GLP 1 weight loss genetics study ties one receptor variant to slightly more loss and more nausea

A Nature paper on GLP 1 weight loss genetics has drawn fresh coverage this week, six months after it appeared in print. The study, run by researchers at 23andMe, links one letter change in the receptor gene targeted by semaglutide and tirzepatide to slightly more weight lost and to more nausea and vomiting. The effect per copy is small, and the data rest largely on what users reported about themselves.

What the survey measured

Starting in August 2024, 23andMe asked its research participants about weight loss medicines, and by August 2025 it had answers from 27,885 people using Ozempic, Wegovy, Mounjaro, Zepbound or compounded versions of semaglutide or tirzepatide. Most respondents were women (82.4%), the median age was 52, and 78.3% were of European ancestry. Participants reported a median of 8.3 months on treatment and a median loss of 11.7% of their starting weight. They also rated eleven side effects as none, mild, moderate or severe. The team then compared these reports with genotypes from the saliva samples that customers had already provided, looking across the genome for variants that lined up with bigger or smaller changes in body mass index.

Before any genes entered the analysis, simple factors already explained a good share of the spread. Women lost more than men, with a median fall in body mass index of 12.2% against 10.0%, and tirzepatide users lost more than semaglutide users over a similar time on drug. People with type 2 diabetes lost about 2.87 percentage points less. A model built on age, sex, starting weight, drug, dose and time on treatment explained about 21.4% of the variation in weight change. The authors also checked the self reports against electronic health records shared by a subset of participants. The records showed smaller losses, a median of 5.79% against 11.8% in the surveys, so the absolute numbers in the paper are probably generous.

One GLP1R variant and the weight loss genetics signal

The genome scan covered 15,237 unrelated people of European ancestry with complete data. One signal stood out on chromosome 6 at a variant called rs10305420, which sits inside GLP1R, the gene for the receptor these drugs activate. It changes the seventh amino acid of the receptor's signal peptide from proline to leucine. Each copy of the T allele went with about 0.76 kilograms of extra weight loss, with a confidence interval running from roughly 0.34 to 1.27 kilograms. The effect looked additive, so two copies meant about twice the gain. Among 42 candidate variants in the region, this was the only one that changes the protein, and the authors judge it the likely cause.

Each copy of the T allele went with about 0.76 kilograms of extra weight loss, with a confidence interval running from roughly 0.34 to 1.27 kilograms.

The team tested the variant again in 4,855 people from the US All of Us program, whose weights came from health records rather than surveys, and the link held in the same direction. It did not replicate in UK Biobank, where the data predate semaglutide and tirzepatide and statistical power was low. The T allele is common in people of European (40%) and Middle Eastern (38%) ancestry and rare in people of African ancestry (7%). In smaller groups of other ancestries the effect pointed the same way without reaching significance on its own. That pattern matters for anyone hoping to explain differences between populations, because the discovery scan itself included only one ancestry group.

Nausea vomiting and the GIPR finding in tirzepatide users

The same stretch of GLP1R also showed up when the team looked at side effects. Variants near the gene were tied to moderate or severe nausea and to vomiting, and a statistical test for shared causes suggested the efficacy and side effect signals probably trace back to the same underlying variant. In plain terms, the version of the receptor linked to more weight loss also went with feeling sicker. That fits earlier work suggesting that the gut and brain pathways behind fullness and nausea are related but separable, a theme covered in a mouse study on different brain wires for each effect. The human data here cannot say which nerve cells are involved, only that the genetic signals overlap.

A second finding applied only to tirzepatide, which acts on both the GLP 1 and GIP receptors. Among tirzepatide users, a protein changing variant in GIPR, rs1800437, was linked to vomiting, while no such link appeared in people taking semaglutide. The C allele carried about 1.83 times the odds of moderate or severe vomiting. People who carried two risk copies at both GLP1R and GIPR had 14.8 times the odds of vomiting on tirzepatide compared with people who carried none, although the confidence interval was wide, from 6.2 to 35.8. The GIPR variant did not change how much weight tirzepatide users lost. These estimates come from a few hundred people at the extremes, so they need independent replication.

What the weight loss genetics model can and cannot predict

The authors combined clinical, demographic and genetic factors into one prediction model. It explained about 25% of the variance in weight change in both the training data and a held out test set, and most of that came from non genetic factors rather than the variants. Models for side effects reached an area under the curve of 65.4% for nausea and 68.0% for vomiting, which means modest ability to rank people by risk. When the efficacy model was applied to health records from 642 people who had not taken the survey, those predicted to lose more did lose more on average. A table of the main figures from the Nature genome study follows.

The study has clear limits. All authors are current or former employees of 23andMe, the company that holds the data. Weight, dose and side effects were self reported, and the health record comparison showed reported losses ran higher than measured ones. Doses reported by users often exceeded those in their records. The discovery scan included only people of European ancestry, and a 0.76 kilogram difference is small next to the wide spread in how individuals respond. The paper offers no GLP 1 weight loss genetics test, and nothing here supports choosing or changing a medicine based on a consumer genotype. Readers following how outcomes vary across large groups may also want the veterans records study on benefits and harms.

Conclusion

This piece of GLP 1 weight loss genetics research points to the drug's own target, the GLP1R receptor, as one source of variation in response, and to GIPR as a source of vomiting with tirzepatide. The effects are real in the data but small, and most of the predictable spread still comes from sex, drug, dose and time on treatment. Independent cohorts with measured weights and more ancestries would show whether GLP 1 weight loss genetics can ever guide care.

Source Su QJ et al. Genetic predictors of GLP1 receptor agonist weight loss and side effects. Nature 653, 770 to 775 (2026). DOI 10.1038/s41586-026-10330-z

Amino Weekly reports research. It is not medical advice and never gives doses.