GLP1R and GIPR Genetics: Why GLP-1 Drugs Work for Some and Not Others

Important Notice: This article is intended exclusively for scientific information and research purposes. All substances mentioned are not intended for human consumption. Always consult qualified professionals before using peptides.

Introduction: A 27,885-Person Answer to a Long-Standing FAQ

On 8 April 2026, Nature published a genome-wide association study (GWAS) led by 23andMe in collaboration with academic clinical centres. Drawing on de-identified survey and genetic data from 27,885 self-reported users of semaglutide, liraglutide, dulaglutide and tirzepatide, the analysis is the largest pharmacogenomic study of GLP-1 receptor agonists conducted to date. It directly addresses one of the most common questions in metabolic medicine: why does the same drug, at the same dose, produce 6 to 20 percent weight loss in different patients, with side effect rates ranging from 5 to 78 percent?

The short answer, according to the authors, is that common variants in the genes encoding the GLP-1 receptor (GLP1R) and the GIP receptor (GIPR) account for a meaningful share of the variance. The longer answer is more nuanced, and it carries implications for how researchers should think about non-responders, dose escalation, and combination therapies such as tirzepatide or the triple agonist retatrutide.

What Did the Study Find?

The 23andMe-led analysis cross-referenced participant genotype data with self-reported outcomes after at least 12 weeks of GLP-1 therapy. Three primary findings stand out:

1. GLP1R variants are associated with weight loss

Several common single-nucleotide polymorphisms (SNPs) in or near the GLP1R locus on chromosome 6p21 reached genome-wide significance for weight loss response. Carriers of the more responsive haplotype lost roughly twice as much body weight on average as carriers of the less responsive haplotype, with the population spread ranging from approximately 6 percent to 20 percent.

2. GIPR variants modulate tirzepatide response

Among tirzepatide users, variants in GIPR independently predicted weight loss in addition to GLP1R variants. This is mechanistically consistent with tirzepatide's dual GLP-1 plus GIP agonism: a patient can carry a favourable GLP1R haplotype but a less responsive GIPR haplotype, or vice versa, producing intermediate phenotypes.

3. Side effect liability tracks the same loci

Self-reported nausea ranged from 5 percent to 78 percent across the cohort. Some of the variants associated with greater weight loss were also associated with higher gastrointestinal side effect rates, hinting that efficacy and tolerability share at least part of the same receptor-level signalling pathway.

How GLP1R and GIPR Variants Work

To understand why a single nucleotide change can shift drug response so substantially, it helps to revisit basic class B GPCR pharmacology.

Class B G-protein coupled receptors

GLP1R and GIPR are both class B (secretin-family) G-protein coupled receptors. They have a large extracellular N-terminal domain that binds the C-terminus of the peptide ligand, and a seven-transmembrane domain that engages the N-terminus of the ligand and propagates the conformational change to intracellular G-proteins (primarily Gs, leading to cAMP) and beta-arrestins.

Where variants matter

Common variants identified in GWAS analyses of GLP1R and GIPR cluster in three functionally relevant regions:

Extracellular domain: alters peptide binding affinity
Transmembrane helices: alters the conformational change that follows binding
Intracellular loops and C-terminus: alters G-protein versus beta-arrestin coupling, the ratio of which influences both efficacy and desensitisation

A single amino acid substitution in any of these regions can shift the dose-response curve, the maximal response, or the bias between G-protein and arrestin signalling, all without abolishing receptor function. For a chronic therapy taken at relatively narrow doses, this is enough to produce a clinically visible spread in outcomes.

Why nausea tracks efficacy

Both anorexigenic effects (via the area postrema and hypothalamus) and nausea (via the area postrema) are mediated by the same central GLP-1 receptors. A receptor variant that increases on-target signalling tends to increase both desired and undesired effects in parallel, which is exactly what the new GWAS observed.

What This Means for Non-Responders

The most provocative implication of the Nature study is that a meaningful fraction of GLP-1 non-responders are not failing because of dose, adherence or lifestyle. They are failing because their receptors do not signal as efficiently in response to the drug.

The classical clinical assumption

Until now, when a patient lost only 4 to 6 percent on semaglutide 2.4 mg, the standard clinical playbook was: titrate to maximum dose, address adherence, address lifestyle, then consider switching to tirzepatide. This still applies. But the 23andMe data suggest that a subgroup of patients carries inherently low-responsiveness alleles in GLP1R, and that simply pushing the dose higher will not close the gap because the bottleneck is at the receptor, not at drug exposure.

What this does not mean

It does not mean a patient is "immune" to GLP-1 therapy. Even low-responsiveness haplotypes still produce some weight loss.
It does not mean genetic testing should be used to gate access to therapy. The effect sizes are population-level, not deterministic at the individual level.
It does not mean current dosing is wrong. The recommended titration schedules remain appropriate; genetics simply explains some of the residual variance.

Implications for Combination Therapies

If the GLP1R is suboptimal for a given patient, does adding a second hormonal axis recover the response? The new data offer a tentative yes, with caveats.

Tirzepatide (GLP-1 + GIP)

Patients with low-responsiveness GLP1R haplotypes but favourable GIPR haplotypes appear to benefit disproportionately from tirzepatide compared with semaglutide, because the GIP component compensates for the weaker GLP-1 signal. Conversely, patients with low-responsiveness on both loci are unlikely to be rescued by tirzepatide alone.

CagriSema (GLP-1 + amylin)

Amylin signals through calcitonin and amylin receptors, which are genetically independent of GLP1R and GIPR. In principle this means cagrilintide-semaglutide combinations open an entirely orthogonal satiety axis. The 23andMe study did not include CagriSema users in significant numbers, so this remains hypothesis-generating.

Retatrutide (GLP-1 + GIP + glucagon)

The triple agonist retatrutide adds glucagon receptor (GCGR) agonism on top of the dual incretin effect. In theory, patients with low-responsiveness on both GLP1R and GIPR could still benefit through GCGR-mediated energy expenditure, although no published GWAS yet stratifies retatrutide response by GCGR genotype.

The "third axis" hypothesis

Taken together, the data support a clinical research hypothesis that the more independent receptor axes a therapy engages, the smaller the fraction of patients who fail to respond. This is one of the strongest arguments for the GLP-1 plus GIP plus glucagon triple agonist class.

Can I Get Tested?

This is the question every patient and many researchers ask after reading press coverage of the Nature paper. The honest answer in April 2026 is: not in a clinically actionable way.

What 23andMe provides

23andMe holds genotype data for millions of customers and routinely returns reports on traits and ancestry. The company has hinted at a future "GLP-1 response" research report that surfaces relevant SNPs to consenting customers. As of this writing, that report has not been launched.

What clinicians can do today

No regulatory body has approved a clinical genetic test for GLP-1 response. The American Diabetes Association, the European Association for the Study of Diabetes and the Endocrine Society have not issued guidelines recommending pre-treatment GLP-1 pharmacogenomic testing. The effect sizes, while real at the population level, do not yet meet the threshold for individual-level clinical decision making.

What researchers can do

For preclinical and observational research, polygenic risk scores constructed from the publicly released summary statistics of the Nature paper can be used to stratify cohorts, predict response distributions and design trials enriched for predicted responders or non-responders.

What to Watch in 2026 and 2027

The 23andMe paper is the opening chapter, not the conclusion. Several follow-on studies and decisions are expected in the next 12 to 24 months.

Independent replication

Large national biobanks (UK Biobank, FinnGen, All of Us, Estonian Biobank) are well positioned to replicate the GLP1R and GIPR findings in independent cohorts. Replication is the prerequisite for any guideline change.

Receptor-level functional validation

In vitro studies of the variant receptors expressed in cell systems can confirm whether the alleles actually shift binding affinity, cAMP response or beta-arrestin recruitment. Without this, the GWAS associations remain statistical rather than mechanistic.

Label updates and pharmacogenomic flags

The FDA Pharmacogenomic Biomarker Table currently does not list GLP1R or GIPR for any GLP-1 agonist. A label update is unlikely before independent replication, but it is the eventual destination if the findings hold.

Trial design innovations

Phase 3 obesity and diabetes trials may begin to pre-specify pharmacogenomic subgroup analyses, similar to what oncology has done for years. This is the most likely near-term impact on the industry.

Conclusion: Genetics Closes Part of the Variance Gap

The 23andMe-led Nature study does not solve the GLP-1 non-responder problem, but it changes the framing. For the first time, the field has a population-scale answer to the question of why the same drug at the same dose produces such different outcomes, and that answer points at the receptors themselves.

Key findings:

27,885 self-reported GLP-1 and tirzepatide users analysed
GLP1R variants associated with weight loss spread of roughly 6 to 20 percent
GIPR variants independently associated with tirzepatide response
Nausea liability ranged from 5 percent to 78 percent and partly tracks the same loci

Open questions:

Independent replication in biobanks is still pending
Functional studies on the variant receptors are needed
No clinical genetic test is currently recommended
Effect of triple agonists in low-responsiveness genotypes is unknown

For researchers, the practical takeaway is that future GLP-1 trials and combination peptide designs should think explicitly about receptor-level variance, not just dose and adherence. For patients and clinicians, the message is patience: the science is real, but it is not yet a test you can order.

Sources

STAT News. "23andMe-led study links genetic variants to weight loss response on GLP-1 drugs." 8 April 2026. https://www.statnews.com/2026/04/08/23andme-genetic-changes-weight-loss-glp1/
Nature. Genome-wide association study of GLP-1 receptor agonist response in 27,885 individuals. Published online 8 April 2026.
FDA Table of Pharmacogenomic Biomarkers in Drug Labeling. Accessed April 2026.
American Diabetes Association. Standards of Care in Diabetes 2026.