Background

A study published in Nature reported associations between variants in drug-target genes and weight loss, nausea, and vomiting during treatment with GLP-1 drugs. Researchers at 23andMe analyzed self-reported data from 27,885 drug users and found an association between GLP1R variants and additional weight loss. Genetic information alone cannot yet accurately predict an individual’s treatment response or identify the best drug for them.

Semaglutide, a glucagon-like peptide-1 (GLP-1) receptor agonist, and tirzepatide, which stimulates both GLP-1 and glucose-dependent insulinotropic polypeptide (GIP) receptors, are used for weight management. Why do weight loss and side effects vary among people taking the same drug? The researchers examined how much genetic differences, alongside factors such as drug type, dose, and duration of use, explain this variation. This was an observational genome-wide association study (GWAS), not a randomized clinical trial.

Key findings

The median duration of drug use among all respondents was 8.3 months. The weight-change GWAS included 15,237 participants of European ancestry who met the analysis criteria. The main outcome was the percentage change in body mass index (BMI) from before to after treatment. With adult height held constant, this is equivalent to percentage change in weight.

The rs10305420 variant in the GLP-1 receptor gene GLP1R was associated with greater weight loss. Each copy of the T allele was associated with an estimated additional loss of about 0.76 kg; the 95% confidence interval for the additional loss was 0.34–1.27 kg. This is an association estimated in a population and should not be interpreted as a guaranteed effect for an individual. The variant changes the seventh amino acid in the protein from proline to leucine.

For side-effect analyses, researchers compared people reporting moderate or severe symptoms with those reporting mild or no symptoms. Variants near GLP1R were associated with nausea and vomiting, while the C allele of rs1800437 in the GIP receptor gene GIPR was linked to increased vomiting risk among tirzepatide users. The effect of the same GIPR variant was not observed among semaglutide users. The findings point to the need to account for differences in the drugs’ targets.

The researchers also built prediction models combining genetic information with treatment, demographic, and clinical data. In test data set aside for evaluation, the weight-change model explained about 25% of the variance in percentage BMI change. The area under the receiver operating characteristic curve (AUC) was 0.654 for the nausea model and 0.680 for the vomiting model. AUC measures how well a model distinguishes groups with and without symptoms; it is not the percentage of correct individual predictions. Non-genetic factors contributed more to all three models: weight change, nausea, and vomiting.

Meaning and outlook

The findings add evidence that differences in drug-target genes contribute to variation in treatment response. One possible mechanism proposed by the researchers is that GLP1R variants alter receptor movement inside cells and the amount of receptor at the cell surface. This mechanism was not directly demonstrated in the study. The researchers also found no evidence that variants regulating gene expression and the weight-loss association signal share the same causal variant.

The association between GLP1R and weight loss was replicated in the same direction in electronic health record (EHR) data from 4,855 people in All of Us. It was not replicated in UK Biobank, which the researchers attributed to the data’s low statistical power. Generalizability is also limited by the participant makeup: 82.4% were women and 78.3% were of European ancestry. Measurement differences also warrant attention, as participants who provided both self-reports and EHR data reported greater weight loss in their self-reports.

The study did not test whether selecting drugs based on genetic testing improves treatment outcomes. The researchers proposed collecting longitudinal data that track both weight change and side effects during dose escalation as a next step. The study design and results are available in the full-text Nature paper.