Can Your Genes Predict Weight Loss on Ozempic or Zepbound? The GLP-1 Genetics Evidence

Why do some people lose 25% of their weight on semaglutide or tirzepatide while others lose 5%, and could a DNA test tell you in advance? A 2026 Nature study of 27,885 people found a genetic variant clearly linked to GLP-1 weight loss, in the gene for the GLP-1 receptor itself.^[1] The effect is real but small: about 0.76 kg (1.7 lb) of extra weight loss for each copy of the variant. The same study found genes linked to nausea and vomiting. Here is what the genetics shows, what it doesn't, and why your sex and how long you stay on treatment matter more than your DNA for now.

The short answer

• Genes matter a little. A missense variant in GLP1R, the gene for the GLP-1 receptor, was associated with about 0.76 kg more weight loss per copy.^[1]
• Side effects may be partly genetic too. Variants in GLP1R and GIPR were linked to nausea or vomiting, and the GIPR link appeared only in people taking tirzepatide.^[1]
• Sex matters more. In 6 randomized trials (19,906 patients) that reported results by sex, women lost 10.9% of body weight on GLP-1 drugs compared with 6.8% for men. In the wider 64-trial analysis, age, race, ethnicity and starting BMI made no significant difference.^[2]
• A DNA test can't tell you yet whether a GLP-1 will work for you. The genetic effect is too small to change a prescribing decision, and no test is validated for this.

The 2026 Nature study

Researchers at the 23andMe Research Institute ran a genome-wide association study (GWAS) of “self-reported weight loss and treatment-related side effects in 27,885 people following GLP1 receptor agonist therapy.” They found three things:^[1]

• A missense variant in GLP1R was “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.”
• Variation in both GLP1R and GIPR was linked to nausea or vomiting, “with the GIPR association being restricted to people using tirzepatide.” That fits biology: tirzepatide acts on the GIP receptor and semaglutide does not.
• Combining these findings into a broader model, they could “stratify patients by efficacy and side effect risk.”

The authors call this a foundation “for precision medicine approaches in the treatment of obesity.”^[1] That is a long way from a test you could use today. Weight loss and side effects were self-reported by 23andMe customers, not measured in a clinic, and the study did not test whether choosing a drug by genotype leads to better results.

How big is 0.76 kg?

Put next to typical results, it is small. In the STEP 1 trial, people taking semaglutide 2.4 mg lost 14.9% of body weight on average, about 15 kg for someone starting at 100 kg.^[5] Someone with two copies of the variant would be expected to lose about 1.5 kg (3.3 lb) more than someone with none. That is worth knowing scientifically, but it is far smaller than the differences people see in practice, and it would not change which drug or dose a doctor chooses.

A second variant: GLP1R R131Q and blood sugar

A separate 2026 Diabetes Care study looked at a “gain-of-function” GLP1R variant called R131Q. In people with type 2 diabetes, each copy came with an extra 0.53-point drop in HbA1c after six months on a GLP-1 drug. In lab experiments the variant boosted the insulin response to the drugs.^[3] This is about blood-sugar response rather than weight, but it points the same way: small differences in the receptor gene change how well the drugs work.

What predicts weight loss better than genes

Sex

A 2026 JAMA Internal Medicine meta-analysis looked for differences in treatment effect across 64 randomized trials. In six trials that reported results by sex (19,906 patients), women lost 10.9% of their body weight and men 6.8%. The authors found “no significant HTE [heterogeneity of treatment effect] by age,” race, ethnicity, baseline BMI or baseline HbA1c.^[2] Real-world telehealth data point the same way: in one US cohort, women had more than four times the odds of losing at least 5% (odds ratio 4.26).^[4]

Staying on treatment

Whether people keep taking the drug at an effective dose also matters. See our review of real-world telehealth outcomes and why you might not be losing weight on a GLP-1.

Should you get a DNA test before starting a GLP-1?

Not for this purpose, based on current evidence. The variant's effect is 0.76 kg per copy, the data were self-reported, and no study has shown that genotype-guided prescribing improves outcomes. If you aren't responding, the practical steps are reviewing dose, side effects and consistency, or switching drugs. See GLP-1 non-responders and switching strategies and switching from semaglutide to tirzepatide.

Bottom line

Your genes play a measurable but small part in how much weight you lose on a GLP-1 and how likely you are to feel sick. The first clear variant adds less than a kilogram per copy. Being female, staying on treatment and reaching an effective dose matter far more. Genetic prediction may become useful as more variants are found, but no test today should decide whether you start, or which drug you choose.