Risk of atrial fibrillation in patients with pneumoconiosis: A nationwide study in Taiwan

Abstract Background To investigate the incidence of new‐onset atrial fibrillation (AF) among subjects with pneumoconiosis using the Taiwan National Health Insurance Research Database. Hypothesis Pneumoconiosis patients are at an increased risk of AF. Methods A total of 12 209 pneumoconiosis patients were in the exposure cohort. Patients without pneumoconiosis were included as the comparison cohort. Both cohorts were matched by gender, age, comorbidity, and index year in a 1:1 manner. Multivariable cox proportional hazard model was used to calculate the adjusted hazard ratios (HRs) after adjustment for age, sex, and all comorbidities. Results The risk of AF in pneumoconiosis patients was 1.30‐fold higher than that of controls (95% CI = 1.17‐1.44) was the key finding. Conclusions Pneumoconiosis is associated with increased risk of incident AF.


| INTRODUCTION
The importance of atrial fibrillation (AF) is increasing recognized since the impact of AF on thromboembolism, adverse cardiovascular outcomes and mortality has been evaluated and well documented. [1][2][3] Indeed, CHA2DS2VASC score, which contained several critical comorbidities, is reported be a good predictor of poor outcomes in AF patients. [4][5][6][7] Pneumoconiosis is a chronic condition with multiple triggering factors and includes multiple clinical presentations. [8][9][10][11][12][13] Moreover, the course can be influenced by permanent exposure to toxins, including tobacco. [8][9][10][11][12][13] Indeed, pneumoconiosis, one kind of environmental lung disease, has been postulated across a number of clinical studies to commonly coexist with cardiovascular diseases. [8][9][10][11][12][13] However, there is no data presented regarding pneumoconiosis and AF. Using a large administrative database in Taiwan, we argued whether this association exists and to explore whether pneumoconiosis is a risk factor for (or "an independent predictor of") AF.

| Data source
All data were acquired from the National Health Insurance Research Database (NHIRD), which was established by National Health Research Institutes and insured over 99% residents in Taiwan. 14 The database contained abundant health and medical treatment information of the insurant. In the present study, we used inpatient file to explore the association between pneumoconiosis (ICD-9-CM code 500, 501, 502, 503, and 505) and AF (ICD-9-CM code 427.31). We further confirmed the diagnosis of pneumoconiosis using Registry of Catastrophic Illness Patient Database (RCIPD), which was a subset of the NHIRD. RCIPD

| Study population
The exposure cohort was enrolled from RCIPD with a diagnosis of pneumoconiosis during 2000 to 2011, and the diagnosis date was defined as the index date. We excluded patients who were younger than 20 years old or who had preexisting AF prior to the index date.
A total of 12 209 pneumoconiosis patients were in the exposure cohort.
Patients without pneumoconiosis were included as the comparison cohort. The comparison cohort was collected using the same criteria as exposure cohort and individually matched by gender, age, comorbidity, and index year in a 1:1 manner. The endpoint of followup was the date of AF diagnosis, death, withdrawal from NHI, or December 31, 2013, whichever came first. Comorbidities included hypertension, diabetes mellitus, hyperlipidemia, coronary artery disease (CAD), peripheral arterial occlusive disease (PAOD), chronic obstructive pulmonary disease (COPD), heart failure, asthma, interstitial lung disease, bronchiectasis, tuberculosis, liver cirrhosis, cancer, chronic kidney disease, hyperthyroidism, gout, sleep disorders, and stroke.

| Statistical analysis
The statistical differences between two cohorts were determined through the chi-square test for categorical variables and t test for

| DISCUSSION
This retrospective study tried to investigate whether pneumoconiosis and AF are associated, and turned up with a positive answer. Patients with pneumoconiosis had a 30% higher risk of incident AF at a mean follow-up of 8 years; this remained significant after adjustment for potential confounders by multivariable cox proportional hazard model analysis.
The greatest strength of this study is its sample size. Moreover, the study methodology is reasonable, the follow-up data was relatively robust, making this finding reliable. The results may provide an original contribution on the risk factors for AF development.
Based upon our results, this novel finding serves to generate an interesting hypothesis for further investigation. The most important question that this study raises is the relevance of this finding to the mechanism of AF. Some might argue that perhaps pneumoconiosis results in a greater vascular insult that may potentiate AF, [9][10][11] or this group identifies a sicker cohort of patients that are more predisposed to development of AF. However, based upon our findings the association between pneumoconiosis and AF is even stronger in the subgroup of no comorbidity after adjustment for the covariates. Furthermore, both cohorts were relatively similar in terms of the prevalence of underlying medical comorbidities. In teasing out these distinctions, it might also reflect on the potential role that pneumoconiosis may play in atrial remodeling. Lung diseases commonly coexist, mostly due to shared common risk factors. [15][16][17][18][19][20][21] In clinical practice, however, we may overlook coexistence of lung diseases and the risk of incident AF. [15][16][17][18][19][20][21][22][23][24][25] There is great evidence in place to link bronchiectasis, COPD, asthma, interstitial lung disease, and AF and the risk of events. [22][23][24][25] In this study, the risk ratio is more profound in the subgroup of male gender and younger age after adjusting for an exhaustive list of covariates; implying that these population deserves more attention and future studies are encouraged to explore the fundamental mechanistic hypothesis in light of the link between pneumoconiosis and AF.

| LIMITATIONS
First, the apparent relationship could be related to presence of several unmeasured confounders or of ascertainment bias. Second, lack of information on body mass index, physical activity, and alcohol use, which are major risk factors for AF and could have played a role as well. Finally, the diagnosis of AF was not made by ECG or Holter.
There is also no information on types of AF (paroxysmal, persistent, or permanent). Neither was the data on treatment performed (medication, pulmonary vein ablation, AV node ablation + pacemaker, flutter ablation) provided. Despite these study limitations which can probably attenuate generalization of the results presented here; our study, a novel and interesting real-world observation, reports on an increased AF rate in patients with pneumoconiosis.

| CONCLUSION
Based on our results, pneumoconiosis and incident AF are associated.

ACKNOWLEDGMENT
This study was supported in part by Taiwan