Frailty, genetic predisposition, and incident atrial fibrillation

Introduction

Atrial fibrillation (AF) is the most prevalent arrhythmia managed in clinical practice especially in the aging population.¹ Meanwhile, frailty is an age-related clinical condition featuring compromised physiological function of multiple systems.² However, little is known regarding the association between frailty and new-onset AF in the middle-aged and elderly and the modification effect of genetic susceptibility for AF.

Methods

This prospective cohort study is based on data from the UK Biobank, a nationwide cohort study.³ The UK Biobank has ethical approval from the North West Multicenter Research Ethics Committee, and all participants provided written consent for the study.

In the present study, we excluded participants with AF at baseline (n = 8376) and with missing values on ≥10 variables of 49 frailty index (FI) factors, any factor of frailty phenotype (FP) (n = 120 574) and any covariate (n = 4605), leaving 368 862 participants for the primary analysis. In addition, participants without genetic data (n = 8439), with any kinship to other individuals in the UK Biobank (n = 108 421), and who were not European descent (n = 15 203) were excluded, leaving 236 799 participants for genetic risk-associated analyses.

Two common methods of determining frailty were used. The Rockwood FI included 49 variables from the clinical examination. Categorical variables were dichotomized (no deficit = 0; deficit = 1), and ordinal variables were mapped into a score between 0 and 1. The sum of deficits was divided by the total number of possible deficits, resulting the FI values between 0 and 1.⁴ Participants with FI ≤ 0.10 were considered as the non-frail, while those with 0.10 < FI ≤ 0.21 and FI > 0.21 were classified as the pre-frail and frail, respectively.⁵ For FP, there are five indicators: weight loss, exhaustion, slow walking speed, low grip strength, and low physical activities. Participants were classified into the frail (three or more frailty criteria), the pre-frail (one or two frailty criteria), or the non-frail (none of the frailty criteria).²

The outcome AF was obtained from the first occurrence of new-onset AF that was defined by the International Classification of Diseases (ICD)-10th Revision code I48 with linkage from primary care, hospital admission electronic health records, and death register records.

The genetic risk of AF was based on the standard polygenic risk score (PRS, field ID 26212 in UK Biobank) supported by external genome-wide association studies data.⁶ We further classified participants into groups with high (the highest PRS quartile), intermediate (the middle two PRS quartiles), or low (the lowest PRS quartile) genetic risk.

Cox proportional hazard regression models were used to evaluate the hazard ratios (HRs) and 95% confidence intervals (CIs). The follow-up time was calculated from the date of baseline recruitment to the date of first-time AF diagnosis, death, or the censoring date (19 December 2022), whichever occurred first. Potential confounders adjusted in the model are listed in Figure 1. The multiplicative interaction analysis between frailty status and categories of genetic susceptibility to AF was performed by using the likelihood ratio test comparing models with and without a cross-product term. All statistical analyses were performed using IBM SPSS Statistics (version 26) and R software (version 4.0.1).

Results

A total of 368 862 participants (mean age 56.2 years, 47.3% male) were assessed. 46.5%, 43.1%, and 10.4% individuals met the criteria of FI-based non-frailty, pre-frailty, and frailty, respectively. For FP criteria, the corresponding numbers were 49.8%, 45.8%, and 4.4%.

During a median follow-up of 13.7 years, 24 269 new AF cases were documented. The HR (95% CI) of incident AF in pre-frailty and frailty was 1.18 (1.15–1.22) and 1.47 (1.41–1.53) according to FI and 1.07 (1.05–1.10) and 1.39 (1.32–1.47) based on FP, respectively. Age- and sex-specific associations between frailty and AF were further shown in Figure 1A. When stratified by 13 traditional risk factors, a stronger association between frailty and AF was observed in women than men, in participants <60 years than >60 years, and in those with low socioeconomic status than those with high socioeconomic status (all P for interaction < .004 after Bonferroni correction).

Compared with those with low genetic risk and non-frailty, participants with high genetic risk and frailty had the highest risk of AF (HR 4.59, 95% CI: 4.15–5.09 in FI and HR 3.44, 95% CI: 3.10–3.82 in FP) (Figure 1B). The association between frailty and AF was strengthened in those with low genetic risk (HR 1.79, 95% CI: 1.55–2.06 in low genetic risk vs. HR 1.31, 95% CI: 1.21–1.43 in high genetic risk) using FI, and the results were similar based on FP (both P for interaction < .001).

In sensitivity analyses, the results remained robust after excluding individuals having incident AF within the first 2 years of follow-up. In the competing risk analysis, the association of frailty with AF did not change due to the causes of death.

Discussion

In this large prospective cohort study, we newly found that compared with non-frail individuals, frail individuals had around 40% higher risk of incident AF. There was a significant interaction between frailty status and genetic risk on AF. The highest AF risk was for participants at high genetic risk and with frailty. However, the prominent association between frailty and AF was observed among individuals with low genetic risk, suggesting that individuals with low genetic risk may be more susceptible to frailty for incident AF. To the best of our knowledge, this is the first study investigating the longitudinal association among frailty, genetic predisposition, and new-onset AF.

Our previous studies reported traditional and novel risk factors of AF,^8,9 and this paper further indicates that frailty could be a novel risk factor of AF and may be incorporated into the evaluation of AF risk. Whether amelioration of frailty might represent another avenue to improve AF development warrants further investigation.

The study limitations include (i) the inability to establish causality; (ii) unknown and unmeasured confounding factors; (iii) the use of ICD-10 coding but not direct diagnosis for AF assessment; (iv) the potential for undiagnosed AF events; and (v) self-selected and not entirely representative nature of UK Biobank indicating replication in other populations.

Declarations

Disclosure of Interest

All authors declare no disclosure of interest for this contribution.

Data Availability

The data set used and analysed during the current study are available from UK Biobank (www.ukbiobank.ac.uk). This research has been conducted using the UK Biobank Resource under Application Number 77740.

Funding

N.W. is supported by the National Natural Science Foundation of China (82170870), Shanghai Municipal Health Commission (2022XD017), and Shanghai Municipal Human Resources and Social Security Bureau (2020074). The funders played no role in the design or conduction of the study; in the collection, management, analysis, or interpretation of the data; or in the preparation, review, or approval of the article.

Ethical Approval

The UK Biobank has ethical approval from the North West Multicenter Research Ethics Committee (REC reference 11/NW/0382), and all participants provided written consent for the study.

Pre-registered Clinical Trial Number

None supplied.

References

Author notes