Two-Year Biomarker Trends and Kidney/Mortality Outcomes After Semaglutide Versus Phentermine–Topiramate and Naltrexone–Bupropion in People With Obesity and Without Diabetes

BACKGROUND: Laboratory changes after antiobesity drugs in routine care are not well-described. STUDY QUESTION: In adults with obesity and no diabetes, how do 2-year laboratory trends and major clinical outcomes compare after starting semaglutide versus naltrexone-bupropion or phentermine-topiramate? STUDY DESIGN: Retrospective cohort study in the TriNetX US Collaborative Network (2021-2025). New users were propensity-score matched 1:1 and followed for up to 24 months in an intention-to-treat framework. A prespecified 3-month landmark analysis and a global network replication were performed. MEASURES AND OUTCOMES: Lipids, HbA1c, ALT, AST, lipase, creatinine, and estimated glomerular filtration rate were assessed at baseline and at 6, 12, and 24 months. Outcomes were all-cause mortality, major adverse kidney events (MAKEs), major adverse liver outcomes, and acute pancreatitis. Cox models estimated hazard ratios (HRs); 24-month absolute risk differences (ARDs) were derived from Kaplan-Meier estimates. RESULTS: Matching yielded 11,484 semaglutide-naltrexone-bupropion pairs and 16,316 semaglutide-phentermine-topiramate pairs. Compared with each comparator, semaglutide users had lower total cholesterol and LDL-C across follow-up; triglycerides were lower versus naltrexone-bupropion. HDL-C was higher with phentermine-topiramate at 12-24 months. HbA1c was lower with semaglutide versus naltrexone-bupropion at 12 and 24 months and slightly higher versus phentermine-topiramate. Between-group differences in creatinine and estimated glomerular filtration rate were limited. Semaglutide use was associated with lower hazards of mortality (vs. naltrexone-bupropion: HR 0.51; ARD -0.4%; vs. phentermine-topiramate: HR 0.62; ARD -0.2%) and MAKE (vs. naltrexone-bupropion: HR 0.47; ARD -0.5%; vs. phentermine-topiramate: HR 0.48; ARD -0.4%). Major adverse liver outcome and pancreatitis estimates were near the null, and landmark results were similar. CONCLUSIONS: Over 2 years in routine care, semaglutide use was associated with lower atherogenic lipid levels and lower hazards of death and MAKEs than naltrexone-bupropion or phentermine-topiramate, while kidney laboratory differences were limited.