Abstract
BACKGROUND AND AIMS: Individuals with heart failure (HF), other forms of cardiovascular disease, or kidney disease are at increased risk for the development and adverse health effects of diabetes. As such, prevention or delay of diabetes is an important treatment priority in these groups. The aim of this meta-analysis was to determine the effect of sodium-glucose co-transporter 2 inhibitors (SGLT2i) on incident diabetes in HF across the spectrum of left ventricular ejection fraction (LVEF) and across the broader spectrum of cardiovascular or kidney disease. METHODS: First, the effects of dapagliflozin vs. placebo on new-onset diabetes were assessed in a pooled, participant-level analysis of the DAPA-HF and DELIVER trials. New-onset diabetes was defined as the new initiation of glucose-lowering therapy during follow-up, and time from randomization to new-onset diabetes was evaluated using Cox proportional hazards models. Second, PubMed and Embase were searched to identify large-scale randomized clinical outcomes trials (RCTs) comparing SGLT2i with placebo among adults with cardiovascular or kidney disease. A trial-level meta-analysis was then conducted to summarize the treatment effects of SGLT2i on the incidence of new-onset diabetes. RESULTS: In the pooled analysis of DAPA-HF and DELIVER including 5623 participants with HF but without diabetes at baseline, dapagliflozin reduced the incidence of new-onset diabetes by 33% [hazard ratio (HR), 0.67; 95% confidence interval (CI), .49-.91; P = .012] when compared with placebo. There was no evidence of heterogeneity across the spectrum of continuous LVEF or key subgroups. Among seven complementary RCTs including 17 855 participants with cardiovascular or kidney disease, SGLT2i reduced the of new-onset diabetes by 26% (HR, 0.74; 95% CI .65-.85; P < .001), with consistent effects across trials. CONCLUSIONS: SGLT2i reduced the incidence of new-onset diabetes among individuals with cardiovascular or kidney disease. These findings suggest that SGLT2i implementation may have an important ancillary benefit on prevention or delay of diabetes in these high-risk populations.