Consistent excess mortality risk after heart failure onset in patients with atrial fibrillation, survival trends over 30 years of follow up, a nationwide registry-based study

Abstract

Abstract Background Atrial fibrillation and heart failure often coexist which is strongly associated with worse outcomes. Overall, the management and prognosis of heart failure has improved over time, however, to which extent this applies to low-risk patients with AF who were subsequently diagnosed with heart failure remains unclear. Purpose To investigate temporal survival trend of patients with AF and without other cardiovascular risk factors after HF onset from 1987 to 2017. Methods All patients from the Swedish National Patient Register, with a first-time diagnosis of AF between 1987 and 2017 were identified and compared with two matched controls without AF from the Total Population Register. Patients < 18 years, or any previous cardiovascular disease, diabetes mellitus and renal failure at baseline were excluded, along with their matched controls. The observation time was divided into two periods: 1987-2002 and 2003-2017. The study population was divided in following age categories: 18-34, 35-54, 55-54 and >75 years. Hazard ratios for survival were estimated using Cox regression models. Results In total we identified 227,811 patients with AF defined as low risk and 452,712 controls, 55.5% men, mean age (SD) in men 65.5 (15.0) and in women 72.7 (13.0) years. Mean follow up time for patients and controls was 8.7 (6.8) years. Between 1987 and 2002 mortality during the first year after onset in patients who were diagnosed with HF was 27.1% in AF patients while patients with AF who remained free from heart failure had a corresponding mortality of 10.6%. All-cause mortality rates within one year for controls were significantly lower in all groups. Compared with 1987-2002, one-year mortality in 2003-2017 decreased for in those with AF and HF (17.5%) as well as in those with AF who remained free of HF (7.5%). Compared to controls the relative mortality hazard within one year was 4.78 (95% CI 4.39-5.20) during the period 1987-2002 for patients with AF and HF-onset and 2.56 (95%CI 2.48-2.64) for AF patients without HF. HRs during 2003-2017 HR for 1-year all-cause mortality were 4.38 (95% CI 4.00-4.79) and 2.96 (95% CI 2.87-3.06), respectively. The highest hazard associated with HF in patients with AF 1987-2002 was found among patients 18-34 years [HR 13.87 (95%CI: 1.71-112.72) that decreased to HR 5.57 (95%CI: 2.01-15.47) 2003-2017. The mortality hazard decreased with increasing age in both periods, more in the later period. However, compared to controls in Cox regression adjusted for age and sex, patients who developed HF had HR 4.92 (95% CI:4.52-5.36) between 1987-2002 that declined only marginally in the later period to HR 4.51 (95%CI:4.12-4.93). Conclusion In low-risk patients with AF, the addition of HF raises 1-year mortality substantially. The 1-year mortality rates declined between the two periods but excess risk in patients with AF and HF compared to controls remained almost constant four times higher in patients with AF and HF than in controls.