Genetic predictors of GLP1 receptor agonist weight loss and side effects
- Design
- Design unknown · 15237 participants · Intermediate outcome
- Match to healthy normal-weight adults aged 55–75
- Different populationINDIRECT: people taking semaglutide or tirzepatide mostly for weight loss, 96.8% with BMI of at least 25 and median BMI 35.1, mostly women (82.4%) with median age 52; 23.1% used the drug for blood sugar. No results are reported for normal-weight people.
- Could weight loss explain it?
- Not applicableNot applicable: the efficacy outcome is percentage BMI change, so weight loss cannot mediate itself, and the side-effect outcomes are genotype associations rather than a benefit to attribute. The authors' co-localization of the weight and nausea signals at GLP1R is a genetic association, not a mediation analysis.
- Study tier
- Study tier 4VERY_LOW: a genome-wide association study within one-off self-reported survey data from treated 23andMe research participants, with no untreated comparator (IN-063: an uncontrolled cross-sectional study of treated users is VERY_LOW); the weight signal replicated in All of Us health-record data but not in UK Biobank.
- Assessment
- Version 3 · Two independent AI reviews agreed · Nov 11, 2026
Study facts come from the paper. Population match, weight-loss explanation and study tier are our judgments, made against the reference group of healthy normal-weight adults aged 55–75.
Researchers at 23andMe surveyed research participants who had taken semaglutide or tirzepatide (27,885 people, median BMI 35.1; 15,237 in the main genetic analysis) and searched their genomes for variants linked to self-reported weight loss and side effects. A variant in the GLP-1 receptor gene was linked to slightly more weight loss, about 0.76 kg per copy, and variants near the same gene were linked to nausea and vomiting; a variant in the GIP receptor gene was linked to vomiting only in people taking tirzepatide. The weight-loss link was repeated in one health-records cohort but not in UK Biobank. Everyone in the study took a drug and reported their own weight and side effects, so these are associations that show who may respond differently, not whether the drugs work compared with not taking them.
01Findings
What the study reported
- Drugs
- Semaglutide, Tirzepatide
- Dose
- various (self-reported most recent weekly dose)
- Comparator
- none: genotype groups compared within treated users, with no untreated group (the variant effect was also compared between semaglutide and tirzepatide users)
- Treatment duration
- median 8.3 months (self-reported)
- Primary outcome
- percentage BMI change on treatment (self-reported) and moderate-to-severe side effects (nausea, vomiting) by genotype
- Effect
- GLP1R missense variant rs10305420: 0.641% more BMI loss (about 0.76 kg) per T allele, P = 2.9 × 10-10; replicated in All of Us (effect -0.47%, P = 0.001) but not in UK Biobank; GLP1R-region variants linked to vomiting (OR 1.57) and nausea (OR 1.36); GIPR variant rs1800437 linked to vomiting only in tirzepatide users and not to weight loss (P = 0.73).
- 95% confidence interval
- 95% CI −1.27 to −0.34 kg (per-allele weight effect of rs10305420)
- P value
- P = 2.9 × 10-10 (rs10305420 and BMI loss)
- Follow-up
- Not stated
- Adverse events
- Self-reported nausea and vomiting were associated with variants near GLP1R (vomiting OR 1.57; nausea OR 1.36), and a GIPR variant with vomiting in tirzepatide users only; people homozygous for risk alleles at both genes had 14.8-fold odds of tirzepatide-related vomiting.
- Limitations
- Weight change and side effects were self-reported in a one-off survey with no untreated comparator; health records showed smaller weight loss than self-report (−5.79% vs −11.8%), the sample was mostly female and of European ancestry, genetic effects were modest beside non-genetic factors, and the efficacy signal did not replicate in UK Biobank, where power was low and the data predate semaglutide and tirzepatide.
Who was studied
- Mean age
- median 52 years
- Age min
- 18
- Sex distribution
- 82.4% female
- Bmi mean
- median 35.1 before treatment
- Obesity status
- mixed (96.8% had BMI ≥25 at baseline; median BMI 35.1)
- Diabetes status
- mixed (23.1% reported using the drug to improve blood sugar)
- Baseline condition
- use of semaglutide or tirzepatide, mostly for weight loss
- Sample size
- 15237
- Inclusion exclusion
- 23andMe research participants who had taken prescription weight-loss medication and reported using a semaglutide or tirzepatide product; GWAS limited to unrelated people of European ancestry with complete data; quality-control filters on weight, height, BMI and age under 18
Study quality details
- Study design
- Unknown
- Sample size
- 15237
- Randomization
- no
- Blinding
- n/a
- Comparator
- none: genotype groups compared within treated users, with no untreated group (the variant effect was also compared between semaglutide and tirzepatide users)
- Follow up duration
- not stated
- Outcome type
- intermediate
- Replication
- not assessed
- Consistency with other evidence
- not assessed
- Population applicability
- INDIRECT
- Statistical precision
- genome-wide P values and one 95% CI reported; GWAS n 15,237; replicated in 4,855 All of Us participants; not replicated in UK Biobank, where expected power was low
- Risk of bias
- not assessed
- Peer review status
- yes
02Funding
Funding and conflicts
- Funding
- not reported
- Industry funded
- Unclear
- Manufacturer
- not stated in source
- Sponsor role
- not reported
- Author conflicts
- All authors are current or former 23andMe employees; one also holds an academic appointment at the Icahn School of Medicine at Mount Sinai and one is now affiliated with Genentech.
- Independent replication
- unknown
Funding is shown on every study and never used to score it.
04Source
The source, as retrieved
Abstract
The development of glucagon-like peptide 1 (GLP1) receptor agonists, including semaglutide and tirzepatide, has transformed the clinical management of overweight and obesity. However, substantial inter-person variability exists in both weight loss efficacy and the incidence of side effects1. To investigate the genetic basis of this variability, here we conduct a genome-wide association study of self-reported weight loss and treatment-related side effects in 27,885 people following GLP1 receptor agonist therapy. We identify a missense variant in GLP1R that is associated significantly with increased efficacy of GLP1 medications (P = 2.9 × 10-10), with an additional -0.76 kg of weight loss expected per copy of the effect allele. Furthermore, we identify associations linking variation in both GLP1R and GIPR to GLP1 medication-related nausea or vomiting, with the GIPR association being restricted to people using tirzepatide. We incorporate these findings into a broader model of GLP1 medication response, and demonstrate the ability to stratify patients by efficacy and side effect risk. These findings provide direct genetic evidence that variation in the drug target genes contributes to inter-person variability in response and lay the foundation for precision medicine approaches in the treatment of obesity.
Where this record came from
| Source | Retrieved | Identifier |
|---|---|---|
| pubmed | Nov 11, 2026 | 41951734 first ingestion |