Association of preoperative sodium-glucose cotransporter 2 (SGLT2) inhibitor use with postoperative delirium and long-term cognitive decline in older adults: a target trial emulation

Postoperative delirium is an important postoperative neurocognitive complication associated with adverse short- and long-term outcomes, particularly in older adults with type 2 diabetes. Current perioperative guidelines recommend withholding sodium-glucose cotransporter 2 (SGLT2) inhibitors 3–4 days before surgery to prevent euglycemic diabetic ketoacidosis. We emulated a target trial to investigate whether recent preoperative SGLT2 inhibitor use was associated with clinically coded postoperative delirium and subsequent cognitive diagnostic outcomes. Using the TriNetX global health research network, we identified adults aged ≥ 60 years with type 2 diabetes undergoing major operative procedures between January 2016 and December 2025. Applying a new-user design and 1:1 propensity score matching, we compared patients with recent preoperative prescriptions for SGLT2 inhibitors within 90 days preoperatively with matched non-users. The primary outcome was clinically coded postoperative delirium within 7 days. Secondary outcomes included 30-day mortality and long-term coded cognitive diagnostic outcomes. The matched cohort included 126,190 patients (63,095 per group). SGLT2 inhibitor use was associated with a lower risk of clinically coded postoperative delirium compared with non-use (3.98% vs. 4.53%; hazard ratio [HR] 0.87, 95% CI 0.83–0.92; P < .001). Exploratory surgical-category analyses suggested that the association was more pronounced in cardiovascular and respiratory procedures than in several other surgical categories, although these findings should be interpreted cautiously. Long-term follow-up (> 5 years) revealed lower risks of subsequent coded delirium and coded cognitive diagnostic outcomes among SGLT2 inhibitor users. Findings were consistent across several sensitivity analyses and were supported by negative control analyses, although residual confounding cannot be excluded. Preoperative SGLT2 inhibitor use was associated with lower risks of postoperative delirium and long-term cognitive outcomes in older adults with type 2 diabetes undergoing surgery. These findings identify a clinically important perioperative neurocognitive signal in a population increasingly exposed to SGLT2 inhibitors. Because residual confounding, exposure misclassification, and outcome underascertainment cannot be excluded, prospective studies with adjudicated exposure, standardized delirium assessment, and perioperative safety monitoring are needed. Postoperative delirium is a major complication in older adults with diabetes. In a target trial emulation of 126,190 patients, preoperative SGLT2 inhibitor use was associated with a lower risk of clinically coded postoperative delirium. The association appeared more pronounced in cardiovascular and respiratory procedures, but subgroup findings were exploratory. These observational findings are hypothesis-generating and do not support changes to current perioperative discontinuation recommendations without prospective validation.