# Taste Got Better, Not Worse

By Nora Bissett, Pricing Editor (https://glpwatchdog.org/authors/nora-bissett) | Published: 2026-08-20 | Updated: 2026-10-27

Canonical page: https://glpwatchdog.org/research/taste-got-better-not-worse

> People on these drugs routinely say food tastes wrong. A trial measured taste objectively, biopsied participants' tongues and scanned their brains, and found taste recognition improved.

## Key points

- Taste recognition improved 11.9 → 14.4 points; treatment difference 2.5 (CI 1.7–3.3).
- That is compatible with food tasting wrong, because recognition measures sensitivity, not enjoyment.
- Three levels of evidence: taste strips, tongue BIOPSIES, and fMRI.
- Gene changes involved taste-bud renewal and plasticity, a pattern that fits adaptation more than damage.
- Single-blinded: participants did not know their assignment, the investigators did, and the taste test was given in person.
- 30 women with PCOS, 16 weeks, semaglutide 1.0 mg. Enjoyment was never measured.

People on these drugs routinely say food tastes wrong: metallic, flat, off. A randomized trial measured taste objectively, took biopsies of participants’ tongues, and scanned their brains while they tasted. Taste recognition **improved**.[1] The two are less contradictory than they look, and the reason explains something patients have been describing for years.

## What was measured, and how deeply

Thirty women with polycystic ovary syndrome, averaging 33.7 years and a BMI of 36.4, were randomized to weekly semaglutide 1.0 mg or placebo for 16 weeks. The study then measured taste at three independent levels, which is unusual at this sample size.

- **Taste recognition**, using 16 strips carrying four concentrations of each of the four basic tastes.
- **Gene expression in the tongue**, from biopsies.
- **Brain activity**, by functional MRI, in response both to pictures of food and to a sweet solution dripped onto the tongue.

*What each level found.[1]*

| Level | Finding |
| --- | --- |
| Taste recognition score | **11.9 → 14.4 points**; treatment difference 2.5 (95% CI 1.7–3.3) |
| Tongue gene expression | Differential RNA expression in *EYA*, *PRMT8*, *CRLF1*, *CYP1B1* — genes tied to taste signaling, neural plasticity and taste-bud renewal |
| Brain response | Reduced putamen activation to food images; increased angular gyrus activity to sweet solution after eating (P < .001) |

Three separate kinds of measurement point the same way. For a 30-person trial that is a lot of converging evidence, and it makes the central finding harder to dismiss than a questionnaire result would be.

## Why 'better' and 'unpleasant' are both true

> Taste recognition improved by 2.5 points against placebo. Nobody measured whether food tasted better.

> **The test measured recognition: whether someone could correctly identify a taste at a given concentration.** Scoring higher means the system is more sensitive. It says nothing about whether the resulting experience is welcome. Sharper detection of bitterness in coffee, or of sweetness in something already sweet, would register as improvement on this test and as food tasting wrong at the table.

That resolves the apparent conflict without dismissing anyone. When someone says food tastes strange on this drug, the plausible reading is that their perception has changed rather than dulled, possibly by becoming more acute, and the familiar no longer matches what they expect.

The brain findings fit that. Reduced putamen activation to pictures of food is consistent with lower *reward anticipation*, or wanting food less on sight. A separate change appeared in response to tasting sweetness after a meal. Wanting less and perceiving more are different systems, and they moved differently.

Enjoyment itself has not been measured directly. Nobody has run a trial asking whether food is more or less *pleasant* on these drugs, which is the question patients are raising. Smell, which carries most of what people call flavor, is covered in [smell changes and body odor on a GLP-1](https://glpwatchdog.org/research/glp1-smell-changes-and-body-odor). What patients themselves say about the experience is in [eight things patients said](https://glpwatchdog.org/research/eight-things-patients-said).

## The design flaw

> **This was single-blinded, not double.** Participants did not know their assignment; the investigators did.[2] And the primary measurement was a taste-strip test given in person. That is the specific design where an experimenter’s knowledge can leak into a result, and it is the main reason to hold the recognition score more loosely than the biopsies and the scans.

The gene expression and imaging findings are far less vulnerable to that, because neither is scored by a human judgment in the room. Their agreement with the taste score is what keeps the overall picture credible.

- **Thirty participants**, all women, all with polycystic ovary syndrome.
- **Sixteen weeks**, on semaglutide 1.0 mg (the diabetes dose, not the 2.4 mg used for weight).
- **No measure of enjoyment**, palatability, or whether people changed what they ate as a result.

## What it is useful for

Mostly for reassurance of a specific kind. Altered taste on these drugs is a real, physically grounded phenomenon, with measurable changes in the tongue and in the brain. Someone reporting that food is different is describing something with biology behind it.

It also suggests taste changes are unlikely to reflect damage. Differential expression of genes governing taste-bud renewal and neural plasticity fits a system adapting rather than degrading.

## Frequently asked questions

### Do GLP-1 drugs change how food tastes?

Yes, measurably. In a randomized trial, taste recognition improved by 2.5 points against placebo, gene expression in tongue biopsies changed in taste-related pathways, and brain responses to food images and to sweetness shifted.

### People say food tastes bad on these drugs. Does this contradict that?

No. The test measured whether someone could correctly identify a taste, which is sensitivity rather than enjoyment. More acute detection of bitterness or sweetness would score as improvement and could easily register as food tasting wrong.

### Has anyone measured whether food is less enjoyable?

Not directly, which is the gap. This trial measured recognition, gene expression and brain activity. Palatability and enjoyment were not assessed.

### Are taste buds being damaged?

The gene expression changes involved taste signaling, neural plasticity and taste-bud renewal, which fits a system adapting rather than degrading.

### How reliable is the study?

It was single-blinded, so investigators knew the assignments while administering a scored taste test, which is the weakest link. The tongue biopsies and brain imaging are far less vulnerable to that, and their agreement with the taste score is what supports the overall picture. Thirty women with PCOS, over 16 weeks.

## References

1. Jensterle M, Kovac J, Vovk A, et al. Semaglutide and Taste in Women With Obesity and Polycystic Ovary Syndrome: A Randomized Placebo-Controlled Study. The Journal of Clinical Endocrinology and Metabolism. 2025. PMID: 40341357. https://pubmed.ncbi.nlm.nih.gov/40341357/
2. University Medical Centre Ljubljana The Effects of Semaglutide on Taste, Tongue Tissue Transcriptome, Gastric Emptying and Central Neural Response in Women With PCOS and Obesity (NCT04263415). ClinicalTrials.gov. 2023. https://clinicaltrials.gov/study/NCT04263415

---

Cite as: GLP Watchdog, Nora Bissett, "Taste Got Better, Not Worse", published 2026-08-20, updated 2026-10-27. https://glpwatchdog.org/research/taste-got-better-not-worse

Canonical HTML page: https://glpwatchdog.org/research/taste-got-better-not-worse

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.
