RESUMO
AIMS: Impact of type 2 diabetes mellitus (T2DM) on non-thromboembolic outcomes in atrial fibrillation (AF) is insufficiently explored. This prospective cohort study of AF patients aimed (i) to analyse the association between T2DM and heart failure (HF) events (including new-onset HF), and all-cause and cardiovascular mortality, (ii) to assess the impact of baseline T2DM treatment on outcomes, and (iii) to explore characteristics of new-onset HF phenotypes in relation to T2DM status. METHODS AND RESULTS: Of 1803 AF patients (515/1288, with/without prior HF), 389 (22%) had T2DM at baseline. After 5 years of median follow-up, T2DM patients had an 85% greater risk of HF events [adjusted hazard ratio (aHR) 1.85; 95% confidence interval (CI) 1.51-2.28; P < 0.001], including a 45% increased risk for new-onset HF (1.45; 1.17-2.28; P = 0.015). T2DM conferred a 56% higher all-cause (1.56, 1.22-2.01; P = 0.003) and a 48% higher cardiovascular mortality (1.48; 1.34-1.93; P = 0.007). Fine-Gray analysis, with mortality as a competing risk, confirmed greater HF risk among T2DM patients. All risks were highest among insulin-treated patients. The prevalence of new-onset HF phenotypes was as follows: 67% preserved ejection fraction (HFpEF), 20% mid-range ejection fraction (HFmrEF) and 13% reduced ejection fraction (HFrEF). On time-dependent Cox regression, adjusted for baseline characteristics and an interim acute coronary event, T2DM increased aHRs for new-onset HFpEF (2.38; 1.30-4.58; P <0.001) and the combined HFmrEF/HFrEF (1.77; 1.11-3.62; P = 0.017). CONCLUSIONS: Atrial fibrillation patients with T2DM have independently increased risk of new-onset/recurrent HF events, cardiovascular and all-cause mortality, particularly when insulin-treated. The prevailing phenotype of new-onset HF was HFpEF; T2DM conferred higher risk of both HFpEF and HFmrEF/HFrEF.