Relationship between frailty and all-cause mortality in patients with atrial fibrillation: a report from the ESC-EHRA EURObservational research programme AF general long-term registry

Abstract Introduction Frailty is a major health determinant for cardiovascular disease. Thus far, data on frailty in patients with atrial fibrillation (AF) are limited. Aims To evaluate frailty in a large contemporary cohort of European AF patients, the relationship with oral anticoagulant (OAC) prescription and with risk of all-cause death. Methods We analyzed patients enrolled in the ESC-EHRA EORP-AF General Long-Term Registry. A 38-items frailty index (FI) was derived from baseline characteristics according to the accumulation of deficits model proposed by Rockwood and Mitnitsky. All-cause mortality was the primary study outcome. Results Out of the 11096 AF enrolled patients, data for evaluating frailty were available for 6557 (59.1%) patients who have been included in this analysis (mean [SD] age 68.9 [11.5], 37.7% females). Baseline median [IQR] CHA2DS2-VASc and HAS-BLED were 3 [2–4] and 1 [1–2], respectively. At baseline, median [IQR] FI was 0.16 (0.12–0.23), with 1276 (19.5%) patients considered “not-frail” (FI<0.10), 4033 (61.5%) considered “pre-frail” (FI 0.10–0.25) and 1248 (19.0%) considered “frail” (FI≥0.25). Age, female prevalence, CHA2DS2-VASc and HAS-BLED progressively increased across the FI classes (all p<0.001). Use of OAC progressively increased among FI classes; after adjustments FI was not associated with OAC prescription (odds ratio [OR]: 1.09, 95% confidence interval [CI]: 0.98–1.19 for each 0.10 FI increase). Conversely, FI was directly associated with vitamin K antagonist (VKA) use (OR: 1.26, 95% CI: 1.18–1.34 for each 0.10 FI increase) and inversely associated with non-VKA OACs (NOACs) use (OR: 0.82, 95% CI: 0.77–0.88). FI was significantly correlated with CHA2DS2-VASc (Rho= 0.516, p<0.001). Over a median [IQR] follow-up of 731 [704–749] days, there were 569 (8.7%) all-cause death events. Kaplan-Meier curves [Figure] showed an increasing cumulative risk for all-cause death according to FI categories. A Cox multivariable analysis, adjusted for age, sex, type of AF and use of OAC, found that increasing FI as a continuous variable was associated with an increased risk of all-cause death (hazard ratio [HR]: 1.56, 95% CI: 1.40–1.73 for each 0.10 FI increase). An association with all-cause death risk was found across the FI categories (HR: 1.71, 95% CI: 1.23–2.38 and HR: 2.88, 95% CI: 2.02–4.12, respectively for pre-frail and frail patients compared to non-frail ones). FI was also predictive of all-cause death (c-index: 0.660, 95% CI: 0.637–0.682; p<0.001). Conclusions In a European contemporary cohort of AF patients the burden of frailty is significant, with almost 1 out of 5 patients found to be “frail”. Frailty influenced significantly the choice of OAC therapy and was associated with (and predictive of) all-cause death at follow-up. Kaplan-Meier Curves Funding Acknowledgement Type of funding source: Private company. Main funding source(s): Since the start of EORP programme, several companies have supported it with unrestricted grants.