Scientists Identify Hidden Biological Reasons Why Ozempic Doesn’t Work for Everyone

Understanding Ozempic and Its Mechanism of Action

Ozempic, whose active ingredient is semaglutide, belongs to a class of drugs known as glucagon-like peptide-1 receptor agonists (GLP-1RAs). These mimic the GLP-1 hormone, naturally released in the gut after eating, which signals the brain to reduce appetite, slows gastric emptying, and boosts insulin secretion while curbing glucagon release. This multi-faceted action leads to better blood sugar control and substantial weight reduction, with pivotal trials like STEP demonstrating 15-20% body weight loss in non-diabetic obese individuals over 68 weeks.

However, real-world data reveals stark variability. In Novo Nordisk's semaglutide trials, about 23% of participants lost less than 5% of their body weight after three months, qualifying as non-responders. Factors like dosage, adherence, and lifestyle play roles, but emerging evidence points to innate biological differences as the core issue.

The Scale of Ozempic Efficacy Variability in Clinical Practice

Average outcomes mask individual disparities. Self-reported data from over 27,000 GLP-1 users showed median BMI reductions of 4.1 kg/m² (roughly 11.7% body weight) after 8 months, yet ranges spanned from negligible to over 20% loss. Electronic health records corroborate this, with self-reports often overstating losses but confirming genetic influences. Non-response rates hover at 20-32%, higher in certain ancestries like African Americans, underscoring the need for biological predictors.

Patient stories abound: some shed dozens of pounds effortlessly, while others gain or plateau despite compliance. This inconsistency has fueled academic scrutiny, with universities racing to decode the 'why' behind these failures.

Groundbreaking Genetic Discoveries: The GLP1R Variant Unveiled

A landmark study published in Nature in April 2026 by researchers from the 23andMe Research Institute, in collaboration with Icahn School of Medicine at Mount Sinai and Genentech scientists, analyzed 27,885 GLP-1 users. They identified a missense variant in the GLP1R gene, rs10305420 (p.Pro7Leu), significantly boosting efficacy. Carriers of the effect allele (T) experienced an extra 0.76 kg weight loss per copy, translating to 0.64% more BMI reduction.

This variant, in the receptor's signal peptide, likely enhances cell surface expression by improving hydrophobic stability, amplifying semaglutide's signaling without altering ligand binding. The effect held across ancestries and drugs, though slightly stronger with tirzepatide. A predictive model incorporating this genetics explained 25% of response variance, far surpassing non-genetic factors alone at 21%.Read the full Nature study.

Microbiome's Role: University of Tartu Predicts Response from Baseline Gut Flora

Complementing genetics, a January 2026 Scientific Reports paper from the University of Tartu's Estonian Genome Centre examined 20 type 2 diabetes patients starting semaglutide. Baseline fecal microbiomes, sequenced via 16S rRNA, predicted HbA1c reductions, with genera like Alistipes linked to better glycemic control and reduced inflammation markers such as neutrophils.

Post-treatment, microbial shifts occurred indirectly via metabolic improvements, not acutely. Higher Alistipes abundance correlated with greater neutrophil drops, suggesting microbiome modulation of inflammation aids semaglutide's effects. This positions gut bacteria as a pre-treatment biomarker, enabling responder stratification.Access the University of Tartu microbiome research.

Phenotypic Insights from Leading U.S. Institutions

Mayo Clinic researchers Andres Acosta and colleagues classified non-responders into 'hungry brain' (high-calorie needs for satiety) and 'hungry gut' (rapid gastric emptying) phenotypes, genetically influenced. These predict poor liraglutide response but better alternatives. Cleveland Clinic studies link neurobeachin gene scores to 50% higher non-response risk via hypothalamic signaling disruptions.

Harvard's Andrea Ganna notes multifactorial causes, while University of Copenhagen's Ruth Loos emphasizes obesity's polygenic nature—thousands of variants contributing incrementally.

Interplay of Genetics, Microbiome, and Environment

Biological reasons converge: GLP1R variants alter receptor function, microbiomes influence drug metabolism and inflammation, phenotypes reflect brain-gut axis wiring. Environment amplifies—diet shapes microbiota, ancestry affects allele frequencies. A Japanese study of 92 diabetics found external cue-driven overeaters responded best, hinting at behavioral-genetic links.

Clinical Implications: Toward Personalized Obesity Therapies

These findings herald pharmacogenomics. Genetic testing could guide semaglutide prescriptions, reserving it for high-responder profiles while directing others to alternatives like tirzepatide (dual GLP-1/GIP). Microbiome profiling offers non-invasive prediction, potentially via stool tests.

Experts like University of Cambridge's Marie Spreckley caution effects are modest clinically, but combined models stratify risks effectively. Side effects like nausea also tie to GLP1R/GIPR variants, aiding tolerance predictions.

Challenges in Implementation and Ongoing Academic Trials

Barriers include testing costs, diversity gaps (most data European ancestry), and integration into guidelines. Universities lead: ongoing trials at Mount Sinai explore GLP1R polygenic scores; Tartu expands microbiome cohorts.

Real-world validation via UK Biobank and All of Us enhances generalizability.

Photo by Artfox Photography on Unsplash

Future Outlook: University Research Paving Precision Medicine

Horizons include CRISPR-edited microbiomes, multi-omics integration, and AI-driven predictors. Projections: by 2030, 50% of prescriptions genetically guided, slashing non-responder rates. Academic collaborations promise equitable access, addressing global obesity epidemics.

For patients, actionable steps: consult providers on genetic/microbiome tests, combine drugs with diet/exercise. Research underscores no one-size-fits-all—biology demands customization.