METABO-PREDICT

Background

The obesity pandemic affects over 650 million adults worldwide, driving demand for personalized treatment strategies. GLP-1 receptor agonists (GLP-1-RA), including Semaglutide and the dual GIP/GLP-1 agonist Tirzepatide, represent a breakthrough in pharmacological obesity treatment. However, weight loss outcomes show extreme inter-individual variability, ranging from 5% to 25% of body weight. Currently, no validated biomarkers exist to predict treatment response before initiating therapy, leading to suboptimal outcomes and inefficient healthcare spending estimated at €200 billion annually in Europe alone. Identifying non-responders before treatment would enable personalized therapy selection and improve cost-effectiveness.

Research Question

METABO-PREDICT addresses this critical gap by testing whether baseline diet-induced thermogenesis (DIT), measured by the low-protein meal test (LPMT), can predict 12-month weight loss in patients treated with GLP-1 receptor agonists. Our central hypothesis is that individuals with LOW baseline DIT ("thrifty phenotype") will show LESS weight loss response to incretin-based therapy, identifying them as potential non-responders who may require alternative or intensified treatment strategies. This hypothesis is grounded in the concept that individuals with efficient energy conservation (low DIT) may be metabolically adapted to resist weight loss.

Methods

We will conduct a translational research program combining a clinical human study with mechanistic mouse experiments across five European countries (Germany, Spain, Netherlands, Belgium, France).

Human Study (n=120): A prospective, multicenter observational cohort across three centers (Kiel, Granada, Maastricht) will recruit 120 patients with obesity (BMI ≥30 kg/m² or ≥27 kg/m² with comorbidity), stratified by sex (60 women, 60 men) to enable sex-specific subanalysis. Patients with diabetes are excluded. Before initiating GLP-1-RA treatment (Semaglutide or Tirzepatide), all participants undergo comprehensive baseline phenotyping including the LPMT with indirect calorimetry (primary predictor), targeted blood biomarkers (FGF21, GDF15, 12,13-diHOME, FGF21 proteoforms), skin temperature buttons for BAT activation surrogate, high-fat meal challenge for metabolic flexibility, ECS profiling, VO2peak testing, and continuous activity monitoring via wearables. Follow-up visits at 3, 6, 9, and 12 months assess weight loss, body composition, and metabolic parameters.

Mouse Study (~200 mice): To elucidate mechanisms underlying the DIT-weight loss relationship, we will study multiple mouse strains at two centers (Antwerp, Toulouse). Outbred CD-1 mice provide genetic diversity mimicking human populations; inbred C57BL/6J and BALB/c strains compare obesity-prone vs obesity-resistant phenotypes; ob/ob (Leptin-KO) mice model severe genetic obesity; and UCP1-KO mice test whether BAT thermogenesis is required for DIT and GLP-1-RA efficacy. All strains undergo simulated LPMT with BAT temperature monitoring, followed by 8-week treatment with Semaglutide or Tirzepatide. Comprehensive tissue analyses (BAT, WAT, skeletal muscle biopsies, liver, plasma) will identify molecular pathways linking DIT to weight loss response.

Expected Outcomes

Primary Outcome: Correlation between baseline DIT (from LPMT) and percentage total body weight loss at 12 months in humans treated with GLP-1-RA. We hypothesize that low DIT will predict poor weight loss response.

Secondary Outcomes: Exploratory analyses of metabolic flexibility, ECS markers, FGF21/GDF15, and VO2peak as additional predictors; sex-stratified analysis (n=60 men vs n=60 women); mechanistic insights from mouse tissue analyses including UCP1-dependency of DIT; development of a predictive algorithm integrating multiple biomarkers for personalized treatment selection.