strong enough signal to take seriously and far too thin to treat as established, which is roughly where most small studies in this field sit. What would settle it is measuring the same people before and after starting.
In the meantime it is a question worth asking yourself and answering honestly, and one nobody selling you the drug will raise — the list of things they leave to you is long and this belongs on it.