Abstract
Background
Emerging epidemiological evidence implicates pulmonary dysfunction in cardiovascular pathogenesis, yet its arrhythmogenic potential remains poorly defined.
Objectives
We aimed to assess the link between ventilatory parameters, pulmonary disease phenotypes and risk of incident arrhythmias across diverse populations.
Methods
We analyzed data from 17,684 adults in two prospective cohort studies-the Atherosclerosis Risk in Communities (ARIC; n = 12,929) and Cardiovascular Health Study (CHS; n = 4,755). Adjudicated arrhythmia diagnoses (atrial fibrillation/flutter [AF/AFL], ventricular arrhythmias [VAs], high-grade atrioventricular [AV] block, and premature atrial/ventricular complexes [PAC/PVC]) were identified via hospitalization records and mortality data. Multivariable-adjusted Cox proportional hazards models quantified associations between forced expiratory volume in 1 s (FEV1%) predicted and forced vital capacity (FVC%) predicted quartiles with arrhythmia risk, adjusting for traditional cardiovascular risk factors.
Results
Over a median follow-up of 12.6 years, impaired FEV1% and FVC% corresponded to a graded increase in arrhythmia risk. Compared to the highest quartile, the lowest FEV1% predicted quartile had elevated hazards for any arrhythmias (HR 1.32, 95% CI 1.23–1.42), AF/AFL (HR 1.68, 1.52–1.85), VAs (HR 1.55, 1.29–1.86), high-grade AV block (HR 1.37, 1.08–1.73), and PAC/PVC (HR 1.42, 1.20–1.69). Similar trends were observed for FVC% predicted quartiles. These associations remained consistent in never-smoking individuals and across cohorts. Obstructive spirometry pattern was associated with the strongest arrhythmia risk, while restrictive ventilatory patterns showed relatively lower risk elevations. No association was observed with sick sinus syndrome.
Conclusions
Reduced pulmonary function suggested independent associations with incident arrhythmias across supraventricular, ventricular, and conduction system pathologies in two historical cohorts. These findings suggest that spirometric indices could potentially represent novel independent indicators for arrhythmia development worthy of further validation in contemporary settings,, with associations distinct from conventional cardiometabolic risk factors.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12916-025-04345-y.
Keywords: Cardiac arrhythmia, Lung function, COPD, Restrictive impairment
Background
Atrial fibrillation and flutter (AF/AFL) represent a major global health challenge, accounting for substantial arrhythmia-related morbidity and imposing a staggering clinical and economic burden on healthcare systems worldwide [1]. Notably, sudden cardiac death-a catastrophic outcome predominantly driven by malignant ventricular arrhythmias (VAs) such as sustained ventricular tachycardia-claims responsibility for nearly half of all cardiovascular mortality [2]. Clinically significant bradyarrhythmias predispose individuals to hemodynamically significant syncopal episodes, often necessitating transcatheter pacemaker implantation to restore physiological cardiac conduction [3]. These arrhythmic disorders collectively underscore an urgent need to refine risk stratification paradigms.
Impaired lung function is a significant predictor of cardiovascular disease mortality [4]. Large prospective and retrospective studies of chronic obstructive pulmonary disease (COPD) patients have shown that the causes of their demise were at least partly due to arrhythmias [5]. Recent studies have indicated that, unlike COPD, proportional reductions in expiratory lung volumes without obstruction, otherwise known as preserved ratio impaired spirometry were associated with a small but statistically significant increased risk for mortality and adverse cardiovascular outcomes [6].
Although previous studies have explored the associations between impaired lung function and arrhythmias, most of which were limited in patients with obstructive lung function and have mainly focused on atrial fibrillation, evidence on the risk of overall ventilatory parameters and incident arrhythmias, including VAs and bradyarrhythmia, is scarce [7]. With this in mind, in the present study, we aimed to (1) explore the link between several ventilatory function parameters, including forced expiratory volume in 1 s (FEV1%), forced vital capacity (FVC%) and FEV1/FVC, as well as pulmonary disease phenotypes, with AF/AFL, VAs, high-grade atrioventricular block (high-grade AV block), sick sinus syndrome (SSS), and premature atrial/ventricular complexes (PAC/PVC); (2) measure the spirometry-arrhythmia relationship according to race and ethnicity, sex and smoking status; (3) study dose–response patterns of ventilatory function parameters on the risk of incident arrhythmias; (4) calculate the population attributable fractions (PAFs) of incident arrhythmias associated with different pulmonary disease phenotypes.
Methods
Study population
The study included subjects from two cohorts, the ARIC (Atherosclerosis Risk in Communities) and the CHS (The Cardiovascular Health Study). Details of the study have been described previously [8, 9]. The ARIC study is a community-based, biracial cohort established to investigate atherosclerosis etiology, clinical outcomes, and variations in cardiovascular risk factors, medical care, and disease. From 1987–1989, 15,792 participants aged 45–64 years were recruited from four US communities: Forsyth County, NC; Jackson, MS; suburban Minneapolis, MN; and Washington County, MD. Participants underwent baseline cardiovascular risk assessment and completed four in-person follow-up visits (1990–1992, 1993–1995, 1996–1998, 2011–2013). Semiannual telephone follow-up ascertained study endpoints. The CHS study is a prospective cohort study of coronary heart disease risk factors in older adults. It enrolled community-dwelling adults ≥ 65 years identified via Medicare eligibility lists from four US sites (North Carolina, California, Maryland, and Pennsylvania).An initial cohort (n = 5201) enrolled in 1989/1990, supplemented by a second cohort (n = 687) in 1992/1993. Annual clinic visits assessed demographics, medical history, hospitalizations, and lifestyle through 1998/1999. Continuous semiannual telephone interviews tracked health status, incident events, and mortality. Both of the two studies were approved by the Institutional Review Board, and informed consent signed by all subjects was obtained. The cohort data sets were obtained from the NIH Biologic Specimen and Data Repository Information Coordinating Center (BioLINCC) [10, 11].
Although ARIC and CHS were initiated to study atherosclerosis and cardiovascular risk in aging, respectively, both cohorts systematically collected incident arrhythmia data through standardized ECGs, hospital surveillance, and adjudication. Given the established role of cardiovascular risk factors in arrhythmogenesis, these cohorts provide a well-phenotyped platform for evaluating novel predictors of arrhythmias, including pulmonary function. Following the exclusion of participants with baseline cardiac arrhythmias (88 in ARIC; 62 in CHS), inadequate spirometry data (888 in ARIC; 850 in CHS), and missing follow-up for incident arrhythmias (1,887 in ARIC; 221 in CHS), the final analytical cohort included 12,929 ARIC participants and 4,755 CHS participants. This yielded an aggregate population of 17,684 individuals (9,592 women [54.2%]; 8,092 men [45.8%]).
Assessment of ventilatory parameters
For the present study, the main measures of lung function of interest were FEV1% predicted, FVC% predicted, and FEV1/FVC% predicted. FEV1% predicted was the volume of gas exhaled in the first second of expiration expressed as a percentage of the predicted value based on age, sex, and race and ethnicity according to recommendations from the Epidemiology Standardization Project. FVC% predicted represents the maximal volume of gas exhaled after maximal inspiratory expressed as a percentage of the predicted value. FEV1/FVC% predicted was derived as the ratio between the 2 values. At baseline, spirometry was conducted using a water-sealed Collins Survey Ⅱ volume displacement spirometer (Collins Medical, Inc.) and Pulmo-Screen Ⅱ software (PDS Healthcare Products, 496 Inc), as has been described previously in ARIC [12]. At least 3 acceptable spirograms were obtained from a minimum of 5 forced expirations, and the best single spirogram was identified by a computer and confirmed by a technician. Quality control was conducted carefully throughout the study. The spirometry protocols used in ARIC and CHS followed standardized procedures that are consistent with current American Thoracic Society (ATS) and European Respiratory Society (ERS) recommendations. Therefore, the lung function parameters derived from these cohorts remain directly comparable to those used in present-day clinical and epidemiological practice [13].
Pulmonary disease phenotypes were classified into 4 categories: Obstructive spirometry pattern: (FEV1/FVC ≤ lower limit of normal (LLN) without bronchodilator administration); Restrictive impairment pattern: (FEV1/FVC > LLN and FVC < LLN); Respiratory symptoms with normal spirometric results (without obstructive or restrictive impairment) and normal spirometry (without respiratory symptoms, obstructive, or restrictive impairment) [14].
Assessment of covariates
Interviewers collected information on age, race and ethnicity, sex, smoking status, education level, medical history, and other demographic factors. For smoking history, subjects identified themselves as current, former, or never smokers. Body mass index was calculated by dividing weight (kilograms) by height (meters) squared. In addition, diabetes, hypertension, coronary heart disease, chronic heart failure, and lung diseases at baseline were ascertained based on self-reported diseases and the ICD-9-CM codes. Medication use history was obtained by self-reported medication intake and by reviewing medication brought by subjects to their visit. Each medication was coded by trained and certified interviewers using a computerized medication classification system. The use of calcium antagonist, adrenergic β-agonists, and digoxin was examined as a potential confounder.
Race and ethnicity data were obtained through participant self-identification using fixed categories defined by the original cohort studies. Consistent with epidemiological best practices, these variables were analyzed as social constructs reflecting differential exposure to systemic inequities (e.g., structural racism, healthcare access barriers), not as biological determinants [15].
Outcome ascertainment
Primary outcomes comprised incident arrhythmias, including any arrhythmias, AF/AFL, VAs, high-grade AV block, SSS, and PAC/PVC. Although modern technologies such as wearable monitors have expanded arrhythmia detection, the core definitions of clinically significant arrhythmias-including atrial fibrillation, ventricular tachycardia, and bradyarrhythmia-have remained stable [16, 17]. Incident arrhythmia cases were ascertained from study visit electrocardiograms (ECGs), hospital discharge diagnoses, and death certificates. Using AF/AFL as an illustrative example, a 12-lead resting ECG was obtained at each study examination and transmitted to the ECG Reading Center for automated coding using the Marquette 12-SL program. AF/AFL detected automatically was subsequently adjudicated by a cardiologist. Hospitalizations during follow-up were identified via telephone interviews and surveillance of local hospitals, with trained abstractors collecting discharge diagnoses. The ascertainment of AF/AFL from hospital discharge codes has been validated in epidemiological studies. AF/AFL was defined by ICD-9-CM codes 427.31 (atrial fibrillation) or 427.32 (atrial flutter) in clinical records, or by ICD-9 427.3 or ICD-10 I48 codes listed as underlying or contributing causes of death on death certificates [18, 19]. VAs encompassed ventricular tachycardia (ICD-9-CM: 427.1), ventricular fibrillation/flutter (427.4, 427.41, 427.42), cardiac arrest (427.5), and sudden cardiac death [20]. Sudden cardiac death was defined underlying cardiac origin because of the absence of conditions clearly unrelated to cardiac arrhythmias [21]. High-grade AV block included second-degree (ICD-9-CM: 426.1, 426.10) or complete AV block (426.0) [3]. SSS was classified as ICD-9-CM 427.81 [22]. PAC/PVC were identified using ICD-9-CM codes 427.6, 427.60 (unspecified), 427.61 (atrial), and 427.69 (ventricular or other). The follow-up duration started from the time lung function was measured to the occurrence of the studied outcomes, death, loss to follow-up (informative censoring), or end of follow-up in the original cohorts (administrative censoring).
Statistical analysis
Participant characteristics were described according to quartiles of FEV1% predicted in the pooled cohort. Baseline variables were reported as mean and standard deviation for normally distributed data, and categorical variables were represented by proportion. Baseline characteristics were compared across FEV1% predicted quartiles using the chi-square test and one-way ANOVA test for categorical and continuous variables.
Multivariate Cox proportional hazard models were used to estimate the hazard ratios (HRs) and 95% confidence interval (CI) 1each type of arrhythmias with FEV1% predicted quartiles, FVC% predicted quartiles, and FEV1/FVC% predicted quartiles, respectively. The proportional hazards assumption was tested by plotting log (-log) survival curves and interaction with time, and no significant violation was present. For each arrhythmia endpoint, we ran a multivariate-adjusted model, with adjustment for sex, race and ethnicity, age, education level, history of hypertension, diabetes, prevalent coronary heart disease, heart failure, cigarette smoking, alcohol drinking, total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, triglycerides, fasting glucose, body mass index, systolic blood pressure, diastolic blood pressure, resting heart rate, QTc interval, left ventricular hypertrophy and use of cardiac medications. The covariates in the adjusted model are selected based on the known literature on potential factors that may be associated with abnormalities in cardiac arrhythmias. Kaplan–Meier survival curves were constructed to describe the cumulative incidence of individual outcome with different FEV1% predicted strata. With ventilatory parameters as continuous variables, we used restricted cubic splines with 5 knots placed at the 5th, 27.5th, 50th, 72.5th, and 95th percentile to assess the potential non-linear association of FEV1% predicted, FVC% predicted, and FEV1/FVC% predicted on arrhythmias. Adjusted Cox proportional hazard models were used to evaluate the association between pulmonary disease phenotypes and each type of arrhythmias separately.
In our primary analysis, missing data were addressed using full information maximum likelihood, assuming missingness at random. The proportion of missing values is detailed in Additional file 1: Table S1. The missing data of covariates were imputed via Markov Chain Monte Carlo (MCMC) multiple imputation prior to inclusion in fully adjusted models [23]. The imputation model incorporated total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, triglycerides, fasting glucose, body mass index, systolic blood pressure, diastolic blood pressure, resting heart rate, and QTc interval. Results from 10 imputation cycles were pooled to generate the final estimates.
To better evaluate the impact of different pulmonary disease phenotypes on the studied outcomes, PAFs were calculated using the equation pdi*[(HR-1)/HR] where pdi represents the proportion of total events in the population arising from the ith exposure category in comparison to the normal spirometry [24]. To further verify the robustness of our findings, we performed sensitivity analyses that included: (1) exclusion of participants with prevalent heart failure or coronary heart disease to mitigate confounding by these conditions; (2) restriction to individuals with complete covariate data; (3) reevaluation of adjusted associations between quartiles of FEV1, FVC, and FEV1/FVC with incident arrhythmia risk; (4) incorporation of time-updated covariates (measured during follow-up visits prior to arrhythmia onset) as time-dependent variables in extended Cox regression to address time-varying confounding; and (5) implementation of Fine-Gray subdistribution hazards models treating mortality as a competing risk, with censoring at death dates [25].
Consistent with established arrhythmia disparities across ethnic groups and by sex, stratified analyses were performed to evaluate potential effect modification by race and ethnicity and sex on ventilatory parameter-arrhythmia associations. Given smoking's potential mediating role between pulmonary function and cardiovascular outcomes, subgroup analyses further examined smoking status. Multivariable Cox regression models separately assessed adjusted associations of FEV1% predicted, FVC% predicted, and FEV1/FVC% predicted quartiles with incident arrhythmia risk in ARIC and CHS cohorts. Likelihood ratio tests evaluated interactions between ventilatory parameters and (a) race and ethnicity, (b) sex, (c) smoking status, and (d) cohort. Where significant interactions emerged, subgroup-specific hazard ratios were derived from models incorporating relevant interaction terms [26].
All statistical analyses were performed with the Stata V.15.0. All p values were two-sides and p < 0.05 was considered statistically significant.
Results
Population characteristics
In general, of the 17,684 subjects included in this study, the mean age was 59.4 ± 9.7 years old, 45.76% were male, and 79.76% were White participants. The baseline characteristics of the subjects according to FEV1% predicted quartiles were presented in Table 1. Compared with subjects in the higher quartile of FEV1% predicted, subjects in the lower quartiles were more commonly males and had higher smoking prevalence and higher cardiovascular disease prevalence. Similarly, self-reported pulmonary symptoms and diagnosis were more common among individuals with lower FEV1% predicted quartiles. The use of cardiac medications was more prevalent in subjects with lower quartiles of FEV1% predicted. Baseline characteristics of the study population in ARIC and CHS cohort were presented in Additional file 1: Table S2.Compared with subjects in the ARIC cohort, a higher proportion of subjects were smokers and commonly White participants in the CHS cohort.
Table 1.
Baseline characteristics of the study population across quartiles of FEV1% predicted
| Quartiles of FEV1% predicted | |||||
|---|---|---|---|---|---|
| Characteristics | Q1 (lowest) | Q2 | Q3 | Q4 (highest) | Total |
| Age, yrs | 62.8 (9.9) | 59.2 (9.5) | 57.9 (9.2) | 57.9 (9.2) | 59.4 (9.7) |
| Males, % | 2540 (57.5) | 2267 (51.3) | 1911 (43.2) | 1373 (31.1) | 8091 (45.8) |
| White participants, % | 3589 (81.2) | 3448 (78.0) | 3534 (79.9) | 3532 (79.9) | 14,103 (79.8) |
| Education levels, % | |||||
| Low | 1498 (33.9) | 1210 (27.4) | 970 (21.9) | 871 (19.7) | 4549 (25.7) |
| Intermediate | 1896 (42.9) | 1876 (42.4) | 1985 (44.9) | 1940 (43.9) | 7697 (43.5) |
| High | 1026 (23.2) | 1335 (30.2) | 1466 (33.2) | 1610 (36.4) | 5437 (30.8) |
| Smoking status, % | |||||
| Never | 1064 (24.1) | 1678 (38.0) | 2082 (47.1) | 2520 (57.0) | 7344 (41.5) |
| Former | 1736 (39.3) | 1574 (35.6) | 1527 (34.5) | 1395 (31.6) | 6232 (35.2) |
| Current | 1620 (36.7) | 1169 (26.4) | 812 (18.4) | 506 (11.5) | 4107 (23.2) |
| Alcohol use, % | 2376 (54.8) | 2416 (54.7) | 2483 (56.2) | 2424 (54.8) | 9699 (54.9) |
| Body mass index, kg/m2 | 27.4 (5.3) | 27.8 (5.3) | 27.6 (5.1) | 27.2 (4.8) | 27.5 (5.1) |
| Systolic blood pressure, mmHg | 131.6 (21.9) | 127.3 (20.7) | 124.5 (19.8) | 123.0 (19.3) | 126.6 (20.7) |
| Diastolic blood pressure, mmHg | 72.8 (12.1) | 73.4 (11.6) | 73.1 (11.2) | 72.6 (10.7) | 73.0 (10.7) |
| Clinical history, % | |||||
| Hypertension | 2013 (45.5) | 1788 (40.4) | 1619 (36.6) | 1406 (31.8) | 6826 (38.6) |
| Diabetes mellitus | 774 (17.5) | 709 (16.0) | 539 (12.2) | 417 (9.4) | 2439 (13.8) |
| Coronary heart disease | 699 (15.8) | 450 (10.2) | 301 (6.8) | 207 (4.7) | 1657 (9.4) |
| Chronic heart failure | 379 (8.6) | 229 (5.2) | 172 (3.9) | 116 (2.6) | 896 (5.1) |
| Laboratory findings | |||||
| Total cholesterol, mmol/L | 5.5 (1.1) | 5.5 (1.1) | 5.6 (1.1) | 5.6 (1.1) | 5.6 (1.1) |
| LDL-cholestrol, mmol/L | 3.4 (1.1) | 3.5 (1.1) | 3.6 (1.0) | 3.6 (1.1) | 3.5 (1.1) |
| HDL-cholestrol, mmol/L | 1.3 (0.4) | 1.3 (0.4) | 1.3 (0.4) | 1.4 (0.4) | 1.3 (0.4) |
| Triglycerides, mmol/L | 1.6 (1.1) | 1.6 (1.0) | 1.5 (1.0) | 1.4 (0.9) | 1.6 (1.0) |
| Fasting blood glucose, mmol/L | 6.4 (2.5) | 6.2 (2.3) | 6.1 (2.2) | 5.9 (1.9) | 6.1 (2.3) |
| Electrographic findings | |||||
| Resting hear rate, bpm | 67.2 (11.0) | 66.1 (10.5) | 65.9 (9.9) | 66.1 (10.0) | 66.3 (10.4) |
| QTc, ms | 421.1 (46.5) | 417.5 (44.8) | 415.6 (49.1) | 416.3 (46.2) | 417.6 (46.7) |
| Left ventricular hypertrophy, % | 185 (4.2) | 137 (3.1) | 98 (2.2) | 79 (1.8) | 499 (2.8) |
| Spirometry | |||||
| FEV1% predicted | 64.7 (13.7) | 86.1 (3.5) | 97.4 (3.1) | 112.2 (8.3) | 90.1 (19.3) |
| FVC% predicted | 78.6 (13.5) | 93.3 (8.8) | 102.4 (8.5) | 115.0 (10.6) | 97.3 (16.9) |
| FEV1/FVC% predicted | 84.9 (17.0) | 96.1 (10.9) | 98.6 (9.4) | 101.0 (9.2) | 95.1 (13.6) |
| FEV1, L | 1.9 (0.6) | 2.5 (0.6) | 2.8 (0.7) | 3.1 (0.8) | 2.6 (0.8) |
| FVC, L | 3.0 (0.9) | 3.5 (0.9) | 3.8 (1.0) | 4.0 (1.0) | 3.6 (1.0) |
| FEV1/ FVC, % | 64.7 (11.9) | 73.4 (6.6) | 75.9 (5.5) | 77.8 (5.2) | 73.0 (9.3) |
| Self-reported symptoms, % | |||||
| Cough | 1061 (24.0) | 579 (13.1) | 435 (9.8) | 338 (7.7) | 2413 (13.7) |
| Phlegm | 1137 (25.7) | 615 (13.9) | 477 (10.8) | 402 (9.1) | 2631 (14.9) |
| Dyspnea | 1285 (29.1) | 703 (15.9) | 554 (12.5) | 436 (9.9) | 2978 (16.8) |
| Self-reported diagnosis, % | |||||
| Bronchitis | 964 (21.8) | 549 (12.4) | 447 (10.1) | 382 (8.6) | 2342 (13.2) |
| Asthma | 459 (10.4) | 258 (5.8) | 171 (3.9) | 148 (3.4) | 1036 (5.9) |
| Emphysema | 341 (7.7) | 71 (1.6) | 47 (1.1) | 23 (0.5) | 482 (2.7) |
| Cardiac medication, % | |||||
| Calcium antagonist | 403 (9.1) | 270 (6.1) | 223 (5.0) | 189 (4.3) | 1085 (6.1) |
| Digoxin | 291 (6.6) | 153 (3.5) | 109 (2.5) | 99 (2.2) | 652 (3.7) |
| β-Blocker | 474 (10.7) | 375 (8.5) | 305 (6.9) | 244 (5.5) | 1398 (7.9) |
Abbreviations: LDL Low-density lipoprotein, HDL High-density lipoprotein, FEV1 Forced expiratory volume in 1 s, FVC Forced vital capacity
Values are mean ± SD or number (percentage) of subjects
Ventilatory parameters and incident arrhythmias
Among the 17,684 study subjects, we identified 7408 cases of any arrhythmias, 4215 of AF/AFL, 1238 of VAs, 698 of high-grade AV block, 634 of SSS, and 1377 of PAC/PVC during a median follow-up period of 12.6 years. By study termination, 9,018 deaths (51.0%), 856 losses to follow-up (4.8%), and 7,810 survivors (44.2%) were documented, with the latter group completing the full follow-up period. As compared with those in the highest quartile (Q4), subjects in the lowest FEV1% predicted quartile (Q1) experienced significantly higher risk of any arrhythmias (HR:1.32[95% CI:1.23–1.42), AF/AFL (HR:1.68[95% CI:1.52–1.85), VAs (HR:1.55[95% CI:1.29–1.86), high-grade AV block (HR:1.37[95% CI:1.08–1.73), and PAC/PVC (HR:1.42[95% CI:1.20–1.69), respectively (Table 2). Cumulative incidence functions also revealed a monotonic association between FEV1% predicted and risk of incident arrhythmias throughout the study period (Fig. 1).
Table 2.
Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function*
| Quartile 1(Lowest) | Quartile 2 | Quartile 3 | Quartile 4 (Highest) | Trend (per 1-SD decrease) | |
|---|---|---|---|---|---|
| FEV1% Predicted | |||||
| Subjects, n | 4421 | 4421 | 4421 | 4421 | |
| Range of FEV1% Predicted (mean) | 11.0–79.6 (64.7) | 79.6–92.0 (86.1) | 92.0–103.0 (97.4) | 103.0–266.0 (112.2) | |
| Any arrhythmias, n | 2249 | 1909 | 1747 | 1503 | |
| Hazard ratio (95% CI) | 1.32 (1.23–1.42) | 1.15 (1.07–1.24) | 1.11 (1.04–1.20) | 1.00 | 1.11 (1.08–1.14) |
| AF/AFL, n | 1367 | 1086 | 962 | 800 | |
| Hazard ratio (95% CI) | 1.68 (1.52–1.85) | 1.26 (1.14–1.38) | 1.17 (1.07–1.29) | 1.00 | 1.24 (1.20–1.28) |
| VAs, n | 468 | 312 | 263 | 195 | |
| Hazard ratio (95% CI) | 1.55 (1.29–1.86) | 1.15 (0.96–1.39) | 1.14 (0.94–1.37) | 1.00 | 1.21 (1.14–1.29) |
| SSS, n | 158 | 174 | 169 | 133 | |
| Hazard ratio (95% CI) | 1.14 (0.89–1.47) | 1.25 (0.99–1.57) | 1.25 (0.99–1.57) | 1.00 | 1.02 (0.93–1.11) |
| High-grade AV block, n | 240 | 180 | 155 | 123 | |
| Hazard ratio (95% CI) | 1.37 (1.08–1.73) | 1.15 (0.91–1.46) | 1.12 (0.88–1.42) | 1.00 | 1.11 (1.02–1.20) |
| PAC/PVC, n | 464 | 360 | 320 | 233 | |
| Hazard ratio (95% CI) | 1.42 (1.20–1.69) | 1.23 (1.03–1.45) | 1.22 (1.02–1.44) | 1.00 | 1.16 (1.10–1.23) |
| FVC% Predicted | |||||
| Subjects, n | 4421 | 4421 | 4421 | 4421 | |
| Range of FVC% Predicted (mean) | 18.0–87.0 (75.8) | 87.0–98.0 (92.8) | 98.0–108.3 (103.0) | 108.3–311.0 (117.7) | |
| Any arrhythmias, n | 2212 | 1950 | 1720 | 1526 | |
| Hazard ratio (95% CI) | 1.26 (1.17–1.35) | 1.13 (1.06–1.21) | 1.07 (1.00–1.15) | 1.00 | 1.09 (1.06–1.12) |
| AF/AFL, n | 1345 | 1094 | 958 | 818 | |
| Hazard ratio (95% CI) | 1.48 (1.35–1.63) | 1.17 (1.06–1.28) | 1.11 (1.01–1.22) | 1.00 | 1.19 (1.15–1.23) |
| VAs, n | 452 | 338 | 248 | 200 | |
| Hazard ratio (95% CI) | 1.50 (1.25–1.79) | 1.21 (1.01–1.45) | 1.04 (0.86–1.26) | 1.00 | 1.19 (1.12–1.27) |
| SSS, n | 171 | 167 | 153 | 143 | |
| Hazard ratio (95% CI) | 1.06 (0.84–1.35) | 1.04 (0.82–1.31) | 1.02 (0.81–1.29) | 1.00 | 1.04 (0.95–1.13) |
| High-grade AV block, n | 253 | 175 | 169 | 101 | |
| Hazard ratio (95% CI) | 1.65 (1.29–2.12) | 1.29 (1.01–1.67) | 1.42 (1.11–1.82) | 1.00 | 1.16 (1.07–1.25) |
| PAC/PVC, n | 452 | 362 | 316 | 247 | |
| Hazard ratio (95% CI) | 1.31 (1.11–1.55) | 1.14 (0.96–1.35) | 1.12 (0.95–1.33) | 1.00 | 1.10 (1.03–1.16) |
| FEV1/FVC% Predicted | |||||
| Subjects, n | 4421 | 4421 | 4421 | 4421 | |
| Range of FEV1/FVC% Predicted (mean) | 27.0–88.8 (78.5) | 88.8–95.8 (92.5) | 95.8–102.3 (98.8) | 102.3–160.0 (111.0) | |
| Any arrhythmias, n | 2144 | 1879 | 1801 | 1584 | |
| Hazard ratio (95% CI) | 1.17 (1.08–1.26) | 1.04 (0.96–1.13) | 1.00 (0.92–1.08) | 1.00 | 1.07 (1.04–1.10) |
| AF/AFL, n | 1283 | 1102 | 993 | 837 | |
| Hazard ratio (95% CI) | 1.33 (1.20–1.47) | 1.08 (0.97–1.19) | 0.98 (0.89–1.09) | 1.00 | 1.16 (1.12–1.20) |
| VAs, n | 415 | 305 | 251 | 267 | |
| Hazard ratio (95% CI) | 1.26 (1.04–1.53) | 1.03 (0.85–1.26) | 0.88 (0.72–1.07) | 1.00 | 1.13 (1.06–1.20) |
| SSS, n | 157 | 160 | 194 | 123 | |
| Hazard ratio (95% CI) | 1.18 (0.89–1.57) | 1.14 (0.87–1.50) | 1.38 (1.07–1.79) | 1.00 | 0.97 (0.87–1.07) |
| High-grade AV block, n | 204 | 153 | 191 | 150 | |
| Hazard ratio (95% CI) | 1.09 (0.85–1.40) | 0.86 (0.66–1.11) | 1.09 (0.86–1.39) | 1.00 | 0.98 (0.90–1.07) |
| PAC/PVC, n | 444 | 354 | 343 | 236 | |
| Hazard ratio (95% CI) | 1.29 (1.07–1.56) | 1.11 (0.92–1.34) | 1.12 (0.93–1.35) | 1.00 | 1.15 (1.08–1.22) |
Abbreviations: FEV1,forced expiratory volume in 1 s, FVC forced vital capacity, AF/AFL atrial fibrillation/flutter, VAs ventricular arrhythmias, SSS sick sinus syndrome, AV atrioventricular, PAC/PVC premature atrial/ventricular complexes, SD standard deviation
*Cox regression adjustment for sex, race and ethnicity, age, education level, history of hypertension, diabetes, prevalent coronary heart disease, heart failure, cigarette smoking, alcohol drinking, total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, triglycerides, fasting glucose, body mass index, systolic blood pressure, diastolic blood pressure, resting heart rate, QTc interval, left ventricular hypertrophy, use of calcium antagonist, digoxin, and β-blocker
Fig. 1.
Cumulative incidence of incident arrhythmias by quartiles of FEV1% predicted. Q1 is the lowest quartile, and Q4 is the highest quartile. Curves were compared using a log-rank test. Abbreviations: FEV1, forced expiratory volume in 1 s; AF/AFL, atrial fibrillation/flutter; VAs, ventricular arrhythmias; SSS, sick sinus syndrome; AV, atrioventricular; PAC/PVC, premature atrial/ventricular complexes
Similar patterns of association were also noted in the FVC% predicted quartiles, with higher risks of any arrhythmias, AF/AFL, VAs, high-grade AV block and PAC/PVC observed among subjects in the lowest FVC% predicted quartile. However, neither the FEV1% predicted nor FVC% predicted seems to be associated with risk of SSS. Notably, although there was no monotonic association between FEV1/FVC% predicted and risk of SSS, subjects with slightly decreased FEV1/FVC% predicted (Q3) did seem to have increased risk of SSS than those in the highest quartile (Q4) (HR:1.38[95% CI:1.07–1.79) (Table 2).
The graded risk-decreasing association of ventilatory parameters with any arrhythmias, AF/AFL, VAs, high-grade AV block and PAC/PVC were further observed in the dose–response analysis (Fig. 2).The adjusted HRs of 1-SD decreases of FEV1% predicted were 1.11 [95% CI, 1.08–1.14] for any arrhythmias, 1.24 [95% CI, 1.20–1.28] for AF/AFL, 1.21 [95% CI, 1.14–1.29] for VAs, 1.11 [95% CI, 1.02–1.20] for high-grade AV block and 1.16 [95% CI, 1.10–1.23] for PAC/PVC (Table 2). Similar results were obtained for FVC% predicted and FEV1/FVC% predicted.
Fig. 2.
Dose–response analysis of FEV1% predicted, FVC% predicted, and FEV1/FVC% predicted and risk of incident arrhythmias. The curves (solid or dotted lines) are plotted using restricted cubic splines and are presented together with 95% CIs (corresponding shaded area). Adjusted for sex, race and ethnicity, and age, plus education level, history of hypertension, diabetes, prevalent coronary heart disease, heart failure, cigarette smoking, alcohol drinking, total cholesterol, high-density lipoprotein cholesterol, low density lipoprotein cholesterol, triglycerides, fasting glucose, body mass index,systolic blood pressure, diastolic blood pressure, resting heart rate, QTc interval,left ventricular hypertrophy, use of calcium antagonist, digoxin, and β-Blocker. Abbreviations: FEV1, forced expiratory volume in 1 s; FVC, forced vital capacity; AF/AFL, atrial fibrillation/flutter; VAs, ventricular arrhythmias; SSS, sick sinus syndrome; AV, atrioventricular; PAC/PVC, premature atrial/ventricular complexes
Sensitivity analyses confirmed the robustness of our findings. Consistent risk estimates were observed after excluding participants with prevalent coronary heart disease or chronic heart failure (Additional file 1: Table S3), and restricting analyses to individuals with complete covariate data yielded materially unchanged results (Additional file 1: Table S4). Comparable associations emerged when evaluating alternative pulmonary function metrics-including FEV1, FVC and FEV1/FVC (Additional file 1: Table S5). Extended Cox regression incorporating time-dependent covariates, including incident antiarrhythmic medication use, addressed time-varying confounding, demonstrating results concordant with primary analyses (Additional file 1: Table S6). Similarly, Fine-Gray subdistribution hazards regression accounting for mortality as a competing risk produced materially unaltered effect estimates despite modest attenuation of hazard ratios (Additional file 1: Table S7).
FEV1% predicted demonstrated significant multiplicative interactions with study cohort (P < 0.001), race and ethnicity (P < 0.001) on arrhythmia risk, but not with sex (P = 0.25) or smoking status (P = 0.51). Stratified analyses revealed that each 1-SD decrement in FEV1% predicted increased any arrhythmias in both CHS (HR 1.16, 95% CI 1.11–1.20) and ARIC (HR 1.08, 95% CI 1.05–1.11) cohorts, with greater effect magnitude in CHS versus ARIC (interaction HR 1.10, 95% CI 1.05–1.15) (Additional file 1: Table S8). Similarly, FEV1% predicted decrements conferred higher risk of any arrhythmias in White versus Black participants (White participants: HR 1.13, 95% CI 1.10–1.16; Black participants: HR 1.00, 95% CI 0.94–1.07; ratio of HRs 1.12, 95% CI 1.06–1.19) (Additional file 1: Table S9). In addition, both FEV1% predicted and FVC% predicted were consistently associated with increased risk of AF/AFL, VAs, high-grade AV block and PAC/PVC in White participants; while in Black participants, FEV1% predicted and FVC% predicted were only associated with risk of AF/AFL (Additional file 1: Table S9). Sex-stratified and smoking status-stratified results are presented in Additional file 1: Table S10-11.
Pulmonary disease phenotypes and pncident prrhythmias
The Kaplan–Meier curves in Fig. 3 showed a clear association between pulmonary disease phenotypes and types of arrhythmias studied. The association between almost all types of arrhythmias was most pronounced with the obstructive spirometry pattern, except for SSS and high-grade AV block (Fig. 4). The presence of the obstructive spirometry pattern, compared with normal spirometry, was after multivariable adjustments associated with HRs of 1.28 (95% CI, 1.21–1.36) for any arrhythmias, 1.41 (95% CI, 1.30–1.53) for AF/AFL, 1.50 (95% CI, 1.30–1.73) for VAs and 1.30 (95% CI, 1.13–1.49) for PAC/PVC. Despite the magnitudes of association were smaller than those with obstructive spirometry pattern, subjects with restrictive impairment pattern experienced increased risk of any arrhythmias (HR: 1.16 [95% CI, 1.07–1.26]), AF/AFL (HR: 1.24 [95% CI, 1.11–1.40]), VAs (HR: 1.22 [95% CI, 1.01–1.50]) and high-grade AV block (HR: 1.50 [95% CI, 1.18–1.92]). Notably, respiratory symptoms with normal spirometric results was not associated with incident arrhythmias. We observed no significant relationship between different pulmonary disease phenotypes and SSS (Fig. 4).
Fig. 3.
Cumulative incidence of incident arrhythmias by pulmonary disease phenotypes. Abbreviations: AF/AFL, atrial fibrillation/flutter; VAs, ventricular arrhythmias; SSS, sick sinus syndrome; AV, atrioventricular; PAC/PVC, premature atrial/ventricular complexes
Fig. 4.
Hazard ratios of incident arrhythmias associated with pulmonary disease phenotypes. Adjusted for sex, race and ethnicity, and age, plus education level, history of hypertension, diabetes, prevalent coronary heart disease, heart failure, cigarette smoking, alcohol drinking, total cholesterol, high-density lipoprotein cholesterol, low density lipoprotein cholesterol, triglycerides, fasting glucose, body mass index,systolic blood pressure, diastolic blood pressure, resting heart rate, QTc interval,left ventricular hypertrophy, use of calcium antagonist, digoxin, and β-blocker. Abbreviations: AF/AFL, atrial fibrillation/flutter; VAs, ventricular arrhythmias; SSS, sick sinus syndrome; AV, atrioventricular; PAC/PVC, premature atrial/ventricular complexes
Pulmonary disease phenotypes on any arrhythmias were significantly modified by race and ethnicity (P for interaction = 0.001), but not by study cohorts (P = 0.06), sex category (P = 0.50), and smoking status (P = 0.43). In White participants, obstructive spirometry pattern and restrictive impairment pattern were associated with all types of incident arrhythmias, except for SSS; while in Black participants, only restrictive impairment pattern was associated with increased risk of AF/AFL (Additional file 1: Table S12). Obstructive spirometry pattern conferred a greater risk of any arrhythmias in White versus Black participants (ratio of HRs 1.15, 95% CI 1.02–1.30), indicating race and ethnicity-specific differential risk. Stratified results by study cohort, sex, and smoking status are presented in Additional file 1: Table S13-15.
To better understand the impact of pulmonary disease on the studied outcomes, PAFs were calculated and are presented in Table 3. Generally, the attributable effect of pulmonary disease accounted for 10.7% (7.8%−13.5%) of any arrhythmias, 16.7% (12.4%−20.8%) of AF/AFL, 16.1% (9.0%−22.8%) of VAs, 10.1% (0–19.4%) of high-grade AV block, and 9.4% (2.3%−17.0%) of PAC/PVC. With regard to pulmonary disease phenotypes, obstructive spirometry pattern, followed by restrictive impairment pattern were major contributing risk factors. Obstructive spirometry pattern was responsible for 18.7% (14.4%−22.7%) of AF/AFL, 18.2% (11.5%−24.3%) of VAs, 11.3% (8.5%−13.9%) of any arrhythmias, and 10.9% (4.3%−17.0%) of PAC/PVC. Of note, restrictive impairment pattern could also explain 10.5% (3.4–17.0%) of high-grade AV block, and 5.0% (1.4%−8.3%) of AF/AFL.
Table 3.
Population attributable fraction (%) and 95% confidence intervals for incident arrhythmias across pulmonary disease phenotypes
| Respiratory symptoms with normal spirometric results | Restrictive impairment pattern | Obstructive spirometry pattern | Pooled | |
|---|---|---|---|---|
| Any arrhythmias | 1.7 (0–3.7) | 3.0 (1.1–4.8) | 11.3 (8.5–13.9) | 10.7 (7.8–13.5) |
| AF/AFL | 2.0 (0–5.1) | 5.0 (1.4–8.3) | 18.7 (14.4–22.7) | 16.7 (12.4–20.8) |
| VAs | 0.9 (0–0.6) | 4.7 (0–9.5) | 18.2 (11.5–24.3) | 16.1 (9.0–22.8) |
| SSS | 0 | 1.6 (0–7.6) | 0 | 0 |
| High-grade AV block | 2.2 (0–8.7) | 10.5 (3.4–17.0) | 5.4 (0–14.5) | 10.1 (0–19.4) |
| PAC/PVC | 0 | 3.2 (0–7.7) | 10.9 (4.3–17.0) | 9.4 (2.3–17.0) |
Abbreviations: AF/AFL atrial fibrillation/flutter, VAs ventricular arrhythmias, SSS sick sinus syndrome, AV atrioventricular, PAC/PVC premature atrial/ventricular complexes
Discussion
This study characterized relationships between ventilatory function indices and incident arrhythmias. Four main observations were noted. Baseline pulmonary function showed an inverse relationship with incident arrhythmic events (excluding SSS), with this pattern being seen across different sex and smoking status groups. White participants with pulmonary impairment displayed higher arrhythmia risk compared to Black participants. The obstructive spirometry pattern appeared as the primary risk factor for incident arrhythmias, with restrictive ventilatory impairment patterns following in significance. Pulmonary dysfunction contributed to a proportion of incident arrhythmias, including AF/AFL, VAs, high-grade AV block and PAC/PVC. To our knowledge, this is the first detailed examination of the relationship between progressive pulmonary function decline and various arrhythmia subtypes. The analysis combines mechanistic and epidemiological viewpoints, addressing an important gap in knowledge regarding cardiopulmonary interactions across different arrhythmia types.
Previous studies have suggested impaired lung function as an independent correlate of AF/AFL, with COPD potentially amplifying risks for AF/AFL progression, recurrence, and adverse prognosis [27, 28]. In this context, our findings indicate that restrictive pulmonary impairment may represent a novel and independent predictor of incident AF/AFL, broadening the epidemiological perspective beyond COPD-centric associations. Mechanistically, we propose a tripartite pathophysiological framework potentially connecting pulmonary dysfunction to AF/AFL pathogenesis: (1) Chronic systemic inflammation/oxidative stress: Hypoxemia-associated activation of inflammatory cytokines (e.g., IL-6, TNF-α) and reactive oxygen species generation may contribute to atrial remodeling, potentially facilitating AF/AFL initiation [29]. (2) Autonomic dysregulation: COPD-related hypercapnia and hypoxemia could promote sympathetic dominance, potentially accelerating atrial ectopy and electrical instability, while concurrently associating with pulmonary vasoconstriction—a recognized risk factor for pulmonary hypertension and right ventricular diastolic dysfunction [30].(3) Pulmonary-ventricular interaction: Progressive pulmonary vascular remodeling may exacerbate right ventricular pressure overload, potentially promoting biatrial stretch and serving as a substrate for arrhythmogenesis [31].
While previous studies have primarily examined COPD in the context of ventricular tachycardia or sudden cardiac death, research on ventricular arrhythmias (VAs) as a composite endpoint remains limited. Our study extends this field through three key observations: Restrictive ventilatory patterns showed associations with VAs independent of smoking status or comorbidities, suggesting their potential as a novel predictor; All spirometric indices demonstrated graded inverse relationships with ventricular arrhythmia incidence, consistent with pulmonary dysfunction representing a spectrum of arrhythmia risk; Respiratory pathophysiology accounted for 16.1% of clinically observed VAs, indicating pulmonary impairment may represent a modifiable factor in arrhythmogenesis. We further propose a conceptual framework in which chronic hypoxia, systemic inflammation, and sympathetic hyperactivation may collectively contribute to myocardial electrical instability [32, 33]. This work positions pulmonary dysfunction beyond a comorbidity, highlighting its observed association with ventricular arrhythmogenesis and suggesting that integrated cardiopulmonary risk assessment merits further investigation.
Analyses examining potential links between lung function and bradyarrhythmia remain relatively limited. This study appears to be the first to explore associations between pulmonary dysfunction and both high-grade atrioventricular (AV) block and sick sinus syndrome (SSS). Key observations include: Declines in FEV1 and FVC showed dose-dependent patterns with high-grade AV block incidence, whereas no significant correlations were observed with SSS; Among pulmonary pathologies, restrictive ventilatory impairment was the only pattern demonstrating independent associations with high-grade AV block, contrasting with non-significant findings for COPD; Restrictive ventilatory patterns accounted for 10.5% of incident high-grade AV block cases, suggesting their potential contribution distinct from traditional ischemic or degenerative pathways. We hypothesize that restrictive lung disease might influence high-grade AV block development through several potential mechanisms: hypoxia-mediated fibrosis in the conduction system, chronic inflammatory interactions, and altered intrathoracic pressure dynamics affecting atrial mechanoelectrical feedback [34, 35]. These findings indicate an association between impaired lung function (particularly restrictive patterns) and incident high-grade AV block, supporting consideration of pulmonary assessment in patients with conduction disorders-especially where pulmonary pathology may be relevant. Further research could determine whether targeted pulmonary evaluation benefits specific subgroups of conduction disorder patients.
Our findings highlight pulmonary dysfunction as a potentially modifiable factor associated with cardiac ectopy, suggesting that prevention strategies incorporating FEV1% predicted may merit consideration. Our study observed novel electrophysiological patterns between pulmonary dysfunction and premature PAC/PVC occurrence: PAC/PVC burden showed an inverse relationship with FEV1% predicted quartiles, with the lowest quartile exhibiting 42% higher ectopic burden compared to the highest quartile; Patients with obstructive spirometry patterns demonstrated a 30% increased PAC/PVC incidence relative to those with normal spirometry, independent of smoking or coronary disease. Impaired pulmonary function may contribute to ectopic activity through several potential pathways: Ventilation-perfusion mismatch and pulmonary hypertension could associate with elevated atrial pressure, potentially altering atrial electrophysiological properties [36]. Additionally, hypoxia, hypercapnia, and heightened adrenergic activity-frequently observed in lung dysfunction-are recognized factors that may influence cardiac automaticity [34]. Pulmonary hypertension has also been linked to right atrial enlargement, which might further contribute to automaticity development [37]. Notably, COPD patients using BiPAP therapy showed reduced PAC/PVC incidence compared to conventionally treated counterparts, with improvements potentially mediated through autonomic modulation [38].These findings suggest pulmonary dysfunction as a potentially modifiable factor associated with cardiac ectopy, indicating that prevention strategies incorporating FEV1% predicted could be considered for further evaluation.
Our study has a number of strengths. First, this investigation leverages two large, multiethnic prospective cohorts, providing robust statistical power for sensitivity analyses across arrhythmia subtypes and stratified subgroup analyses by race and ethnicity, sex, smoking status, and pulmonary impairment phenotypes. Second, by restricting primary analyses to never-smokers and validating findings in smoking-stratified models, we disentangle pulmonary pathophysiology from tobacco-related confounders-a critical advancement over prior studies conflating these effects. Third, integration of black and white cohorts reveals race and ethnicity-specific pulmonary-arrhythmic risk gradients, challenging universal pathophysiological models.
The study results should be interpreted along with several limitations. First, despite extensive covariate adjustments, unmeasured factors (e.g., subclinical sleep apnea) may partially mediate observed associations. Second, while baseline ventilatory parameters predicted arrhythmia onset, longitudinal spirometry trajectories-potentially reflecting cumulative cardiopulmonary injury-require dedicated analysis in future trials. Third, absence of quantitative CT-based emphysema scoring precluded assessment of structural lung damage-arrhythmia dose responses. Emerging AI-driven parenchymal analysis tools could bridge this gap. Fourth, this study was limited to individuals aged 45 to 92 years; therefore, the influence of lung function outside this age range could not be assessed. Additionally, because the study population comprised only White and Black participants, the findings may not be generalizable to other racial/ethnic groups. Fifth, our results are based on ARIC and CHS data, which were collected in the 1980s-1990s. While diagnostic and therapeutic approaches have indeed evolved since then, the key physiological measures in our analysis-such as FEV1% predicted and FVC% predicted-were obtained using standardized spirometry protocols that remain consistent with current ATS/ERS guidelines. Likewise, the identification of arrhythmia outcomes, including atrial fibrillation and ventricular tachyarrhythmias, relied on clinically adjudicated events and ICD codes whose definitions have remained largely stable over time. Thus, both the exposure and outcome assessments retain relevance in modern clinical settings. Moreover, the fundamental pathophysiological mechanisms linking pulmonary impairment to arrhythmic risk-such as autonomic dysfunction, hypoxia, and myocardial remodeling-are time-independent and biologically plausible. While the absolute event rates or treatment patterns may differ today, the associations we observed provide important insights that merit validation in newer cohorts. Sixth, the inherent differences in cohort design, population characteristics, and data collection protocols still introduce heterogeneity for ARIC and CHS study. However, we performed cohort-stratified analyses, and report consistent associations within each cohort, although effect sizes varied. We tested for interaction terms, including between cohort and FEV1% predicted, with findings suggesting effect modification but not invalidation. Furthermore, we conducted sensitivity analyses including Fine-Gray competing risk models and time-updated covariates, all of which yielded consistent results, reinforcing the robustness of our conclusions. Finally, incident arrhythmias were ascertained via hospitalization records and study ECGs, potentially undercapturing events diagnosed solely in outpatient primary care (e.g., paroxysmal AF or NSVT). Future studies with continuous ambulatory monitoring are needed to evaluate subclinical arrhythmias.
Conclusions
Impaired pulmonary function shows observable associations with incident cardiac arrhythmias, including AF/AFL, ventricular arrhythmias (VAs), high-grade AV block, and PAC/PVC. These associations were still observed after accounting for established arrhythmia risk factors, including traditional cardiovascular comorbidities. However, given the reliance on historical cohorts (ARIC/CHS) with data collected in the 1980s-1990s, the generalizability of these findings to contemporary clinical practice requires validation in modern populations. Our findings indicate that pulmonary ventilation metrics may represent potential biomarkers worthy of further investigation for risk stratification and preventive approaches in arrhythmia development.
Supplementary Information
Additional file 1: Tables S1-S15. Table S1. Proportion of missing values in the main analysis. Table S2. Baseline characteristics of the study population in ARIC cohort, and CHS cohort. Table S3. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function, excluding prevalent coronary heart disease or heart failure. Table S4. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function with complete data. Table S5. Hazard ratios and 95% confidence intervals for incident arrhythmias according to FEV1, FVC and FEV1/FVC. Table S6. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function, using time-dependent covariates. Table S7. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function with death as competing risk. Table S8. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function, ARIC cohort and CHS cohort. Table S9. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function, stratified by race and ethnicity. Table S10. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function, stratified by sex. Table S11. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function, stratified by smoking status. Table S12. Hazard ratios and 95% confidence intervals for incident arrhythmias according to lung disease categories, by race and ethnicity. Table S13. Hazard ratios and 95% confidence intervals for incident arrhythmias according to lung disease categories, ARIC cohort and CHS cohort. Table S13. Hazard ratios and 95% confidence intervals for incident arrhythmias according to lung disease categories, ARIC cohort and CHS cohort. Table S15. Hazard ratios and 95% confidence intervals for incident arrhythmias according to lung disease categories, stratified by smoking status.
Additional file 2. STROBE Statement.Checklist of items that were included in reports of the cohort studies.
Acknowledgements
We would like to thank the staff and participants of the ARIC and CHS studies for their important contributions.
Abbreviations
- ARIC
Atherosclerosis Risk in Communities
- CHS
Cardiovascular Health Study
- COPD
Chronic obstructive pulmonary disease
- ECG
Electrocardiogram
- AF
Atrial fibrillation
- AFL
Atrial flutter
- VA
Ventricular arrhythmia
- AV
Atrioventricular
- PAC
Premature atrial complexes
- PVC
Premature ventricular complexes
- SSS
Sick sinus syndrome
- FEV1
Forced expiratory volume in 1 s
- FVC
Forced vital capacity
- HR
Hazard ratio
- CI
Confidence interval
- PAF
Population attributable fraction
Authors' contributions
Designed the experiments: YJC LPQ YJL SLL MPL QH JBF JCL WJ D LJL. Analyzed the data: LPQ YJC YJL WJD. Wrote the manuscript: LPQ YJC LJL. All authors contributed to the revision of the manuscript by providing comments on the analyses and interpretation of the findings. All authors read and approved the final manuscript.
Funding
The study was financially supported by the grants from National Natural Science Foundation of China (82270333;81600260), the Natural Science Foundation of Guangdong Province, China (2024A1515013067;2022A1515012358), Guangzhou Science and Technology Program (2024B03J1344), High-level Talents Introduction Plan of Guangdong Provincial People’s Hospital (KY012023007), Clinical Research Special Fund of Guangdong Medical Association (2024HY-A6002), and National Science and Technology Innovation Major Project-Research Project on Prevention and Treatment of Cancer, Cardiovascular, Respiratory and Metabolic Diseases (2023ZD0504202; 2023ZD0504204).
Data Availability
The datasets used and/or analyzed during the current study are available from the BioLINCC website on reasonable application.
No datasets were generated or analysed during the current study.
Declarations
Ethics approval and consent to participate
Each of the studies included in this investigation has already received ethical approval from an individual institutional review board, and all participants provided written informed consent. This research has been conducted using publicly available datasets. De-identified data were used, and no additional ethical approval was required.
Consent for publication
Not applicable. This manuscript does not contain any individual person’s data in any form.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Yun-Jiu Cheng, Li-Ping Qu, Yi-Jian Liao, and Si-Long Lu contributed equally to this work.
Contributor Information
Yun-Jiu Cheng, Email: cheng831011@sina.com.
Wen-Juan Duan, Email: duanwenjuan@gdph.org.cn.
Li-Juan Liu, Email: liulj35@mail.sysu.edu.cn.
References
- 1.Ko D, Chung MK, Evans PT, Benjamin EJ, Helm RH. Atrial fibrillation: a review. JAMA. 2025;333(4):329–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Marijon E, Narayanan K, Smith K, Barra S, Basso C, Blom MT, et al. The Lancet commission to reduce the global burden of sudden cardiac death: a call for multidisciplinary action. Lancet. 2023;402(10405):883–936. [DOI] [PubMed] [Google Scholar]
- 3.Zhang J, Liu J, Ye M, Zhang M, Yao F, Cheng Y. Incidence and risk factors associated with atrioventricular block in the general population: the atherosclerosis risk in communities study and Cardiovascular Health Study. BMC Cardiovasc Disord. 2024;24(1): 509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Bhatt SP, Wu C, Sun Y, Balte PP, Schwartz JE, Divo MJ, et al. Low lung function is associated with high population attributable fraction for cardiovascular mortality. Ann Am Thorac Soc. 2025;22(3):359–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pham HN, Kanaan C, Ibrahim R, Abdelnabi M, Soin S, Bcharah G, et al. Incidence of arrhythmias in chronic obstructive pulmonary disease, obstructive sleep apnea, and overlap syndrome: A retrospective cohort study. Heart Rhythm. 2025;S1547-5271(25):00213-9. [DOI] [PubMed]
- 6.Robertson NM, Centner CS, Tejwani V, Hossen S, Karmali D, Liu S, et al. Preserved Ratio Impaired Spirometry Prevalence, Risk Factors, and Outcomes: A Systematic Review and Meta-Analysis. Chest. 2025;167(6):1591-614. [DOI] [PubMed]
- 7.Romiti GF, Corica B, Pipitone E, Vitolo M, Raparelli V, Basili S, et al. Prevalence, management and impact of chronic obstructive pulmonary disease in atrial fibrillation: a systematic review and meta-analysis of 4,200,000 patients. Eur Heart J. 2021;42(35):3541–54. [DOI] [PubMed] [Google Scholar]
- 8.The ARIC investigators. The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. Am J Epidemiol. 1989;129(4):687–702. [PubMed] [Google Scholar]
- 9.Fried LP, Borhani NO, Enright P, Furberg CD, Gardin JM, Kronmal RA, et al. The cardiovascular health study: design and rationale. Ann Epidemiol. 1991;1(3):263–76. [DOI] [PubMed] [Google Scholar]
- 10.Giffen CA, Carroll LE, Adams JT, Brennan SP, Coady SA, Wagner EL. Providing contemporary access to historical biospecimen collections: development of the NHLBI biologic specimen and data repository information coordinating center (BioLINCC). Biopreserv Biobank. 2015;13(4):271–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Giffen CA, Wagner EL, Adams JT, Hitchcock DM, Welniak LA, Brennan SP, et al. Providing researchers with online access to NHLBI biospecimen collections: The results of the first six years of the NHLBI BioLINCC program. PLoS One. 2017;12(6): e178141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cheng YJ, Chen ZG, Yao FJ, Liu LJ, Zhang M, Wu SH. Airflow obstruction, impaired lung function and risk of sudden cardiac death: a prospective cohort study. Thorax. 2022;77(7):652–62. [DOI] [PubMed] [Google Scholar]
- 13.Graham BL, Steenbruggen I, Miller MR, Barjaktarevic IZ, Cooper BG, Hall GL, et al. Standardization of spirometry 2019 update. An official American Thoracic Society and European Respiratory Society technical statement. Am J Respir Crit Care Med. 2019;200(8):e70–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Vaz FC, McAvay G, Van Ness PH, Casaburi R, Jensen RL, MacIntyre N, et al. Phenotype of Spirometric Impairment in an Aging Population. AM J Resp Crit Care. 2016;193(7):727–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Flanagin A, Frey T, Christiansen SL. Updated guidance on the reporting of race and ethnicity in medical and science journals. JAMA. 2021;326(7):621–7. [DOI] [PubMed] [Google Scholar]
- 16.Van Gelder IC, Rienstra M, Bunting KV, Casado-Arroyo R, Caso V, Crijns H, et al. 2024 ESC Guidelines for the management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS). EUR HEART J. 2024;45(36):3314–414. [DOI] [PubMed] [Google Scholar]
- 17.Al-Khatib SM, Stevenson WG, Ackerman MJ, Bryant WJ, Callans DJ, Curtis AB, et al. 2017 AHA/ACC/HRS Guideline for Management of Patients With Ventricular Arrhythmias and the Prevention of Sudden Cardiac Death: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Rhythm Society. J Am Coll Cardiol. 2018;72(14):e91–220. [DOI] [PubMed] [Google Scholar]
- 18.Maheshwari A, Norby FL, Roetker NS, Soliman EZ, Koene RJ, Rooney MR, et al. Refining prediction of atrial fibrillation-related stroke using the P(2)-CHA(2)DS(2)-VASc score. Circulation. 2019;139(2):180–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zhao D, Bartz TM, Sotoodehnia N, Post WS, Heckbert SR, Alonso A, et al. Mitochondrial DNA copy number and incident atrial fibrillation. BMC Med. 2020;18(1): 246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Norby FL, Alonso A, Rooney MR, Maheshwari A, Koene RJ, Zhang M, et al. Association of ventricular arrhythmias with dementia: the Atherosclerosis Risk in Communities (ARIC) study. Neurology. 2021;96(6):e926–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zhao D, Post WS, Blasco-Colmenares E, Cheng A, Zhang Y, Deo R, et al. Racial differences in sudden cardiac death. Circulation. 2019;139(14):1688–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Jensen PN, Gronroos NN, Chen LY, Folsom AR, DeFilippi C, Heckbert SR, et al. Incidence of and risk factors for sick sinus syndrome in the general population. J Am Coll Cardiol. 2014;64(6):531–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Schafer JL, Olsen MK. Multiple imputation for multivariate missing-data problems: a data analyst’s perspective. Multivar Behav Res. 1998;33(4):545–71. [DOI] [PubMed] [Google Scholar]
- 24.Rockhill B, Newman B, Weinberg C. Use and misuse of population attributable fractions. Am J Public Health. 1998;88(1):15–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Austin PC, Fine JP. Practical recommendations for reporting Fine-Gray model analyses for competing risk data. Stat Med. 2017;36(27):4391–400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Cotter J, Schmiege S, Moss A, Ambroggio L. How to interact with interactions: what clinicians should know about statistical interactions. Hosp Pediatr. 2023;13(10):e319–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Li J, Agarwal SK, Alonso A, Blecker S, Chamberlain AM, London SJ, et al. Airflow obstruction, lung function, and incidence of atrial fibrillation: the Atherosclerosis Risk in Communities (ARIC) study. Circulation. 2014;129(9):971–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Romiti GF, Corica B, Mei DA, Frost F, Bisson A, Boriani G, et al. Impact of chronic obstructive pulmonary disease in patients with atrial fibrillation: an analysis from the GLORIA-AF registry. Europace. 2023. 10.1093/europace/euae021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Guo Y, Lip GY, Apostolakis S. Inflammation in atrial fibrillation. J Am Coll Cardiol. 2012;60(22):2263–70. [DOI] [PubMed] [Google Scholar]
- 30.Manolis AA, Manolis TA, Apostolopoulos EJ, Apostolaki NE, Melita H, Manolis AS. The role of the autonomic nervous system in cardiac arrhythmias: the neuro-cardiac axis, more foe than friend? Trends Cardiovasc Med. 2021;31(5):290–302. [DOI] [PubMed] [Google Scholar]
- 31.Parikh RR, Norby FL, Wang W, Thenappan T, Prins KW, Van’T HJ, et al. Association of Right Ventricular Afterload With Atrial Fibrillation Risk in Older Adults: The Atherosclerosis Risk in Communities Study. Chest. 2022;162(4):884–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Theofilogiannakos EK, Anogeianaki A, Tsekoura P, Glouftsios P, Ilonidis G, Hatzitolios A, et al. Arrhythmogenesis in patients with stable chronic obstructive pulmonary disease. J Cardiovasc Med. 2008;9(1):89–93. [DOI] [PubMed] [Google Scholar]
- 33.Sourour N, Riveland E, Naesgaard P, Kjekshus H, Larsen AI, Rosjo H, et al. Associations Between Biomarkers of Myocardial Injury and Systemic Inflammation and Risk of Incident Ventricular Arrhythmia. Jacc-Clin Electrophy. 2024;10(9):2021–32. [DOI] [PubMed] [Google Scholar]
- 34.Chen Y, Lou L, Zhang X, Jin L, Chen Y, Chen L, et al. Association between circulating leukocytes and arrhythmias: Mendelian randomization analysis in immuno-cardiac electrophysiology. Front Immunol. 2023;14:1041591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Kusumoto FM, Schoenfeld MH, Barrett C, Edgerton JR, Ellenbogen KA, Gold MR, et al. 2018 ACC/AHA/HRS guideline on the evaluation and management of patients with bradycardia and cardiac conduction delay: a report of the American College of Cardiology/American Heart Association task force on clinical practice guidelines and the Heart Rhythm Society. Circulation. 2019;140(8):e382-482. [DOI] [PubMed] [Google Scholar]
- 36.Medi C, Kalman JM, Ling LH, Teh AW, Lee G, Lee G, et al. Atrial electrical and structural remodeling associated with longstanding pulmonary hypertension and right ventricular hypertrophy in humans. J Cardiovasc Electrophysiol. 2012;23(6):614–20. [DOI] [PubMed] [Google Scholar]
- 37.Vahdatpour CA, Luebbert JJ, Palevsky HI. Atrial arrhythmias in chronic lung disease-associated pulmonary hypertension. Pulm Circ. 2020;10(1): 765637027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Marvisi M, Brianti M, Marani G, Turrini G, Zambrelli P, Ajolfi C, et al. Acute antiarrhythmic effects of bi-level positive airway pressure ventilation in patients with acute respiratory failure caused by chronic obstructive pulmonary disease: a randomized clinical trial. Respiration. 2004;71(2):152–8. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Additional file 1: Tables S1-S15. Table S1. Proportion of missing values in the main analysis. Table S2. Baseline characteristics of the study population in ARIC cohort, and CHS cohort. Table S3. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function, excluding prevalent coronary heart disease or heart failure. Table S4. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function with complete data. Table S5. Hazard ratios and 95% confidence intervals for incident arrhythmias according to FEV1, FVC and FEV1/FVC. Table S6. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function, using time-dependent covariates. Table S7. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function with death as competing risk. Table S8. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function, ARIC cohort and CHS cohort. Table S9. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function, stratified by race and ethnicity. Table S10. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function, stratified by sex. Table S11. Hazard ratios and 95% confidence intervals for incident arrhythmias according to objective indices of lung function, stratified by smoking status. Table S12. Hazard ratios and 95% confidence intervals for incident arrhythmias according to lung disease categories, by race and ethnicity. Table S13. Hazard ratios and 95% confidence intervals for incident arrhythmias according to lung disease categories, ARIC cohort and CHS cohort. Table S13. Hazard ratios and 95% confidence intervals for incident arrhythmias according to lung disease categories, ARIC cohort and CHS cohort. Table S15. Hazard ratios and 95% confidence intervals for incident arrhythmias according to lung disease categories, stratified by smoking status.
Additional file 2. STROBE Statement.Checklist of items that were included in reports of the cohort studies.
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the BioLINCC website on reasonable application.
No datasets were generated or analysed during the current study.