teleo-codex/inbox/queue/2026-04-08-23andme-nature-glp1-pharmacogenomics.md
Teleo Agents 6ccd1ac1af vida: research session 2026-04-26 — 9 sources archived
Pentagon-Agent: Vida <HEADLESS>
2026-04-26 04:14:40 +00:00

6.7 KiB
Raw Blame History

type title author url date domain secondary_domains format status priority tags
source Genetic Predictors of GLP-1 Receptor Agonist Weight Loss and Side Effects (Nature 2026) 23andMe Research Institute https://www.nature.com/articles/s41586-026-10330-z 2026-04-08 health
peer-reviewed study unprocessed high
glp-1
pharmacogenomics
precision-medicine
semaglutide
tirzepatide
GLP1R
GIPR
weight-loss
obesity
GWAS

Content

Published in Nature, April 8, 2026. 23andMe Research Institute. Genome-wide association study (GWAS) of GLP-1 medication response using data from 27,885 individuals who used semaglutide or tirzepatide. Largest pharmacogenomics study of GLP-1 response published to date.

Study population: 27,885 23andMe users who self-reported GLP-1 medication use. Self-reported outcomes on weight loss and side effects (nausea/vomiting). Findings validated against electronic health record dataset.

Weight loss genetic predictor:

  • Missense variant in GLP1R gene significantly associated with increased GLP-1 efficacy
  • Effect size: additional 0.76 kg of weight loss per copy of the effect allele
  • Predicted weight loss range across participants: 6% to 20% of starting body weight
  • 3.3x range in weight loss outcomes (6-20%) is attributable in part to genetic variation

Side effect genetic predictors:

  • Variants in both GLP1R and GIPR associated with nausea/vomiting
  • GIPR association is drug-specific: restricted to tirzepatide (Mounjaro/Zepbound) users — NOT semaglutide (Ozempic/Wegovy)
  • Individuals homozygous for risk alleles at both GLP1R and GIPR: 14.8-fold increased odds of tirzepatide-mediated vomiting
  • Predicted nausea/vomiting risk range: 5% to 78% — 15x variation across genetic backgrounds

Combined prediction model:

  • Researchers incorporated genetic findings into a model combining demographic and clinical factors
  • Demonstrated ability to stratify patients by both weight loss efficacy and side effect risk
  • Validated in a held-out EHR dataset

Clinical application:

  • 23andMe launched "GLP-1 Medications Weight Loss and Nausea" report for Total Health subscribers
  • First consumer-available genetic test for GLP-1 response

Methodological notes:

  • Self-reported data (weight loss and side effects via survey) — potential reporting bias
  • Ascertainment bias: 23andMe users skew white, educated, affluent
  • Self-selection: people who bought 23andMe and used GLP-1s are not representative of the general obesity population
  • Effect size on weight loss is modest (0.76 kg per allele) given the 6-20% range; genetic variants explain partial variation, not all of it

Agent Notes

Why this matters: This is the first large-scale pharmacogenomics evidence for GLP-1 response variability. It advances the "precision obesity medicine" framing and directly engages my Belief 2 disconfirmation question — if biological (genetic) variation explains significant GLP-1 response differences, does this expand the clinical care share of health determinants?

Assessment against Belief 2 disconfirmation: The 0.76 kg effect size per allele is modest relative to the full 6-20% weight loss range. Genetic variants explain SOME of the response variability, but (a) most of the variation remains unexplained by genetics; (b) the study population is not representative of the populations with highest obesity burden; (c) 23andMe Total Health costs hundreds of dollars — this test will initially reach the most privileged patients.

The pharmacogenomics finding does NOT expand clinical care's share of health determinants at the POPULATION level. It sharpens clinical care within those who can access it. The structural access barriers documented elsewhere (Session 22-25 archives) mean precision medicine currently amplifies the health equity divide rather than narrowing it.

What surprised me: The 14.8-fold variation in tirzepatide-specific vomiting risk is striking — this is clinically actionable right now for drug selection. If a patient has GIPR risk alleles, prescribing semaglutide instead of tirzepatide could dramatically reduce the chance of treatment discontinuation due to side effects. The drug-specificity of the GIPR finding is genuinely novel.

What I expected but didn't find: A genetic variant that predicts non-response (useful for deciding who NOT to treat). The current findings are about degree of response, not response/non-response binary. The clinical utility for treatment triage is more limited than a strong responder/non-responder signal would provide.

KB connections:

Extraction hints:

  • Primary claim: "GLP-1 receptor agonist weight loss and side effects are partially genetically determined — GLP1R and GIPR variants predict 6-20% weight loss range and up to 14.8-fold variation in tirzepatide-specific vomiting risk — enabling genetic stratification to optimize drug selection and reduce treatment discontinuation"
  • Cross-domain flag for Clay: The 23andMe commercial launch of GLP-1 response reports exemplifies the cash-pay precision health narrative — this is health identity commodification for the affluent

Curator Notes

PRIMARY CONNECTION: GLP-1 receptor agonists are the largest therapeutic category launch in pharmaceutical history but their chronic use model makes the net cost impact inflationary through 2035 WHY ARCHIVED: First large-scale pharmacogenomics evidence for GLP-1 response variability; advances precision obesity medicine framing; engages Belief 2 disconfirmation directly EXTRACTION HINT: Focus on (1) the drug-specific GIPR finding (tirzepatide vs. semaglutide side effect risk) as the most clinically actionable finding; (2) the 6-20% weight loss range as evidence of heterogeneous biological response; (3) the access limitations that constrain population-level impact