Glucagon-like peptide-1 receptor agonists (GLP-1RAs), widely used for managing type 2 diabetes (T2D) and weight, are gaining attention for treatment of a broad set of conditions. Large-scale, real-world evidence of their broader clinical impact is needed. We assessed the phenome-wide risks and benefits of new users of GLP-1RA versus sulfonylureas, SGLT-2 inhibitors (SGLT-2i), and DPP-4 inhibitors (DPP-4i) in adults with T2D in a multi- healthcare center, 1:1 high-dimensional propensity score matched new user cohort study utilizing principles of target trial emulation with electronic health records (EHRs) of over 9 million patients followed up to 730 days. Primary outcomes included 239 clinical endpoints across 14 organ systems. Among 86,790 patients, mean age was 58-66 years and 44-62% were female across cohorts. Compared to DPP-4i, GLP-1RA use was associated with a reduced risk of acute myocardial infarction (sHR 0.61, 95% CI 0.43–0.85) and chronic kidney disease (sHR 0.71, 95% CI 0.62–0.81). Compared to sulfonylurea, GLP-1RA use was associated with reduced acute renal failure risk (sHR 0.77, 95% CI 0.65–0.90). GLP-1RA use was associated with reduced heart failure risk compared to SGLT-2i (sHR 0.66, 95% CI 0.55–0.80). GLP-1RA also was associated with lower epilepsy risk versus DPP-4i (sHR 0.49, 95% CI 0.32–0.76). GLP1RA was associated with elevated risks of nausea/vomiting: sHR 1.37 vs SGLT2i, 95% CI 1.15–1.62; sHR 1.46 vs sulfonylureas, 95% CI 1.24–1.71). Head-to-head real-world comparisons with established T2D therapies confirmed the broad cardiorenal and metabolic benefits and known on-target adverse effects of GLP-1RAs –consistent with randomized clinical trials - but also suggest potential risks for musculoskeletal and genitourinary adverse events, warranting continued real-world post-market surveillance.

RV, AP, LD, PM, DG, RF, CH, MS have no conflict of interests. KS is a consultant for Google and serves on their consumer health advisory panel. AB reported serving as a cofounder and consultant to Personalis and NuMedii; serving as a consultant to Mango Tree Corporation, Samsung, 10x Genomics, Helix, Pathway Genomics, and Verinata (Illumina); serving on paid advisory panels or boards for Geisinger Health, Regenstrief Institute, Gerson Lehman Group, AlphaSights, Covance, Novartis, Genentech, and Merck, and Roche; owning stock in Personalis, NuMedii, Apple, Meta, Alphabet, Microsoft, Amazon, Snap, 10x Genomics, Illumina, Regeneron, Sanofi, Pfizer, Royalty Pharma, Moderna, Sutro, Doximity, BioNtech, Invitae, Pacific Biosciences, Editas Medicine, Nuna Health, Assay Depot, and Vet24seven; receiving personal fees from Johnson and Johnson, Roche, Genentech, Pfizer, Merck, Lilly, Takeda, Varian, Mars, Siemens, Optum, Abbott, Celgene, AstraZeneca, AbbVie, Westat, Stanford University, NuMedii, Personalis, Dartmouth University, Boston Children's Hospital, Johns Hopkins University, Endocrine Society, Alliance for Academic Internal Medicine, Children's Hospital of Philadelphia, University of Pittsburgh Medical Center, Cleveland Clinic, University of Utah, Society of Toxicology, Mayo Clinic, Washington University in Saint Louis, and University of Michigan; receiving grants from the National Institutes of Health, Peraton, Genentech, Johnson and Johnson, US Food and Drug Administration (FDA), Robert Wood Johnson Foundation, Leon Lowenstein Foundation, Intervalien Foundation, Priscilla Chan and Mark Zuckerberg, the Barbara and Gerson Bakar Foundation, March of Dimes, Juvenile Diabetes Research Foundation, California Governor's Office of Planning and Research, California Institute for Regenerative Medicine, L'Oreal, and Progenity outside the submitted work. No other disclosures were reported.

This research received primary support from the Clinical and Translational Science Institute at the University of California, San Francisco (UCSF), under award number UL1TR001872. Additional funding was partially provided by the U.S. Food and Drug Administration (FDA) through grant number U01FD005978, awarded to the UCSF-Stanford Center of Excellence in Regulatory Science and Innovation. Further support was contributed by the UCSF Bakar Computational Health Sciences Institute and the University of California Health Center for Data-driven Insights and Innovation.

I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.

Yes

The details of the IRB/oversight body that provided approval or exemption for the research described are given below:

The Institutional Review Boards across the University of California Health system determined that the research use of the HIPAA limited data set for this cohort study did not constitute human participants research and was therefore exempt from further approval and informed consent.

I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.

Yes

I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).

Yes

I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.

Yes

The data used for this analysis was derived from the de-identified electronic health records of patients receiving care across the University of California Health. Although de-identified, the individual-level nature of the data used risks individuals being identified, or being able to self-identify, if the data are released publicly.