To investigate the effects of glucose-lowering agents on all-cause mortality, and cardiovascular and renal outcomes in adults with type 2 diabetes. A MEDLINE and EMBASE search was performed to identify randomized controlled trials, published up to 28 February 2022, with a follow-up ≥52 weeks, in which glucose-lowering drugs were compared with either placebo or active comparators. We included only trials reporting formal external adjudication of events. All-cause mortality, 3-point MACE (major cardiovascular events), and hospitalization for heart failure (HHF) were considered as principal outcomes. Doubling of serum creatinine, worsening albuminuria, and renal death were considered as secondary endpoints. We included randomized controlled trials performed on metformin (n=17), pioglitazone (n=20), alpha-glucosidase inhibitors (n=9), insulin secretagogues (n=42), dipeptidyl-peptidase-4 inhibitors (n=67), glucagon-like peptide-1 receptor agonists (n=45) or sodium-glucose co-transporter-2 inhibitors (SGLT-2i; n=42) and insulin (n=18). Glucagon-like peptide-1 receptor agonist and SGLT-2i were associated with a significant reduction in all-cause mortality [Mantel-Haenszel odds ratio (MH-OR), 95% confidence interval: 0.88 (0.83; 0.95) and 0.85 (0.79; 0.91), respectively] and MACE [MH-OR, 95% confidence interval: 0.89 (0.84; 0.94) and 0.90 (0.84; 0.96), respectively]. SGLT-2i was associated with a reduced risk of HHF [MH-OR 0.68 (0.62; 0.75)], worsening albuminuria [MH-OR 0.67 (0.55; 0.80)] and doubling of serum creatinine [MH-OR 0.58 (0.44; 0.79)]. Metformin and pioglitazone were associated with a significantly lower risk of MACE [MH-OR 0.60 (0.47; 0.80) and 0.85 (0.74; 0.97), respectively] and pioglitazone with a higher risk of HHF [MH-OR 1.30 (1.04; 1.62)]. Insulin secretagogues were associated with increased risk of all-cause mortality [MH-OR 1.12 (1.01; 1.24)] and MACE [MH-OR 1.19 (1.02; 1.39)]. The results of this updated meta-analysis need to be considered in the choice of drug treatment for type 2 diabetes mellitus, which cannot be merely based on the effect of glucose-lowering drugs on long-term glycaemic control.
Read full abstract