In a TriNetX target trial emulation of lean adults with type 2 diabetes (BMI below 27 kg/m2), new GLP-1 receptor agonist users were matched 1:1 to DPP-4 inhibitor (n=7,389), SGLT2 inhibitor (n=8,969) or usual-care (n=8,610) groups. Coded heart failure was lower in all three comparisons (HR 0.71 versus DPP-4i; 0.56 versus SGLT2i). MACE was lower versus DPP-4i (HR 0.83) and SGLT2i (HR 0.81). All-cause mortality was lower versus DPP-4i (HR 0.61) but not versus SGLT2i (HR 0.89, CI 0.76-1.05). Gastroparesis was more frequent versus DPP-4i. Lean patients are poorly represented in the major GLP-1 cardiovascular trials, which enrolled mostly overweight or obese participants, so this fills a real evidence gap, particularly for East Asian-style lean diabetes. Treat the mortality signal with suspicion. A 39% mortality reduction versus DPP-4i exceeds anything seen in randomized GLP-1 trials, and the null result against SGLT2i suggests healthy-user bias and residual confounding, which the authors themselves acknowledge. Unintentional weight loss or sarcopenia in already-lean patients is also a legitimate clinical concern this database cannot address. The heart failure finding is consistent across comparators, though diagnosis-code outcomes in electronic records are noisy. The gastroparesis signal deserves attention in thin patients. Overall, this is hypothesis-generating and confirmatory in direction, not practice-changing. Lean adults with diabetes should discuss options with their clinician, as SGLT2 inhibitors remain a well-supported heart failure choice.