Association between the body mass index and risk of cardiovascular events in sodium-glucose cotransporter 2 inhibitor users compared with dipeptidyl-peptidase 4 inhibitor users: A nationwide cohort study in Korea

Aims: There is limited evidence regarding whether obesity modifies the association between the use of sodium-glucose cotransporter 2 inhibitors (SGLT2is) and the risk of cardiovascular events. We assessed whether baseline body mass index (BMI) modifies the association between SGLT2i use and the risk of major adverse cardiovascular events (MACE) and heart failure (HF) in patients with type 2 diabetes (T2D). Materials and Methods: We used the nationwide claims data of Korea (September 2014–December 2022) to construct an active-comparator, new-user cohort of patients with T2D stratified by the Asian BMI categories: normal weight, 18.5–23 kg/m²; overweight, 23–25 kg/m²; and obesity, ≥25 kg/m². New-users of SGLT2i were propensity score (PS)-matched with new-users of dipeptidyl peptidase 4 inhibitor (DPP4i) in a 1:1 ratio. The co-primary outcomes were 4-point MACE and hospitalization for HF (HHF). Patients were followed up using an as-treated exposure definition. PS-matched hazard ratios (HR) with 95% confidence intervals (CI) were estimated using the Cox model. Results: New-users of SGLT2i and DPP4i were PS-matched in a 1:1 ratio (n = 231 332 pairs; normal weight, 21 285 pairs; overweight, 35 372 pairs; and obesity, 174 675 pairs). The overall HR for the risk of MACE with SGLT2i versus DPP4i use was 0.90 (95% CI: 0.86–0.95), with no evidence of effect modification by baseline BMI (p for homogeneity = 0.27). The risk of HHF decreased in the overall cohort (HR: 0.53, 95% CI: 0.44–0.64), as well as in the obesity (HR: 0.47, 95% CI: 0.37–0.58) and overweight (HR: 0.49, 95% CI: 0.31–0.78) groups but not in the normal-weight (HR: 0.88, 95% CI: 0.59–1.31) group, with evidence of effect modification by the BMI (p for homogeneity = 0.01). Conclusions: The association between SGLT2i use and the risk of MACE and HHF was significant in patients with obesity. Baseline BMI was an effect modifier in the association between SGLT2i use and the risk of HHF, with a more pronounced association observed with increasing BMI and with no significant effect modification of the association noted in patients with normal weight.