In a nationwide cohort of 379,858 adults with obesity and type 2 diabetes (32,401 surgical, 347,457 medication), overlap weighting balanced the groups. Metabolic bariatric surgery was associated with lower five-year nephropathy (HR 0.55), retinopathy (HR 0.60), major adverse cardiovascular events (HR 0.78) and all-cause mortality (2.1% vs 3.7%; HR 0.61). One-year weight loss was 25.4% after surgery versus 6.4% with obesity medications, and 76.7% versus 47.2% of patients with baseline HbA1c ≥6.5% reached HbA1c below 6.5%.
The direction matches earlier surgery-versus-medication comparisons, but the scale here is what stands out: this is one of the largest head-to-head analyses in real-world practice. Several caveats matter. Only 6.4% weight loss on medication is well below what trials of semaglutide or tirzepatide show at full dose, which suggests many real-world patients stopped early, underdosed or switched agents, and that likely widens the gap. Observational data cannot rule out residual confounding, since patients who choose surgery may differ in motivation, health-seeking behaviour and access in ways claims data miss. Follow-up of roughly five years also predates widespread use of the most potent drugs. The practical message for adults with obesity and diabetes is that surgery remains the most effective option for weight loss, glycaemic control and complication prevention, and worth discussing even in the GLP-1 era. Confirmatory rather than paradigm-shifting, it still needs randomised trials against modern incretin dosing.