Atrial fibrillation and heart failure often coexist and are closely intertwined, each condition worsening the other. However, the temporal relationships between these two conditions have not yet been fully explored. We aimed to assess, on a national scale, the prognosis of patients hospitalised with HF and AF based on the timing of the onset of the two conditions.

All consecutive patients with both AF and HF admitted to hospital between 2010 and 2018 were included, i.e. 1,237,927. Patients were identified in the French administrative hospital discharge database.

Overall, most patients developed HF before AF (n=860,937; 69.5%). At follow-up (median [IQR]: 1.4 [0.2–3.7] years), patients with HF first had an increased risk of all-cause death (yearly incidence: 19.9% vs. 10.1%; [RR (95% CI)]: [1.98 (1.96–1.99)]; P<0.0001), CV death (7.5% vs. 3.4%; [RR: 2.22 (2.19–2.24)]; P<0.0001), and non-CV death (12.4 vs. 6.7%/year; OR 1.85, 95% CI: 1.84–1.87; P<0.0001). By contrast, the difference in RR for ischemic stroke was less obvious between the 2 groups (2.3 vs. 2.50%/year; OR: 1.11, 95% CI: 1.09–1.12; P<0.0001). In the propensity score-matched population, patients with HF first also had worse outcomes than patients with AF first (all-cause death: 17.2% vs. 9.6% [RR: 1.79 (1.78–1.81)], P<0.0001; CV death rates: 6.3% vs. 3.3% [RR: 1.94 (1.91–1.97)]; P<0.0001). Surprisingly, the RR for ischemic stroke varied considerably between the 2 groups and higher in the AF first group (2 vs. 0.9%/year; OR: 0.42, 95% CI: 0.41–0.44; P<0.0001).

In this large study, we identified two distinct clinical entities based on the chronological sequence of AF and HF onset. Our results confirm that the prognosis for HF preceding AF was much worse than the opposite, but further studies are needed to investigate the thromboembolic results highlighted in this study.