Science

Your Genes May Predict How Well GLP-1 Medications Work for You

GLP-1 Companion · 9 min read

One of the most frustrating realities of GLP-1 treatment is how unpredictably it works across individuals. Two people can start the same medication on the same day at the same dose and have dramatically different experiences — one loses 20% of their body weight with minimal side effects, while the other loses 6% and struggles with persistent nausea. For years, clinicians and patients have wondered why. A major study published in Nature in April 2026 has begun to answer that question, pointing directly to differences in our DNA.

The Study: A Genome-Wide Look at 27,885 GLP-1 Users

Researchers at 23andMe conducted the largest genome-wide association study (GWAS) ever performed on GLP-1 medication outcomes. The study enrolled 27,885 individuals who reported using GLP-1 receptor agonists — including semaglutide (Ozempic, Wegovy) and tirzepatide (Mounjaro, Zepbound) — and combined genetic data with self-reported outcomes including weight loss and treatment-related side effects such as nausea and vomiting.

By scanning hundreds of thousands of genetic variants across the genome and comparing them against reported weight loss and side effect outcomes, the team was able to identify specific DNA variants that predict individual responses to these medications. The results were published on April 8, 2026, in Nature.

Finding #1: A Gene That Predicts More Weight Loss

The most significant finding related to weight loss efficacy was a missense variant in the GLP1R gene — the gene that encodes the GLP-1 receptor, which is the direct molecular target of semaglutide and GLP-1-based therapies. The variant, rs10305420, is a naturally occurring amino acid change in the receptor protein.

People carrying this GLP1R variant experienced approximately 0.76 kg (about 1.7 lbs) more weight loss per copy of the effect allele compared to non-carriers. Since humans carry two copies of each gene, someone with two copies of the variant could expect roughly 1.5 kg of additional weight loss compared to someone with no copies — a modest but statistically meaningful difference that maps directly onto the drug's molecular target.

• The variant sits in the GLP1R gene, which codes for the receptor that semaglutide directly binds
• Each copy of the effect allele was associated with approximately 0.76 kg additional weight loss
• The effect appears to reflect differences in how well the receptor responds to drug binding
• This kind of genetic variant — where a drug works better because its target is more responsive — is exactly what precision medicine researchers look for

Finding #2: Genetics of Nausea and Vomiting — and a Drug-Specific Effect

The second major finding concerned the most common and limiting side effect of GLP-1 medications: nausea and vomiting. The study identified genetic associations with nausea/vomiting risk in both GLP1R and GIPR — the latter being the receptor for glucose-dependent insulinotropic polypeptide (GIP), the second hormone targeted by tirzepatide.

Critically, the GIPR association was drug-specific: it only appeared in people using tirzepatide, not in those using semaglutide alone. This makes biological sense — tirzepatide activates both GLP-1 and GIP receptors, while semaglutide targets only GLP-1. People with certain GIPR variants may have a GIP receptor that amplifies the nausea-inducing effects when tirzepatide activates it, a side effect pathway that semaglutide users simply don't experience.

How Wide Is the Range of Predicted Outcomes?

Beyond identifying individual variants, the researchers built a predictive model incorporating genetics alongside demographic and clinical factors (sex, drug type, dose, treatment duration). The model revealed striking individual variation in expected outcomes:

• Predicted weight loss ranged from approximately 6% to 20% of starting body weight across individuals
• Predicted nausea/vomiting risk ranged from approximately 5% to 78%
• Non-genetic factors — sex, which drug was used, dose, and duration of treatment — explained a substantially larger share of the outcome variability than genetics alone
• Genetics contributed meaningfully to predictions but did not dominate them

The broad range in predicted outcomes underscores that GLP-1 response is genuinely heterogeneous — it is not random noise, but rather a structured pattern driven by identifiable biological and clinical factors. For patients currently experiencing disappointing results, this research suggests their experience is likely explainable, not simply bad luck.

Important Caveats: What the Study Does and Does Not Tell Us

External experts reviewing the study for Science Media Centre (UK) highlighted several important limitations that temper clinical enthusiasm:

1. Self-reported outcomes: Weight loss was self-reported, and the study found that self-reported figures (average 11.8% weight loss) were substantially higher than what electronic health records typically show (~5.8%). Self-report bias is a known issue in weight loss research
2. Effect size is modest: While statistically robust, the GLP1R variant effect of ~0.76 kg per allele is small relative to the typical 10–15% weight loss seen with these medications overall. Genetics is not a large fraction of total outcome variability
3. Population limitations: The study cohort was predominantly female and of European ancestry, which limits how well the findings generalize to men and non-European populations
4. Short treatment duration: The median treatment duration was only 8 months — shorter than the typical long-term treatment arc for obesity management
5. Not ready for clinical use: Experts concluded that current evidence is not sufficient to support using genetic testing to guide GLP-1 prescribing decisions in routine clinical practice

"Non-genetic factors such as sex, drug type, dose, and duration appear to explain a substantially larger proportion of variability — genetics is only one component of response variation." — Dr. Marie Spreckley, University of Cambridge

Why This Research Matters Anyway

Despite these limitations, the study represents a genuine scientific advance. Finding genetic variants that predict drug response — especially variants located directly in the gene encoding the drug's target — is the cornerstone of pharmacogenomics. The fact that the GIPR variant effect is drug-specific (tirzepatide but not semaglutide) is a particularly compelling piece of evidence that the finding reflects real biology, not statistical noise.

Previous pharmacogenomics successes — in cancer treatment, psychiatric medication, and anticoagulation — followed this same arc: initial discovery of genetic predictors, followed by years of replication and refinement, followed eventually by clinical implementation. GLP-1 genetics appears to be entering that pipeline.

The 23andMe Clinical Application

In parallel with the Nature publication, 23andMe launched a consumer-facing product called "GLP-1 Medications: Weight Loss and Nausea" through its Total Health service. The report uses the identified genetic variants alongside individual health data to generate personalized predictions of expected weight loss and nausea risk for people considering or already using GLP-1 medications.

This represents an early commercial translation of pharmacogenomic research into patient-facing tools — though it should be noted that such consumer products typically move ahead of clinical validation. The product may be informative for patients curious about their genetic profile, but should not be used to make medication decisions without physician guidance.

What This Means if You Are Currently on a GLP-1 Medication

For patients already using semaglutide or tirzepatide, this research provides useful framing even if it does not change immediate clinical decisions:

• If you are a slow responder, your genetics may be part of the explanation — but so are dose, duration, diet, and other modifiable factors. Don't stop treatment based solely on early results
• If you experience significant nausea on tirzepatide, GIPR genetics may be a contributing factor — this is not a personal failing or something you can simply will away
• If you are considering switching between semaglutide and tirzepatide, the drug-specific nature of the GIPR nausea effect suggests that switching could meaningfully change your side effect profile — this is worth discussing with your prescriber
• Genetic testing for GLP-1 response is not yet standard of care — do not delay treatment while waiting for genetic guidance
• Non-genetic factors (how long you stay on the medication, dose escalation approach, dietary changes) still explain more of the outcome variance than genetics — these remain the most actionable variables

Looking Ahead: Precision Medicine for Obesity

The April 2026 Nature study is part of a broader shift in obesity medicine toward precision treatment — matching patients to therapies based on their individual biology rather than applying a one-size-fits-all approach. GLP-1 medications are currently prescribed using BMI thresholds and comorbidity criteria, with no biological biomarkers guiding which medication to choose or what dose to use.

Genetic predictors of response are one piece of that future precision toolkit. The research community will need to replicate these findings in independent cohorts, expand studies to include diverse ancestries, conduct prospective trials where genetic information is used to guide treatment, and ultimately demonstrate that genotype-guided prescribing leads to better outcomes than standard care.

That work will take years. But for the first time, we have robust genetic signals pointing directly at the molecular targets of the most important obesity medications of our era — and that is a meaningful scientific beginning.

Sources

• Belsham et al. (2026). Genetic predictors of GLP1 receptor agonist weight loss and side effects. Nature.
• Wilding et al. (2021). Once-Weekly Semaglutide in Adults with Overweight or Obesity. NEJM.
• Jastreboff et al. (2022). Tirzepatide Once Weekly for the Treatment of Obesity. NEJM.