Epidemiology and impact of frailty in patients with atrial fibrillation in Europe

Abstract

Background

Frailty is a medical syndrome characterised by reduced physiological reserve and increased vulnerability to stressors. Data regarding the relationship between frailty and atrial fibrillation (AF) are still inconsistent.

Objectives

We aim to perform a comprehensive evaluation of frailty in a large European cohort of AF patients.

Methods

A 40-item frailty index (FI) was built according to the accumulation of deficits model in the AF patients enrolled in the ESC-EHRA EORP-AF General Long-Term Registry. Association of baseline characteristics, clinical management, quality of life, healthcare resources use and risk of outcomes with frailty was examined.

Results

Among 10,177 patients [mean age (standard deviation) 69.0 (11.4) years, 4,103 (40.3%) females], 6,066 (59.6%) were pre-frail and 2,172 (21.3%) were frail, whereas only 1,939 (19.1%) were considered robust. Baseline thromboembolic and bleeding risks were independently associated with increasing FI. Frail patients with AF were less likely to be treated with oral anticoagulants (OACs) (odds ratio 0.70, 95% confidence interval 0.55–0.89), especially with non-vitamin K antagonist OACs and managed with a rhythm control strategy, compared with robust patients. Increasing frailty was associated with a higher risk for all outcomes examined, with a non-linear exponential relationship. The use of OAC was associated with a lower risk of outcomes, except in patients with very/extremely high frailty.

Conclusions

In this large cohort of AF patients, there was a high burden of frailty, influencing clinical management and risk of adverse outcomes. The clinical benefit of OAC is maintained in patients with high frailty, but not in very high/extremely frail ones.

atrial fibrillation, frailty, chronicity, oral anticoagulant therapy, outcomes, epidemiology, older people

Key Points

• Data on the relationship between frailty and atrial fibrillation (AF) are scarce. We assessed the epidemiology and impact of frailty, evaluated through a 40-item frailty index (FI), in the contemporary ESC-EHRA EORP-AF General Long-Term Registry.

• Among 10,177 AF patients, 2,172 (21.3%) were frail, and a total of 80% of patients showed a relevant burden of frailty.

• Thromboembolic and bleeding risks were independently associated with increasing FI, and frail patients were also less likely treated with oral anticoagulants (OACs) and with a rhythm control strategy.

• Frailty was associated with a higher risk for all outcomes examined, with a non-linear exponential relationship. OACs reduced the risk of outcomes, except in patients with very/extremely high frailty.

Introduction

Frailty is a medical syndrome characterised by a reduced physiologic function, which increases vulnerability to endogenous and exogenous stressors and the risk of adverse outcomes (including dependency and death) [1]. Frailty may indeed serve as a surrogate for measuring the biological complexity of individuals [2].

In the light of progressive population aging, frailty has rapidly become an emergent public health priority, demanding specific interventions and strategies [3]. While being initially ‘confined’ to geriatric medicine, awareness of frailty has increased in other clinical specialties, including cardiovascular medicine [4–6]. Measuring such a complex phenomenon as frailty poses significant challenges, with several models that have been proposed to identify and evaluate frailty [7]. Among these, the frailty index (FI), proposed by Rockwood and Mitnitski, was designed to capture the accumulation of health deficits over time, to provide an alternative to chronological age [8].

It is well recognised that atrial fibrillation (AF) is closely related to increasing age, multimorbidity and clinical complexity [9–11]. Notwithstanding this, the evidence regarding frailty evaluation in the context of AF is still limited [12].

We aimed to report the epidemiology of frailty in a large European cohort of AF patients and describe its impact on the clinical management and outcomes of these patients.

Methods

The European Society of Cardiology (ESC)—European Heart Rhythm Association (EHRA) EURObservational Research Programme (EORP) Atrial Fibrillation General Long-Term Registry is a prospective multicentre observational registry held by the ESC and endorsed by the EHRA. The study enrolled consecutive AF patients presenting in 250 cardiology practices in 27 participating countries, both in- and outpatient settings. A detailed description of the study design, baseline characteristics and 1-year follow-up results have been provided elsewhere [13–15].

All participants were adults ≥18 years, had AF electrocardiographically documented within 12 months before enrolment and provided written informed consent. Enrolment was undertaken from October 2013 to September 2016, with planned 1-year and 2-year follow-up. The institutional review boards approved the study protocol at each participating centre, according to the EU Note for Guidance on Good Clinical Practice CPMP/ECH/135/95 and the Declaration of Helsinki.

Evaluation of frailty

Frailty was assessed according to a 40-item FI (Supplemental Table 1), built on the cumulative deficits model, as proposed by Rockwood and Mitnitski [8, 16] and according to the standardisation principles described by Searle et al. [17]. FI was computed based on a multidimensional evaluation, including patients’ vital signs, comorbidities, symptoms, biomarkers and functions. For each patients, the FI was calculated as the ratio of the total deficits, and the total number of deficits included in the index (i.e. 40). According to the usual standards, an FI ranging from 0.10 to <0.25 defined the presence of pre-frailty, whereas an FI ≥0.25 denoted the presence of frailty [18].

Details regarding definitions of baseline variables, evaluation of healthcare resources use, quality of life and outcomes and the statistical analysis are reported in the Supplemental Methods.

Results

Among the original 11,096 patients enrolled in the ESC-EHRA EORP-AF General Long-Term Registry, a total of 10,177 (91.7%) had available data to evaluate the FI at baseline. Mean age [standard deviation (SD)] was 69.0 (11.4) years, 4,103 (40.3%) were women. At baseline, median [interquartile range (IQR)] CHA₂DS₂-VASc and HAS-BLED scores were 3 [2–4] and 1 [1, 2], respectively; EHRA score was ≥2 in 5,606 (55.1%) patients.

At baseline, mean (SD) FI was 0.18 (0.09), with a median [IQR] of 0.17 [0.11–0.23]. Accordingly, 6,066 (59.6%) were pre-frail, and 2,172 (21.3%) were frail. The distribution of the overall cohort according to FI is shown in Figure 1.

Figure 1

Distribution of FI in the ESC-EHRA EORP-AF General Long-Term Registry Cohort.

Baseline characteristics according to FI categories are reported in Table 1. For higher levels of the FI, AF patients were more likely to be older and women, present low socio-economic status, live alone and have sedentary behaviour. The prevalence of cardiovascular risk factors, comorbidities and polypharmacy was higher among pre-frail and frail subjects.

Table 1

Baseline characteristics according to frailty classes

N = 10,177	Robust, N = 1939	Pre-Frail, N = 6,066	Frail, N = 2,172	P
Socio-demographic characteristics
Age, years median [IQR]	65 [56–74]	71 [63–77]	73 [66–79]	<0.001
Female, n (%)	603 (31.1)	2,426 (40.0)	1,074 (49.4)	<0.001
European Region, n (%) Northern Europe Western Europe Eastern Europe Southern Europe	368 (19.0) 647 (33.4) 172 (8.9) 752 (38.8)	819 (13.5) 2,152 (35.5) 894 (14.7) 2,201 (36.3)	209 (9.6) 489 (22.5) 618 (28.5) 856 (39.4)	<0.001
Low socio-economic status, n (%) 8,079	721 (47.2)	2,410 (51.4)	1,202 (64.6)	<0.001
Domestic status, n (%) 8,653 Living alone Living with partner/family	240 (14.2) 1,449 (85.8)	885 (17.3) 4,231 (82.7)	359 (19.4) 1,489 (80.6)	<0.001
Physical activity, n (%) 8,862 None/occasional Regular/intense	1,014 (60.2) 671 (39.8)	3,905 (75.1) 1,296 (24.9)	1728 (87.4) 248 (12.6)	<0.001
Clinical characteristics and comorbidities
Site of inclusion, n (%) Outpatient facility Hospital	1,188 (61.3) 751 (38.7)	3,038 (50.1) 3,028 (49.9)	665 (30.6) 1,507 (69.4)	<0.001
Reason for admission, n (%) Other than AF AF	390 (20.1) 1,548 (79.9)	2002 (33.0) 4,064 (67.0)	1,041 (47.9) 1,131 (52.1)	<0.001
BMI, kg/m2 median [IQR]	26.2 [24.0–28.9]	27.7 [24.9–31.2]	28.9 [25.5–32-7]	<0.001
SBP, mmHg median [IQR]	125 [119–134]	130 [120–143]	140 [120–150]	<0.001
DBP, mmHg median [IQR]	80 [70–80]	80 [70–88]	80 [70–90]	<0.001
AF classification, n (%) First detected Paroxysmal Persistent LT persistent Permanent Unknown	346 (17.8) 632 (32.6) 413 (21.3) 64 (3.3) 441 (22.7) 43 (2.2)	957 (15.8) 1,528 (25.2) 1,191 (19.6) 258 (4.3) 2025 (33.4) 105 (1.7)	325 (15.0) 504 (23.2) 397 (18.3) 114 (5.3) 811 (37.4) 19 (0.9)	<0.001
Heart failure, n (%)	157 (8.1)	2,143 (35.3)	1,586 (73.0)	<0.001
Coronary artery disease, n (%)	154 (7.9)	1,669 (27.5)	1,025 (47.2)	<0.001
Hypertension, n (%)	650 (33.5)	3,877 (63.9)	1742 (80.2)	<0.001
Diabetes mellitus, n (%)	95 (4.9)	1,293 (21.3)	949 (43.7)	<0.001
Lipid disorder, n (%)	322 (16.6)	2,485 (41.0)	1,247 (57.4)	<0.001
Previous TE events, n (%)	92 (4.7)	674 (11.1)	409 (18.8)	<0.001
Previous haemorrhagic events, n (%)	28 (1.4)	291 (4.8)	219 (10.1)	<0.001
PAD, n (%)	15 (0.8)	402 (6.6)	386 (17.8)	<0.001
CKD, n (%)	37 (1.9)	520 (8.6)	664 (30.7)	<0.001
COPD, n (%)	34 (1.8)	466 (7.7)	402 (18.5)	<0.001
Anaemia, n (%)	2 (0.1)	198 (3.3)	336 (15.5)	<0.001
Predisposition to bleeding, n (%)	9 (0.5)	81 (1.3)	114 (5.3)	<0.001
Dementia, n (%)	2 (0.1)	41 (0.7)	72 (3.3)	<0.001
Malignancy, n (%)	67 (3.5)	452 (7.5)	242 (11.1)	<0.001
CHA2DS2-VASc, median [IQR]	2 [1–3]	3 [2–4]	4 [3–5]	<0.001
High TE Risk, n (%)	868 (44.8)	4,722 (77.9)	2051 (94.5)	<0.001
HAS-BLED, median [IQR]	1 [0–2]	1 [1–2]	2 [1–3]	<0.001
High bleeding risk, n (%)	70 (3.6)	925 (15.2)	766 (35.3)	<0.001
Polypharmacy, n (%)	433 (22.5)	3,320 (55.2)	1,673 (78.0)	<0.001

BMI = body mass index; CKD = chronic kidney disease; COPD = chronic obstructive pulmonary disease; DBP = diastolic blood pressure; PAD = peripheral arterial disease; SBP = systolic blood pressure; TE = thromboembolic.

Baseline characteristics associated with frailty

A multivariable multinomial logistic model showed that female sex, being enrolled in Eastern Europe and in-hospital and polypharmacy were associated with both pre-frailty and frailty, whereas low socio-economic status was associated with frailty (Supplemental Table 2). Reporting regular/intense exercise, paroxysmal AF and being enrolled in Southern Europe were inversely associated with pre-frailty and frailty (Supplemental Table 2).

Relationship between FI and AF Scores

Both CHA₂DS₂-VASc and HAS-BLED scores increased with FI (Table 1 and Supplemental Figure 1) (both P < 0.001). A linear regression model, adjusted for AF type and EHRA score, showed that both CHA₂DS₂-VASc and HAS-BLED scores were independently associated with FI [Beta 0.025, 95% confidence interval (CI) 0.025–0.026, t = 68.608, P < 0.001 and Beta 0.031, 95% CI 0.029–0.032, t = 46.211, P < 0.001, respectively], with no evidence of collinearity (Variance Inflation Factor [VIF] = 1.064, maximum condition index = 9.625, and VIF = 1.029, maximum condition index = 9.555, respectively). The two scores were independently associated with increasing FI also in a multivariable model containing both (data not shown).

AF management according to frailty

The use of antithrombotic therapies and clinical management according to frailty is reported in Supplemental Table 3. The highest prevalence of oral anticoagulants (OAC) prescription was observed among pre-frail patients (87.6%), followed by frail (83.0%) and robust patients (80.6%) (P < 0.001). Multivariable logistic regression analysis showed that, differently than frail patients (odds ratio [OR] 0.70, 95% CI 0.55–0.89, P = 0.004), pre-frail patients were more likely to receive OAC than robust patients (OR 1.21, 95% CI 1.01–1.44, P = 0.036).

The use of vitamin K antagonists (VKAs) progressively increased with frailty, whereas the opposite was observed for non-vitamin K antagonist OACs (NOACs) (P < 0.001; Supplemental Table 3). Multivariable logistic regression analysis (Table 2) showed that VKAs, compared with no OAC use, were more likely prescribed in pre-frail patients and less likely prescribed in frail patients. Moreover, frail patients were less likely to receive a NOAC than robust ones. Among patients prescribed OAC, a multivariable logistic regression analysis showed that both pre-frail and frail patients were less likely to be prescribed a NOAC than a VKA (OR 0.83, 95% CI 0.72–0.97, P = 0.019 and OR 0.69, 95% CI 0.56–0.84, P < 0.001, respectively).

Table 2

AF management according to frailty classes

OAC prescription	OR*	95% CI	P
VKAs (n = 5,038) versus No OAC (n = 1,496)
Frailty classes   Robust Pre-frail Frail	Ref. 1.24 0.73	Ref. 1.02–1.51 0.56–0.94	Ref. 0.027 0.016
NOACs (n = 3,638) versus No OAC
Frailty classes   Robust Pre-frail Frail	Ref. 1.09 0.54	Ref. 0.90–1.33 0.41–0.70	Ref. 0.370 <0.001
Clinical management strategy	OR†	95% CI	P
Rate control (n = 4,603) versus observation (n = 1,508)
Frailty classes   Robust Pre-frail Frail	Ref. 1.23 1.33	Ref. 1.00–1.51 1.00–1.78	Ref. 0.045 0.052
Rhythm control (n = 4,039) versus observation
Frailty classes   Robust Pre-frail Frail	Ref. 0.98 0.99	Ref. 0.80–1.21 0.73–1.33	Ref. 0.864 0.933

^*Adjusted for CHA₂DS₂-VASc score, European region, low socio-economic status, domestic status, physical activity, site of inclusion, reason for admission, type of AF and polypharmacy; †adjusted for EHRA score, European region, low socio-economic status, domestic status, physical activity, site of inclusion, reason for admission, type of AF and polypharmacy.

The clinical management strategy at discharge is described in Supplemental Table 3. Frail patients were more likely managed according to a rate control strategy rather than a rhythm control strategy (P < 0.001) (Supplemental Table 3). After multivariable adjustments (Table 2), both pre-frailty and frailty were independently associated with the use of a rate control strategy, and no difference was observed regarding the rhythm control strategy. Excluding those patients managed exclusively with an observation strategy, both pre-frailty and frailty were associated with lower odds of receiving rhythm control strategy than rate control (OR 0.75, 95% CI 0.63–0.89, P < 0.001 and OR 0.69, 95% CI 0.55–0.86, P = 0.001, respectively).

Relationship between quality of life and FI

At baseline, median [IQR] health utility score values decreased significantly according to frailty levels (0.95 [0.89–1.00], 0.87 [0.75–0.95], 0.71 [0.54–0.84] for robust, pre-frail and frail patients, respectively; P < 0.001); similar findings were observed for the visual analogue scale (VAS) values (80 [70–90], 70 [55–80], 60 [50–75] for the robust, pre-frail and frail patients, respectively; P < 0.001). Linear multivariable regression models, adjusted for AF type, OAC use and CHA₂DS₂-VASc and EHRA scores, showed that FI (for each 0.10 increase) was inversely and independently associated with both the health utility score (Beta −0.127, 95% CI -0.133/−0.121, t = −42.865, P < 0.001) and VAS (Beta −7.108, 95% CI −7.799/ −6.416, t = −20.147, P < 0.001), with no evidence of collinearity (VIF = 1.918, maximum condition index = 12.619 and VIF = 1.918, maximum condition index = 12.620, respectively).

Impact of frailty on health-resources use

Increasing frailty was broadly associated with higher use of healthcare resources (Supplemental Table 4). After adjustment for potential confounders, increasing frailty was significantly associated with greater use of both internal medicine/general practioner visits and emergency room admissions during follow-up. Similar results were also reported for hospitalisation events (both for cardiovascular and non-cardiovascular causes; Supplemental Table 4).

Relationship between frailty and major adverse events

Among the patients included in the analysis, follow-up data were available for 9,613 (95.5%) participants. During a mean (SD) follow-up of 1.84 (0.51) years, the rate of all major adverse events increased across frailty levels (all P < 0.001; Table 3). Kaplan–Meier curves showed increasing cumulative risk across the level of frailty for all the outcomes (Supplemental Figures 2–6). Cox multivariable adjusted analysis confirmed this finding (for both frailty levels and each 0.10 increase of FI) (Table 3).

Table 3

Rate of adverse clinical events and relationship with frailty

	Robust	Pre-frail	Frail	FI (Each 0.10)	P
All-cause death	56 (3.0)	491 (8.6)	362 (17.9)	–	<0.001
HR [95% CI]*	–	2.13 [1.60–2.84]	3.54 [2.56–4.89]	1.59 [1.45–1.75]
CV death	20 (1.1)	163 (2.8)	161 (8.0)	–	<0.001
HR [95% CI]*	–	1.92 [1.19–3.11]	4.15 [2.44–7.05]	1.89 [1.63–2.20]
Non-CV death	36 (1.9)	328 (5.7)	201 (10.0)	–	<0.001
HR [95% CI]*	–	2.25 [1.57–3.22]	3.17 [2.10–4.76]	1.42 [1.26–1.60]
MACEs	69 (3.8)	484 (8.6)	357 (17.8)	–	<0.001
HR [95% CI]*	–	1.80 [1.35–2.40]	3.41 [2.44–4.77]	1.69 [1.52–1.88]
Major bleeding	16 (0.9)	125 (2.2)	65 (3.9)	–	<0.001
HR [95% CI]†	–	2.25 [1.32–3.85]	2.87 [1.55–5.29]	1.32 [1.09–1.59]

^*Adjusted for type of AF, CHA₂DS₂-VASc score, EHRA score, use of OAC; †adjusted for type of AF, HAS-BLED score, EHRA score, use of OAC.

The analysis of the interaction between frailty and age on the risk of all-cause mortality and major adverse cardiovascular events (MACEs) is reported in Supplemental Table 5. Regression curves describing the association between FI, age strata and risk of outcomes are reported in Supplemental Figures 7 and 8. The risk of all-cause death increased progressively with FI across the age strata, but the difference in risk magnitude was lower with increasing age and FI (Supplemental Figure 7); conversely, while the risk of MACEs increased progressively with FI up until FI = 0.40 across all age classes, the magnitude of risk increase for higher FI values appeared higher in younger patients (Supplemental Figure 8).

Spline curves analysis and impact of OAC

Multivariable adjusted restricted cubic splines of the association between FI and the risk of outcomes are reported in Supplemental Figure 9A–E, with FI = 0.10 as a reference, and showed a non-linear relationship between FI and risk of the examined outcomes was observed.

When stratifying the analysis according to OAC use, all-cause death was significantly lower in patients treated with OAC reporting a FI between 0.05 and 0.36. In patients with a higher FI (~3% of the cohort), no difference was found in the risk of all-cause death between OAC users and non-users (Figure 2A). Patients with a FI between 0.03 and 0.44 (equal to more than 98% percentile of the cohort distribution) treated with OAC showed a significantly lower risk of MACEs than those not treated with OAC (Figure 2B). Similar data were found for the occurrence of cardiovascular death and non-cardiovascular death (Supplemental Figures 10 and 11). No difference was found across the spectrum of FI regarding major bleeding (Supplemental Figure 12).

Figure 2

Association between FI and risk of all-cause death and maces according to oac use. (A) All-cause death; (B) MACEs. (Red Lower Line) OAC prescribed; (Blue Upper Line) OAC not prescribed.

Predictive performance of FI among AF patients

FI had a modest-to-good predictive value for all the major adverse outcomes examined (Supplemental Table 6), with the highest c-index found for the occurrence of cardiovascular death (0.715, 95% CI 0.688–0.741) and the lowest for major bleeding occurrence (0.611, 95% CI 0.575–0.648).

Discussion

In a large and representative cohort of contemporary AF patients, frailty was found in about one out of five persons, whereas the prevalence of pre-frailty was around 60%. Frailty and pre-frailty were associated with greater deprivation, both on health and social aspects, and increasing FI was independently associated with higher thromboembolic and bleeding risks estimates, as reflected by CHA₂DS₂-VASc and HAS-BLED scores. Frailty also impacted the clinical management of AF patients (including OAC prescription) and also had a detrimental effect on quality of life and healthcare resources use. Finally, frailty and pre-frailty were associated with a proportionally higher risk for all major adverse outcomes examined, and FI was non-linearly associated with an increased risk of adverse events. OAC reduced the risks of outcomes, except in patients with very high/extreme frailty, without any significant increase in the risk of major bleeding across the spectrum of FI (Central Illustration).

Recently, the concept of frailty has gained significant attention, even beyond geriatric medicine, where it was initially conceived [3–6]. Given the high burden of multimorbidity, the impact on quality of life, perceived health and the risk for major adverse outcomes, AF appears to be significantly burdened by frailty [9, 10]. However, the relationship between AF and frailty has been investigated in a limited number of studies and cohorts, providing so far only inconsistent evidence [12, 19, 20].

Our study assesses and describes the epidemiology of frailty in a cohort of contemporary European AF patients, and represents the first large validation of the FI tool in this clinical and geographical setting. Our results show that more than 80% of European AF patients present with some degree of frailty. Previous estimates of frailty prevalence among AF patients showed considerable variability, ranging from 1% to over 80% [12, 19, 20]. A sub-analysis of the ENGAGE-AF TIMI 48 trial reported a similar prevalence of frailty and prefrailty, although in the context of a randomised controlled trial [21]. Finally, a recent meta-analysis showed how prevalence of frailty among AF patients is up to 40%, reaching 75% when considering also pre-frailty [22]; these findings are consistent with our results.

Our analysis also shows how several factors associated with a more deprived or susceptible personal and health status (i.e. low socio-economic status, hospitalised patients, increased polypharmacy) were incrementally associated with pre-frailty and frailty. Moreover, the significant associations between frailty and age, female sex and physical activity were expected [23–25]. Consistency with previous evidence further strengthens our results and confirms the estimates of frailty in the present cohort [23–25]. Also, paroxysmal AF was inversely associated with frailty, confirming that more permanent AF is associated with a greater burden of comorbidities [26]. Thromboembolic and bleeding risks contribute to the burden of frailty, further underlining the relationship between AF and frailty.

We also found that frailty was inversely associated with OAC prescriptions and lower likelihood of receiving a rhythm control strategy. Furthermore, frail individuals were less likely to be prescribed with both VKAs and NOACs, whereas VKAs were more likely prescribed in pre-frail patients. Finally, considering only patients prescribed with OAC, we found NOACs less likely prescribed than VKAs in both pre-frail and frail patients.

Thus far, data about the relationship between frailty and OAC prescription have been controversial [12]. Our data reinforce previous evidence on the OAC undertreatment of frail AF patients [22], reflecting the substantial absence of specific evidence and, in turn, of guidelines’ recommendations related to the prescription and management of OAC in frail patients [9]. Nevertheless, the available evidence suggests that frailty very likely represents an obstacle to OAC prescription due to physicians’ concerns about the risk of major bleeding [12, 27]. Moreover, lower NOACs prescription among frail patients may reflect the limited data on the effectiveness and safety of NOACs in this patient group [28–30].

Similarly, limited data are available regarding the relationship between frailty and rate/rhythm control. In a cohort of patients age ≥ 65 years old, rate control was more prescribed than rhythm control, although with no differences between frail and non-frail patients [31], while a survey performed by the EHRA, showed that 40% of cardiologists reported rate control as a unique approach in frail patients, whereas 57% believed that both approaches could be used [27]. In this context, our data prove that AF patients are less managed with rhythm control according to the burden of frailty, which entails conservative management.

Our study also shows how frailty impacts quality of life, healthcare resource use and risk of major adverse events in AF patients. While the relationship between frailty, quality of life and higher use of healthcare resources has already been described [32–34], our study is the first to analyse these relationships in AF patients, demonstrating a detrimental effect of frailty on both. Furthermore, we demonstrated a significant association between increasing frailty and the risk of all major adverse events. The evidence of an association between FI and higher risk of death has been already reported in the general population [35], as well as among AF patients [19]. However, previous studies may have not been able to achieve the same granularity of analysis [36, 37]. For example, Wilkinson et al. [36] found a significant association with all-cause mortality, but did not show any association with cerebrovascular events and only a partial association with bleeding outcomes. Conversely, Gugganig et al. [37] demonstrated a significant association with adverse outcomes (i.e. all-cause death, stroke, bleeding), but did not explore the causes of death, the overall risk of cardiovascular events or the relevant outcome of major bleeding. Furthermore, they did not show a similar ‘exposure-effect’ relationship as we did in our analysis [37]. While a recent meta-analysis confirmed that frailty increases the risk of outcomes in AF patients [22], the granularity of our data allowed us to expand the understanding of this association, describing an ‘exposure-effect’ between the burden of frailty and risk of outcomes. The results of the predictive analysis reinforce these results, showing that FI has a good-to-moderate predictive ability for all the outcomes considered (in line with all the AF scores used in clinical practice) [38].

We also showed how age (i.e. chronological aging) and frailty (i.e. biological aging) are independently able to influence the risk of outcomes. The observed interaction between age and FI suggests that the impact of frailty is even more prominent than age in determining the occurrence of events. The relative impact of FI is particularly important in younger subjects regarding the all-cause death and MACEs, consistently with a clinically relevant role of frailty already observed in younger adults [39]. Nevertheless, our data demonstrate that, although age plays a role as determinant of frailty, chronological and biological aging should be distinguished because only partially overlapping [40, 41].

The spline curves analysis suggests that FI is non-linearly associated with the risk of outcomes, and that OAC reduces the risk even up to high levels of FI, with no difference only in those with very or extremely high FI. These results underline how the use of OAC allows a significant reduction of risk only in patients that have a significant residual capacity. Our data are also reassuring about the positive benefit/risk ratio even in AF patients with moderate to high levels of frailty, since the use of OAC did not increase the risk of bleeding at any level of frailty. These data seem to support a recent EHRA consensus on the use of NOACs in AF patients [42], introducing that in some patients with extreme frailty, the OAC prescription may not be safe. Also, the authors suggested for the first time the use of an objective tool to measure frailty, the Clinical Frailty Scale [42].

Taken together, our data emphasise the need for a routine evaluation of frailty in AF patients. A formal assessment of frailty—through the means of geriatric comprehensive assessment, followed by a personalised intervention—can reduce the burden of frailty, leading to improvement in clinical outcomes [43–46]. More data are needed to elucidate which patients would benefit the most from receiving a formal frailty assessment. Combining the evaluation of frailty with an integrated care management approach, recommended as the ‘Atrial fibrillation Better Care’ pathway [47, 48], could significantly reduce all the primary AF-related adverse outcomes.

Limitations

The main limitation of the current study relates to the observational nature of the registry itself. Consequently, the study was not specifically powered to determine differences between the subgroups examined. The absence of a central events adjudication with an investigator-based reporting of the adverse outcomes represents another limitation, which entails caution in interpreting the current results. Third, since not all the patients included in the analysis had follow-up available represents another limitation. Moreover, since our project is derived from an AF registry we could not evaluate whether frailty could have a specific, stronger impact on AF patients compared with non-AF subjects. Lastly, in the process of building the FI, the absence of specific tools to measure physical and other types of physiological performance, which are instead evaluated by the EQ-5D-5L, represents another limitation that could limit the generalizability of results, particularly in relation to older subjects.

Conclusions

In this large European cohort of unselected AF patients, we found a highly significant burden of frailty, influencing significantly all the main aspects related to the management of AF, comprising OAC prescription and clinical management. A higher burden of frailty (i.e. biological age) was associated with a higher risk for all the major adverse events independently and more prominently than chronological age. The clinical benefit of using OAC was maintained even in patients with high frailty, but not in those with very high/extreme frailty. More data are still needed on the optimal management of this topical issue in AF patients.

Executive committee

G. Boriani (Chair), G.Y.H. Lip, L. Tavazzi, A. P. Maggioni, G-A. Dan, T. Potpara, M. Nabauer, F. Marin, Z. Kalarus, L. Fauchier, R. Ferrari, A. Shantsila.

Steering Committee (National Coordinators)

A. Goda, G. Mairesse, T.Shalganov, L. Antoniades, M. Taborsky, S. Riahi, P. Muda, I. García Bolao, O. Piot, M. Nabauer, K. Etsadashvili, E.Simantirakis, M. Haim, A. Azhari, J. Najafian, M. Santini, E. Mirrakhimov, K.a Kulzida, A. Erglis, L. Poposka, M. Burg, H. Crijns, Ö. Erküner, D. Atar, R. Lenarczyk, M. Martins Oliveira, D. Shah, G-A. Dan, E. Serdechnaya, T. Potpara, E. Diker, G.Y.H. Lip, D. Lane.

Investigators

ALBANIA Durrës: E. Zëra, Tirana: U. Ekmekçiu, V. Paparisto, M. Tase, Tirana: H. Gjergo, J. Dragoti, A. Goda, BELGIUM Bastogne: M. Ciutea, N. Ahadi, Z. el Husseini, M. Raepers, Gilly: J. Leroy, P. Haushan, A. Jourdan, Haine Saint Paul: C. Lepiece, Hasselt: L. Desteghe, J. Vijgen, P. Koopman, G. Van Genechten, H. Heidbuchel, Kortrijk: T. Boussy, M. De Coninck, H. Van Eeckhoutte, N. Bouckaert, La Louviere: A. Friart, J. Boreux, C. Arend, Liege: P. Evrard, Liège: L. Stefan, E. Hoffer, J. Herzet, M. Massoz, Liège: C. Celentano, M. Sprynger, L. Pierard, Liège: P. Melon, Overpelt: B. Van Hauwaert, C. Kuppens, D. Faes, D. Van Lier, A. Van Dorpe, Waremme: A. Gerardy, Yvoir: O. Deceuninck, O. Xhaet, F. Dormal, E. Ballant, D. Blommaert, BULGARIA Pleven: D. Yakova, M. Hristov, T. Yncheva, N. Stancheva, S. Tisheva, Plovdiv: M. Tokmakova, F. Nikolov, D. Gencheva, Sofia: T. Shalganov, B. Kunev, M. Stoyanov, Sofia: D. Marchov, V. Gelev, V. Traykov, Varna: A. Kisheva, H. Tsvyatkov, R. Shtereva, S. Bakalska-Georgieva, S. Slavcheva, Y. Yotov, CZECH REPUBLIC Ústí nad Labem: M. Kubíčková, DENMARK Aalborg: A. Marni Joensen, A. Gammelmark, L. Hvilsted Rasmussen, P. Dinesen, S. Riahi, S. Krogh Venø, B. Sorensen, A. Korsgaard, K. Andersen, C. Fragtrup Hellum, Esbjerg: A. Svenningsen, O. Nyvad, P. Wiggers, Herning: O. May, A. Aarup, B. Graversen, L. Jensen, M. Andersen, M. Svejgaard, S. Vester, S. Hansen, V. Lynggaard, Madrid: M. Ciudad, Tallinn: R. Vettus, Tartu: P. Muda, ESTONIA Elche, Alicante: A. Maestre, Toledo: S. Castaño, FRANCE Abbeville: S. Cheggour, Abbeville: J. Poulard, V. Mouquet, S. Leparrée, Aix-en-Provence: J. Bouet, J. Taieb, Amiens: A. Doucy, H. Duquenne, Angers: A. Furber, J. Dupuis, J. Rautureau, Aurillac: M. Font, P. Damiano, Avignon Cedex: M. Lacrimini, Brest: J. Abalea, S. Boismal, T. Menez, J. Mansourati, Chartres: G. Range, H. Gorka, C. Laure, C. Vassalière, Creteil: N. Elbaz, N. Lellouche, K. Djouadi, Montpellier: F. Roubille, D. Dietz, J. Davy, Nimes: M. Granier, P. Winum, C. Leperchois-Jacquey, Paris: H. Kassim, E. Marijon, J. Le Heuzey, Paris: J. Fedida, C. Maupain, C. Himbert, E. Gandjbakhch, F. Hidden-Lucet, G. Duthoit, N. Badenco, T. Chastre, X. Waintraub, M. Oudihat, J. Lacoste, C. Stephan, Pau: H. Bader, N. Delarche, L. Giry, Pessac: D. Arnaud, C. Lopez, F. Boury, I. Brunello, M. Lefèvre, R. Mingam, M. Haissaguerre, Rennes: M. Le Bidan, D. Pavin, V. Le Moal, C. Leclercq, Saint Denis: O. Piot, T. Beitar, Saint Etienne: I. Martel, A. Schmid, N. Sadki, C. Romeyer-Bouchard, A. Da Costa, Tours: I. Arnault, M. Boyer, C. Piat, L. Fauchier, FYR MACEDONIA Bitola: N. Lozance, S. Nastevska, Ohrid: A. Doneva, B. Fortomaroska Milevska, B. Sheshoski, K. Petroska, N. Taneska, N. Bakrecheski, Skopje: K. Lazarovska, S. Jovevska, V. Ristovski, A. Antovski, Skopje: E. Lazarova, I. Kotlar, J. Taleski, L. Poposka, S. Kedev, Skopje: N. Zlatanovik, Štip: S. Jordanova, T. Bajraktarova Proseva, S. Doncovska, GEORGIA Tbilisi: D. Maisuradze, A. Esakia, E. Sagirashvili, K. Lartsuliani, N. Natelashvili, N. Gumberidze, R. Gvenetadze, Tbilisi: K. Etsadashvili, N. Gotonelia, N. Kuridze, Tbilisi: G. Papiashvili, I. Menabde, GERMANY Aachen: S. Glöggler, A. Napp, C. Lebherz, H. Romero, K. Schmitz, M. Berger, M. Zink, S. Köster, J. Sachse, E. Vonderhagen, G. Soiron, K. Mischke, Bad Reichenhall: R. Reith, M. Schneider, Berlin: W. Rieker, Biberach: D. Boscher, A. Taschareck, A. Beer, Boppard: D. Oster, Brandenburg: O. Ritter, J. Adamczewski, S. Walter, Chemnitz: A. Frommhold, E. Luckner, J. Richter, M. Schellner, S. Landgraf, S. Bartholome, Chemnitz: R. Naumann, J. Schoeler, Dachau: D. Westermeier, F. William, K. Wilhelm, M. Maerkl, Detmold: R. Oekinghaus, M. Denart, M. Kriete, U. Tebbe, Ebersbach: T. Scheibner, Erlangen: M. Gruber, A. Gerlach, C. Beckendorf, L. Anneken, M. Arnold, S. Lengerer, Z. Bal, C. Uecker, H. Förtsch, S. Fechner, V. Mages, Friedberg: E. Martens, H. Methe, Göttingen: T. Schmidt, Hamburg: B. Schaeffer, B. Hoffmann, J. Moser, K. Heitmann, S. Willems, S. Willems, Hartmannsdorf: C. Klaus, I. Lange, Heidelberg: M. Durak, E. Esen, Itzehoe: F. Mibach, H. Mibach, Kassel: A. Utech, Kirchzarten: M. Gabelmann, R. Stumm, V. Ländle, Koblenz: C. Gartner, C. Goerg, N. Kaul, S. Messer, D. Burkhardt, C. Sander, R. Orthen, S. Kaes, Köln: A. Baumer, F. Dodos, Königsbrück: A. Barth, G. Schaeffer, Leisnig: J. Gaertner, J. Winkler, Leverkusen: A. Fahrig, J. Aring, I. Wenzel, Limburg: S. Steiner, A. Kliesch, E. Kratz, K. Winter, P. Schneider, Ludwigsburg: A. Haag, I. Mutscher, R. Bosch, Markkleeberg: J. Taggeselle, S. Meixner, Meissen: A. Schnabel, Meppen: A. Shamalla, H. Hötz, A. Korinth, Merzig: C. Rheinert, Moosburg: G. Mehltretter, Mühldorf: B. Schön, N. Schön, A. Starflinger, E. Englmann, Munich: G. Baytok, T. Laschinger, G. Ritscher, Munich: A. Gerth, Münster: D. Dechering, L. Eckardt, Nienburg: M. Kuhlmann, N. Proskynitopoulos, Paderborn: J. Brunn, K. Foth, Pirna: C. Axthelm, H. Hohensee, K. Eberhard, S. Turbanisch, Plauen: N. Hassler, A. Koestler, Riesa: G. Stenzel, Riesa: D. Kschiwan, M. Schwefer, S. Neiner, S. Hettwer, Rotenburg a.d. Fulda: M. Haeussler-Schuchardt, R. Degenhardt, S. Sennhenn, S. Steiner, Starnberg: M. Brendel, Westerstede: A. Stoehr, W. Widjaja, S. Loehndorf, A. Logemann, J. Hoskamp, J. Grundt, Zorneding: M. Block, Zwiesel: R. Ulrych, A. Reithmeier, V. Panagopoulos, ITALY Bologna: C. Martignani, D. Bernucci, E. Fantecchi, I. Diemberger, M. Ziacchi, M. Biffi, P. Cimaglia, J. Frisoni, G. Boriani, Firenze: I. Giannini, S. Boni, S. Fumagalli, S. Pupo, A. Di Chiara, P. Mirone, Modena: E. Fantecchi, G. Boriani, F. Pesce, C. Zoccali, V.L. Malavasi, KAZAKHSTAN Almaty: A. Mussagaliyeva, B. Ahyt, Z. Salihova, K. Koshum-Bayeva, KYRGYZSTAN Bishkek: A. Kerimkulova, A. Bairamukova, E. Mirrakhimov, LATVIA Riga: B. Lurina, R. Zuzans, S. Jegere, I. Mintale, K. Kupics, K. Jubele, A. Erglis, O. Kalejs, MALTA Birkirkara: K. Vanhear, M. Burg, M. Cachia, E. Abela, S. Warwicker, T. Tabone, R. Xuereb, MONTENEGRO Podgorica: D. Asanovic, D. Drakalovic, M. Vukmirovic, N. Pavlovic, L. Music, N. Bulatovic, A. Boskovic, THE NETHERLANDS Almere: H. Uiterwaal, N. Bijsterveld, Amsterdam: J. De Groot, J. Neefs, N. van den Berg, F. Piersma, A. Wilde, Delfzijl: V. Hagens, Enschede: J. Van Es, J. Van Opstal, B. Van Rennes, H. Verheij, W. Breukers, Heerenveen: G. Tjeerdsma, R. Nijmeijer, D. Wegink, R. Binnema, Hengelo: S. Said, Maastricht: Ö. Erküner, S. Philippens, W. van Doorn, H. Crijns, Rotterdam: T. Szili-Torok, R. Bhagwandien, P. Janse, A. Muskens, s-Hertogenbosch: M. van Eck, R. Gevers, N. van der Ven, Venlo: A. Duygun, B. Rahel, J. Meeder, NORWAY Oslo: A. Vold, C. Holst Hansen, I. Engset, D. Atar, POLAND Bytom: B. Dyduch-Fejklowicz, E. Koba, M. Cichocka, Cieszyn: A. Sokal, A. Kubicius, E. Pruchniewicz, Gliwice: A. Kowalik-Sztylc, W. Czapla, Katowice: I. Mróz, M. Kozlowski, T. Pawlowski, M. Tendera, Katowice: A. Winiarska-Filipek, A. Fidyk, A. Slowikowski, M. Haberka, M. Lachor-Broda, M. Biedron, Z. Gasior, Kielce: M. Kołodziej, M. Janion, Kielce: I. Gorczyca-Michta, B. Wozakowska-Kaplon, Łódź: M. Stasiak, P. Jakubowski, T. Ciurus, J. Drozdz, Łódź: M. Simiera, P. Zajac, T. Wcislo, P. Zycinski, J. Kasprzak, Nysa: A. Olejnik, E. Harc-Dyl, J. Miarka, M. Pasieka, M. Ziemińska-Łuć, W. Bujak, Opoczno: A. Śliwiński, A. Grech, J. Morka, K. Petrykowska, M. Prasał, Opole: G. Hordyński, P. Feusette, P. Lipski, A. Wester, Radlin: W. Streb, Rzeszów: J. Romanek, P. Woźniak, M. Chlebuś, P. Szafarz, W. Stanik, Szczecin: M. Zakrzewski, J. Kaźmierczak, Szczecin: A. Przybylska, E. Skorek, H. Błaszczyk, M. Stępień, S. Szabowski, W. Krysiak, M. Szymańska, Tarnów: J. Karasiński, J. Blicharz, M. Skura, Warsaw: K. Hałas, L. Michalczyk, Z. Orski, K. Krzyżanowski, A. Skrobowski, Warsaw: L. Zieliński, M. Tomaszewska-Kiecana, M. Dłużniewski, Warsaw: M. Kiliszek, M. Peller, M. Budnik, P. Balsam, G. Opolski, A. Tymińska, K. Ozierański, A. Wancerz, Warsaw: A. Borowiec, E. Majos, R. Dabrowski, H. Szwed, Zabrze: A. Musialik-Lydka, Zabrze: A. Leopold-Jadczyk, E. Jedrzejczyk-Patej, M. Koziel, R. Lenarczyk, M. Mazurek, Z. Kalarus, Zabrze: K. Krzemien-Wolska, P. Starosta, E. Nowalany-Kozielska, Zakopane: A. Orzechowska, M. Szpot, M. Staszel, PORTUGAL Almada: S. Almeida, H. Pereira, L. Brandão Alves, R. Miranda, L. Ribeiro, Carnaxide Lisboa: F. Costa, F. Morgado, P. Carmo, P. Galvao Santos, R. Bernardo, P. Adragão, Santarém: G. Ferreira da Silva, M. Peres, M. Alves, M. Leal, Vila Real: A. Cordeiro, P. Magalhães, P. Fontes, S. Leão, Viseu: A. Delgado, A. Costa, B. Marmelo, B. Rodrigues, D. Moreira, J. Santos, L. Santos, ROMANIA Arad: A. Terchet, D. Darabantiu, S. Mercea, V. Turcin Halka, A. Pop Moldovan, Brasov: A. Gabor, B. Doka, G. Catanescu, H. Rus, L. Oboroceanu, E. Bobescu, Bucharest: R. Popescu, A. Dan, A. Buzea, I. Daha, G. Dan, I. Neuhoff, Bucharest: M. Baluta, R. Ploesteanu, N. Dumitrache, M. Vintila, Bucharest: A. Daraban, C. Japie, E. Badila, H. Tewelde, M. Hostiuc, S. Frunza, E. Tintea, D. Bartos, Bucharest: A. Ciobanu, I. Popescu, N. Toma, C. Gherghinescu, D. Cretu, N. Patrascu, C. Stoicescu, C. Udroiu, G. Bicescu, V. Vintila, D. Vinereanu, M. Cinteza, R. Rimbas, Iași: M. Grecu, Oradea: A. Cozma, F. Boros, M. Ille, O. Tica, R. Tor, A. Corina, A. Jeewooth, B. Maria, C. Georgiana, C. Natalia, D. Alin, D. Dinu-Andrei, M. Livia, R. Daniela, R. Larisa, S. Umaar, T. Tamara, M. Ioachim Popescu, Târgu Mureș: D. Nistor, I. Sus, O. Coborosanu, Timișoara: N. Alina-Ramona, R. Dan, L. Petrescu, Timișoara: G. Ionescu, I. Popescu, C. Vacarescu, E. Goanta, M. Mangea, A. Ionac, C. Mornos, D. Cozma, S. Pescariu, RUSSIAN FEDERATION Arkhangelsk: E. Solodovnicova, I. Soldatova, J. Shutova, L. Tjuleneva, T. Zubova, V. Uskov, Arkhangelsk: D. Obukhov, G. Rusanova, Arkhangelsk: I. Soldatova, N. Isakova, S. Odinsova, T. Arhipova, Arkhangelsk: E. Kazakevich, E. Serdechnaya, O. Zavyalova, Saint-Petersburg: T. Novikova, Saint-Petersburg: I. Riabaia, S. Zhigalov, Saint-Petersburg: E. Drozdova, I. Luchkina, Y. Monogarova, Vladivostok: D. Hegya, L. Rodionova, L. Rodionova, V. Nevzorova, Vladivostok: I. Soldatova, O. Lusanova, SERBIA Belgrade: A. Arandjelovic, D. Toncev, M. Milanov, N. Sekularac, Belgrade: M. Zdravkovic, S. Hinic, S. Dimkovic, T. Acimovic, J. Saric, Belgrade: M. Polovina, T. Potpara, B. Vujisic-Tesic, M. Nedeljkovic, Belgrade: M. Zlatar, M. Asanin, Belgrade: V. Vasic, Z. Popovic, Belgrade: D. Djikic, M. Sipic, V. Peric, B. Dejanovic, N. Milosevic, Belgrade: A. Stevanovic, A. Andric, B. Pencic, M. Pavlovic-Kleut, V. Celic, Kragujevac: M. Pavlovic, M. Petrovic, M. Vuleta, N. Petrovic, S. Simovic, Z. Savovic, S. Milanov, G. Davidovic, V. Iric-Cupic, Niška Banja: D. Simonovic, M. Stojanovic, S. Stojanovic, V. Mitic, V. Ilic, D. Petrovic, M. Deljanin Ilic, S. Ilic, V. Stoickov, Pirot: S. Markovic, Šabac:S. Kovacevic. SPAIN Alicante: A. García Fernandez, Benalmadena: A. Perez Cabeza, Córdoba: M. Anguita, Granada: L. Tercedor Sanchez, Huarte: E. Mau, J. Loayssa, M. Ayarra, M. Carpintero, Madrid: I. Roldán Rabadan, Murcia: M. Leal, Murcia: M. Gil Ortega, Murcia: A. Tello Montoliu, E. Orenes Piñero, S. Manzano Fernández, F. Marín, A. Romero Aniorte, A. Veliz Martínez, M. Quintana Giner, Pamplona: G. Ballesteros, M. Palacio, O. Alcalde, I. García-Bolao, San Juan de Alicante: V. Bertomeu Gonzalez, Santiago de Compostela: F. Otero-Raviña, J. García Seara, J. Gonzalez Juanatey, SWITZERLAND Geneva: N. Dayal, P. Maziarski, P. Gentil-Baron, D. Shah, TURKEY Adana: M. Koç, Afyon: E. Onrat, I. E. Dural, Ankara: K. Yilmaz, B. Özin, Ankara: S. Tan Kurklu, Y. Atmaca, Ankara: U. Canpolat, L. Tokgozoglu, Ankara: A. K. Dolu, B. Demirtas, D. Sahin, Ankara: O. Ozcan Celebi, E. Diker, Antalya: G. Gagirci, Bayraklı/Izmir: U.O.Turk, Bursa: H. Ari, Diyarbakır: N. Polat, N. Toprak, Gaziantep: M. Sucu, Görükle-Bursa: O. Akin Serdar, Istanbul: A. Taha Alper, Istanbul: A. Kepez, Istanbul: Y. Yuksel, Kurupelit—Samsun: A. Uzunselvi, S. Yuksel, M. Sahin, Merkez/Düzce: O. Kayapinar, Mersin: T. Ozcan, Sivas: H. Kaya, M. B. Yilmaz, Trabzon: M. Kutlu, Yüreğir-Adana: M. Demir, UNITED KINGDOM Barnstaple: C. Gibbs, S. Kaminskiene, M. Bryce, A. Skinner, G. Belcher, J. Hunt, L. Stancombe, B. Holbrook, C. Peters, S. Tettersell, Birmingham: A. Shantsila, D. Lane, K. Senoo, M. Proietti, K. Russell, P. Domingos, S. Hussain, J. Partridge, R. Haynes, S. Bahadur, R. Brown, S. McMahon, G. Y H Lip, Blackburn: J. McDonald, K. Balachandran, R. Singh, S. Garg, H. Desai, K. Davies, W. Goddard, Blackpool: G. Galasko, I. Rahman, Y. Chua, O. Payne, S. Preston, O. Brennan, L. Pedley, C. Whiteside, C. Dickinson, J. Brown, K. Jones, L. Benham, R. Brady, Carlisle: L. Buchanan, A. Ashton, H. Crowther, H. Fairlamb, S. Thornthwaite, C. Relph, A. McSkeane, U. Poultney, N. Kelsall, P. Rice, T. Wilson, Chertsey: M. Wrigley, R. Kaba, T. Patel, E. Young, J. Law, Cramlington: C. Runnett, H. Thomas, H. McKie, J. Fuller, S. Pick, Exeter: A. Sharp, A. Hunt, K. Thorpe, C. Hardman, E. Cusack, L. Adams, M. Hough, S. Keenan, A. Bowring, J. Watts, Great Yarmouth: J. Zaman, K. Goffin, H. Nutt, Harrogate: Y. Beerachee, J. Featherstone, C. Mills, J. Pearson, L. Stephenson, Huddersfield: S. Grant, A. Wilson, C. Hawksworth, I. Alam, M. Robinson, S. Ryan, Macclesfield: R. Egdell, E. Gibson, M. Holland, D. Leonard, Maidstone: B. Mishra, S. Ahmad, H. Randall, J. Hill, L. Reid, M. George, S. McKinley, L. Brockway, W. Milligan, Manchester: J. Sobolewska, J. Muir, L. Tuckis, L. Winstanley, P. Jacob, S. Kaye, L. Morby, Nottingham: A. Jan, T. Sewell, Poole: C. Boos, B. Wadams, C. Cope, P. Jefferey, Portsmouth: N. Andrews, A. Getty, A. Suttling, C. Turner, K. Hudson, R. Austin, S. Howe, Redhill: R. Iqbal, N. Gandhi, K. Brophy, P. Mirza, E. Willard, S. Collins, N. Ndlovu, Rhyl: E. Subkovas, V. Karthikeyan, L. Waggett, A. Wood, A. Bolger, J. Stockport, L. Evans, E. Harman, J. Starling, L. Williams, V. Saul, Salisbury: M. Sinha, L. Bell, S. Tudgay, S. Kemp, J. Brown, L. Frost, Shrewsbury: T. Ingram, A. Loughlin, C. Adams, M. Adams, F. Hurford, C. Owen, C. Miller, D. Donaldson, H. Tivenan, H. Button, South Shields: A. Nasser, O. Jhagra, B. Stidolph, C. Brown, C. Livingstone, M. Duffy, P. Madgwick, Southampton: P. Roberts, E. Greenwood, L. Fletcher, M. Beveridge, S. Earles, Taunton: D. McKenzie, D. Beacock, M. Dayer, M. Seddon, D. Greenwell, F. Luxton, F. Venn, H. Mills, J. Rewbury, K. James, K. Roberts, L. Tonks, Torquay: D. Felmeden, W. Taggu, A. Summerhayes, D. Hughes, J. Sutton, L. Felmeden, Watford: M. Khan, E. Walker, L. Norris, L. O’Donohoe, Weston-super-Mare: A. Mozid, H. Dymond, H. Lloyd-Jones, G. Saunders, D. Simmons, D. Coles, D. Cotterill, S. Beech, S. Kidd, Wolverhampton: B. Wrigley, S. Petkar, A. Smallwood, R. Jones, E. Radford, S. Milgate, S. Metherell, V. Cottam, Yeovil: C. Buckley, A. Broadley, D. Wood, J. Allison, K. Rennie, L. Balian, L. Howard, L. Pippard, S. Board, T. Pitt-Kerby.

Acknowledgements

EORP Oversight Committee, Executive and Steering Committees (National Coordinators) of the EURObservational Research Programme (EORP)—Atrial Fibrillation General Long-Term (EORP-AFGen LT) Registry of the European Society of Cardiology (ESC). Data collection was conducted by the EORP department by Patti-Ann McNeill as Project Officer, Viviane Missiamenou as Data Manager. Overall activities were coordinated and supervised by Doctor Aldo P. Maggioni (EORP Scientific Coordinator).

Declaration of Conflicts of Interest

D.L. has received investigator-initiated educational grants from Bristol-Myers Squibb (BMS), has been a speaker for Bayer, Boehringer Ingelheim and BMS/Pfizer and has consulted for BMS Boehringer Ingelheim; L.F. has been consultant or speaker for Bayer, BMS/Pfizer, Boehringer Ingelheim, Medtronic, Novartis; T.S.P. is consultant for Bayer and Pfizer, with no fees received personally; G.B. received small speaker’s fees from Medtronic, Boston, Boehringer Ingelheim and Bayer; G.Y.H.L. has been consultant and speaker for BMS/Pfizer, Boehringer Ingelheim and Daiichi-Sankyo. No fees are directly received personally. All the disclosures happened outside the submitted work. All other authors have nothing to declare.

Declaration of Sources of Funding

Since the start of EORP, the following companies have supported the programme: Abbott Vascular Int. (2011–21), Amgen Cardiovascular (2009–18), AstraZeneca (2014–21), Bayer (2009–18), Boehringer Ingelheim (2009–19), Boston Scientific (2009–12), The Bristol Myers Squibb and Pfizer Alliance (2011–16), The Alliance Daiichi Sankyo Europe GmbH and Eli Lilly and Company (2011–17), Edwards (2016–19), Gedeon Richter Plc. (2014–17), Menarini Int. Op. (2009–12), MSD-Merck & Co. (2011–14), Novartis Pharma AG (2014–20), ResMed (2014–16), Sanofi (2009–11), SERVIER (2010–21), Vifor (2019–22).

Author notes

Marco Proietti and Giulio Francesco Romiti authors equally contributed to the paper.

Giuseppe Boriani and Gregory Y. H. Lip Joint senior authors.