Cardiovascular safety of glucose‐lowering agents as add‐on medication to metformin treatment in type 2 diabetes: report from the S wedish N ational D iabetes R egister

Aim To investigate the relative safety of various glucose‐lowering agents as add‐on medication to metformin in type 2 diabetes in an observational study linking five national health registers. Research design and methods Patients with type 2 diabetes who had been on metformin monotherapy and started another agent in addition to metformin were eligible for inclusion. The study period was 2005‐2012. Adjusted hazard ratios ( HRs ) with 95% confidence intervals ( CIs ) of mortality, cardiovascular disease ( CVD ), coronary heart disease ( CHD ), stroke and congestive heart failure ( CHF ) were estimated using C ox proportional hazards models, weighted for a propensity score. Results Of the 20 422 patients included in the study, 43% started on second‐line treatment with sulphonylurea ( SU ), 21% basal insulin, 12% thiazolidinedione ( TZD ), 11% meglitinide, 10% dipeptidyl peptidase‐4 ( DPP ‐4) inhibitor, 1% glucagon‐like peptide‐1 ( GLP ‐1) receptor agonist and 1% acarbose. At the index date, the mean patient age was ~60 years for all groups except the GLP ‐1 receptor agonist (56.0 years) and SU (62.9 years) groups. Diabetes duration and glycated haemoglobin levels were similar in all groups. When compared with SU , basal insulin was associated with an 18% higher risk and TZD with a 24% lower risk of mortality [ HR 1.18 (95% CI 1.03‐1.36) and 0.76 (95% CI 0.62‐0.94)], respectively. DPP ‐4 inhibitor treatment was associated with significantly lower risks of CVD , fatal CVD , CHD , fatal CHD and CHF . Conclusions This nationwide observational study showed that second‐line treatment with TZD and DPP ‐4 inhibitor as add‐on medication to metformin were associated with significantly lower risks of mortality and cardiovascular events compared with SU , whereas basal insulin was associated with a higher risk of mortality.