0
Original Research: Pulmonary Vascular Disease |

The Impact of Smoking on Thromboembolism and Mortality in Patients With Incident Atrial FibrillationSmoking and Thromboembolism in Atrial Fibrillation: Insights From the Danish Diet, Cancer, and Health Study FREE TO VIEW

Ida Ehlers Albertsen, BSc; Lars Hvilsted Rasmussen, PhD; Deirdre A. Lane, PhD; Thure Filskov Overvad, BSc; Flemming Skjøth, PhD; Kim Overvad, PhD; Gregory Y. H. Lip, MD; Torben Bjerregaard Larsen, MD, PhD
Author and Funding Information

From the Department of Cardiology (Ms Albertsen, Mr Overvad, and Drs Skjøth and Larsen), Aalborg Atrial Fibrillation Study Group, Aalborg University Hospital, Aalborg, Denmark; Thrombosis Research Centre (Drs Rasmussen, Lip, and Larsen), Aalborg University, Aalborg, Denmark; the Section for Epidemiology (Dr Overvad), Department of Public Health, Aarhus University, Aarhus, Denmark; and the University of Birmingham Centre for Cardiovascular Sciences (Ms Albertsen, Drs Lane and Lip, and Mr Overvad), City Hospital, Birmingham, England.

Correspondence to: Torben Bjerregaard Larsen, MD, PhD, Department of Cardiology, Aalborg AF Study Group, Aalborg Hospital, Aarhus University Hospital, Forskningens Hus, Søndre Skovvej 15, DK-9100 Aalborg, Denmark; e-mail: tobl@rn.dk


Drs Lip and Larsen served as joint senior authors on this article.

This study was reported in a poster presentation at the American Heart Association’s Scientific Sessions, November 6, 2012, Los Angeles, CA, and at the European Congress of Epidemiology, August 12, 2013, Aarhus, Denmark.

Funding/Support: The Danish Cancer Society financially supported the Diet, Cancer and Health study [Grant 91-8501].

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details.


Chest. 2014;145(3):559-566. doi:10.1378/chest.13-1740
Text Size: A A A
Published online

Background:  Smoking and atrial fibrillation (AF) are major health problems worldwide and are responsible for substantial health-care costs. Our aim was to investigate whether smoking impacts the risk of stroke and death in patients with AF. To test this hypothesis, we analyzed data from a large Danish cohort: the Diet, Cancer, and Health study.

Methods:  This was a cohort study of 57,053 people (27,178 men; 29,876 women) aged 50 to 64 years. The risk of thromboembolism (ischemic stroke/arterial thromboembolism) or death according to smoking habits among 3,161 patients with incident AF (mean age, 66.9 years; 2,032 men, 1,129 women) was assessed using Cox proportional hazard models after a median follow-up of 4.9 years.

Results:  Of those with AF, 34% were current smokers and 37% former smokers. After adjustment for vitamin K antagonist treatment, the hazard ratios (HRs) (95% CI) of thromboembolism or death were 3.13 (1.72-6.37) and 2.73 (2.02-3.70) among women and men who currently were heavy smokers (> 25 g/d), respectively. The associations remained after adjustment for well-established risk factors with HRs of 3.64 (1.88-7.07) and 2.17 (1.59-2.95) among women and men, respectively. In a sensitivity analysis, smoking was still strongly associated with thromboembolism or death after censoring people with a cancer diagnosis during follow-up.

Conclusions:  Smoking is associated with a higher risk of thromboembolism or death in patients with AF even after adjusting for well-recognized risk factors used in stroke risk stratification schemes. The associations may be modified by sex, as the associations were strongest among women.

Figures in this Article

Smoking is a dominant health hazard worldwide and responsible for major health-care costs.1 Indeed, smoking is associated with a higher incidence of vascular disease mortality2 and contributes independently to the risk of stroke.35 Furthermore, smoking has been associated with higher risk of atrial fibrillation (AF).69

AF is the most common cardiac arrhythmia, with a lifetime risk for development of one in four.10 AF is an independent risk factor for stroke, thromboembolism, and death.8,11 The risk of thromboembolism and stroke in patients with AF is not homogenous and varies according to the presence of common stroke risk factors.1214

The few prospective studies examining smoking as a risk factor for stroke in patients with AF have been restricted to dichotomous analyses of smoking habits and have revealed inconsistent results.1519 Hence, smoking has not been comprehensively assessed as a risk factor for adverse outcomes among patients with AF.

We hypothesized that smoking would be associated with a higher risk of thromboembolism and death among patients with incident AF. To test this hypothesis, we analyzed data from a large Danish prospective cohort, the Diet, Cancer, and Health (DCH) study, to assess the risk of thromboembolism and/or death among incident AF cases according to smoking status.

The DCH cohort was established between 1993 and 1997. The study design has been reported in detail elsewhere.20 The primary objective of this prospective study was to investigate the etiologic role of diet and lifestyle in the development of cancer; 27,178 men and 29,876 women were enrolled. The participants were aged 50 to 64 years and were without a cancer diagnosis registered in the Danish Cancer Registry21 at baseline. The DCH cohort has detailed information on demographics, existing comorbidities, and individual risk factors, including detailed information on smoking habits.

Case Finding

The DCH study subjects were linked to the National Patient Register21 dating back to 1977, using the Danish Personal Identification number. This is a unique and national identification number, which is part of the personal information stored in the Civil Registration System. The study population in this study included participants who developed incident AF during follow-up: April 1995 until December 30, 2009.

Codes from the International Classification of Diseases (ICD)-10 were used to extract admissions for AF. AF and atrial flutter have one ICD-10 code (I48). Therefore, some atrial flutter cases have been included in the present study.

Exposure

The exposure variable studied was smoking. Information on tobacco consumption was obtained from a standardized questionnaire at baseline and at 5 years after inclusion. In this study, the information most recent and prior to the AF diagnosis was obtained. Current tobacco consumption in grams per day was calculated by equating a cigarette with 1 g, a cheroot or pipe with 3 g, and a cigar with 4.5 g of tobacco. Participants were arbitrarily categorized into four groups: nonsmokers, former smokers, moderate current smokers (1-25 g/d), and heavy current smokers (> 25 g/d). The exposure level for current smokers was also investigated as a continuous variable.

Information on relevant baseline characteristics and comorbid variables was found in the National Patient Register using ICD and Anatomical Therapeutic Chemical classification system codes (e-Table 1). In Denmark, ICD-8 was used until 1995, after which ICD-10 was used.

Medical treatment status was obtained by evaluating the prescription pattern obtained from the Danish National Prescription Registry21 of medics within the specified Anatomical Therapeutic Chemical classification system group. The treatment period was initiated at time of the first prescription and was considered to proceed if the next prescription was within 1 year or not exceeding the previous period length by a factor 1.5. The treatment period was assumed to continue after the last prescription date according to the number of daily doses in the last package, as estimated from the previous period.

Outcomes

We defined our primary outcome as the composite of thromboembolism or death during follow-up, with thromboembolism comprising ischemic stroke and arterial thromboembolism. Separate analyses were performed for the components of the primary composite end point, namely, thromboembolism and death. Information on emigration or death was available from the National Civil Registration System. Incident cases of ischemic stroke and arterial thromboembolism were found in the National Patient Register.

Statistical Methods

Associations between tobacco exposure and risk of event were reported in terms of incidence and subsequently analyzed by Cox proportional regression analyses using time since AF as the underlying time axis. End of study or emigration was considered as outcome independent censoring. For the secondary outcome event thrombosis, death was considered as a censoring event but also as a competing risk.

All analyses were stratified by sex. Since oral anticoagulation therapy significantly reduces stroke/thromboembolism and death in patients with AF,22 vitamin K antagonist (VKA) treatment was incorporated as a time-varying indicator for being in treatment in all analyses using the counting-process approach.23 The multivariate analyses were adjusted for well-known risk factors included in established guidelines: congestive heart failure, hypertension, diabetes, prior stroke/thromboembolism, vascular disease, and age.12,13 The risk factors were derived at the time of AF diagnosis and assumed constant throughout time at risk. Age was included as a cubic spline with four knots.24 The dose-response for tobacco exposure (dose in g/d) was analyzed by allowing the exposure of the current smokers to be modeled using a natural cubic spline with four knots. The results are presented graphically and reported as hazard ratios (HRs) with 95% CIs.

The associations among the low-risk participants, defined as men with a CHA2DS2-VASc (congestive heart failure, hypertension, age ≥ 75 years, diabetes mellitus, previous stroke/transient ischemic attack, vascular disease, age 65 to 74 years, and sex category) score of 0 and women with a score of 1, were analyzed by a Cox regression model adjusted for VKA treatment and age. A sensitivity-analysis was undertaken to see if the association between smoking and event was primarily driven by death because of cancer; therefore, all death recordings for patients with a cancer diagnosis during follow-up were treated as a censoring event. A similar approach was undertaken to investigate the impact of COPD on the risk of death.

The potential of including smoking information in a risk score (one point for current smoker) for predicting stroke was evaluated post hoc by reporting c-statistics,25 integrated discrimination improvement (IDI), and net reclassification improvement (NRI)26 at 5 years’ follow-up. Data were analyzed using Stata, version 12 (Stata Corp). A P value < .05 was considered statistically significant.

Ethics

The DCH study was conducted in accordance with the Helsinki Declaration II and approved by the regional ethics committees and the Danish Data Protection Agency (Journal No. 2011-41-6056). Written informed consent was obtained from all participants.

In the DCH cohort, 3,366 patients were diagnosed with incident AF during follow-up. Of those, 205 (6.1%) were excluded from the final analyses, leaving a study population of 3,161 cases (Fig 1).

Figure Jump LinkFigure 1. Flowchart showing people included in the final study population. AF* = atrial fibrillation; CPR = Central Person Register.Grahic Jump Location

Clinical characteristics of the study population are shown in Table 1. Of these, 29.6% were never smokers, 36.5% were former smokers, 27.5% smoked ≤ 25 g/d, and 6.4% smoked > 25 g/d. No sex differences in patterns of clinical characteristics were evident (e-Table 2).

Table Graphic Jump Location
Table 1 —Clinical Characteristics of 3,161 Incident Atrial Fibrillation Cases According to Smoking Status

Data are given as median (10%-90%) or % unless otherwise indicated. VKA = vitamin K antagonist.

a 

Congestive heart failure covers both heart failure and left ventricular dysfunction.

b 

VKA treatment at time of atrial fibrillation diagnosis.

During the 17,391 person-years of follow-up (median, 4.9 years), 619 patients (19.6%) died, and 215 (6.8%) had an ischemic stroke or arterial thromboembolism. Crude outcome incidence is shown in Table 2.

Table Graphic Jump Location
Table 2 —Crude Incidence per 100 Person-Years (95% CI) for Thromboembolisma and Death After Incident Atrial Fibrillationb
a 

Thromboembolism comprising ischemic stroke and arterial thromboembolism.

b 

Whole cohort and men and women separately.

HRs, with never smokers as the reference, are shown in Table 3. In multivariate analyses, the risk of thromboembolism or death was significantly higher with greater smoking burden, largely driven by the association with death. In the > 25 g/d smokers, the HRs in the multivariate analyses for women and men were 3.64 (95% CI, 1.88-7.07) and 2.17 (95% CI, 1.59-2.95), respectively. Sensitivity analyses show that the associations involving death as an outcome were not explained by excess risk of cancer or fatal COPD (e-Tables 3, 4). The association for the composite end point among heavy smokers for women disappeared, but this subgroup included only 24 patients, and these results should be interpreted with caution, as also reflected by the wide CIs. A post hoc analysis revealed improved discriminative statistics (c-statistics, IDI, and NRI) when smoking was added to the CHA2DS2-VASc score (Table 4).

Table Graphic Jump Location
Table 3 —Hazard Ratios (95% CI) for Thromboembolisma or Death Among 3,161 Women and Men After Incident Atrial Fibrillation

See Table 1 legend for expansion of abbreviation.

a 

Thromboembolism comprising ischemic stroke and arterial thromboembolism.

b 

VKA-adjusted: with time from atrial fibrillation diagnosis as time axis and adjusted for VKA treatment.

c 

Multivariate analysis adjusted for VKA treatment, congestive heart failure, hypertension, diabetes, prior stroke/thromboembolism/transient ischemic attack, vascular disease and age.

Table Graphic Jump Location
Table 4 —Predictive Value (95% CI) of Smoking Over the CHA2DS2-VASc Score Using C-Statistics, NRI, and IDI

CHA2DS2-VASc = congestive heart failure, hypertension, age 75 years, diabetes mellitus, previous stroke/transient ischemic attack, vascular disease, age 65 to 74 years, and sex category; IDI = integrated discrimination improvement; NRI = net reclassification improvement.

a 

Comprising thromboembolism or death.

After multivariate analyses, the HR of death among female patients smoking > 25 g/d was fivefold higher, whereas in men, the HR was 2.6, indicating that the associations were possibly modified by sex, although observed at a relative scale. The association between smoking habits and outcome is shown as a cubic spline in Figure 2. The splines show a dose-response relationship between amount of smoking and risk of “thromboembolism or death” and death. The curves indicate a plateau when looking at thromboembolism.

Figure Jump LinkFigure 2. Continuous analyses of the amount of tobacco and risk of thromboembolism and death. Data are from multivariate analyses on 2,032 men and 1,129 women, adjusted for vitamin K antagonist-treatment, congestive heart failure, hypertension, diabetes, prior stroke/thromboembolism/transient ischemic attack, vascular disease, and age. Dots represent previous smokers with a 95% CI and the horizontal dashed lines indicates hazard ratio = 1.Grahic Jump Location
Low-Risk Subgroup

The incidence and HRs of low-risk subjects are shown in e-Table 5. There was a positive association between smoking and thromboembolism/death after controlling for known risk factors (e-Fig 1).

In this large prospective cohort, we confirmed our hypothesis by demonstrating that smoking is associated with a significantly higher risk of thromboembolism or death, with a dose-response relationship among patients with AF. The association was attenuated but remained even after controlling for well-known risk factors for stroke among patients with AF. The observed association was driven primarily by a difference in mortality.

A history of ischemic stroke or transient ischemic attack, age, hypertension, diabetes, female sex, and vascular disease is taken into account when assessing the risk of stroke in patients with AF.12,13 In this study we demonstrated that not all the risk was accounted for by these well-recognized risk factors used in stroke-risk stratification schemes, and we found indications of better performance of the CHA2DS2-VASc score when smoking was included.

Smoking in people without AF is known to be associated with higher incidence of vascular disease mortality2 and stroke.35 However, the potential associations between smoking and outcomes in patients with AF is sparsely investigated. Furthermore, previous trials15,16,27 that have investigated this subject have all handled smoking as a dichotomous exposure variable with no information on burden/amount, in contrast to this study, in which the amount of smoking was investigated.

An analysis of the Stroke Prevention Using Oral Thrombin Inhibitor in Atrial Fibrillation (SPORTIF) III and V clinical trial cohorts of 7,329 patients with AF receiving anticoagulant treatment showed that smoking was associated with a twofold higher risk of stroke (HR, 2.10; 95% CI, 1.38-3.18) after multivariate analyses adjusting for stroke/transient ischemic attack, coronary artery disease, and alcohol.15 Conversely, a meta-analysis with information on smoking status from three of the historical, randomized controlled trials comparing anticoagulation vs no anticoagulation17,18,19 found a lower risk of stroke in current smokers than in noncurrent smokers (relative risk, 0.4; 95% CI, 0.2-0.96),16 but this was no longer statistically significant after multivariate adjustment; indeed, no numbers were given of the relative risk of current smoking after multivariate adjustment, except for a P value of .08, nor was there information on factors included in the multivariate analyses. The study concluded that the protective effect of smoking may have occurred by chance and that smokers were younger than noncurrent smokers. Indeed, only around 13% were smokers in the clinical trials and very few events occurred in this subgroup, indicating that the study may have been underpowered to investigate smoking as a risk factor. We found clear indications of a dose-response relationship, with stronger associations among current smokers than previous smokers, and the highest risk observed among heavy current smokers (Table 3).

We found that the effect of smoking was possibly modified by sex, as women had relatively higher HRs compared with men. Several studies investigating the risk of stroke in patients with AF have found women to be at higher risk.2830 Two large register-based studies investigating sex differences in stroke risk28,29 found a significantly higher multivariate HR of stroke for women of 18% and 14%, respectively. However, none of the studies took lifestyle factors, such as smoking habits, into account. Our study indicates that smoking could partly explain the association of AF and higher risk of stroke in women, since the association for smoking was possibly stronger in women than in men.

The DCH study population is a selected population in which only about 35% of those invited accepted the invitation to participate, and identification of patients with AF relied only on hospital diagnoses, although these have a high positive predictive value of 92.6%.31 It is, however, unlikely that the association found is systematically different between identified and nonidentified participants with AF. The participants were followed in national registries with almost complete follow-up and, therefore, the study is most likely not subject to selection bias.

Changes in smoking habits may have occurred during follow-up. We sought to minimize the effect of this by using the most recent information on tobacco consumption before AF diagnosis. However, we cannot exclude misclassification stemming from inadequate self-reporting of smoking habits. Underreporting of smoking habits could have led to overestimation of the HRs among the former and moderate current smokers. The predictive value for stroke diagnosis is > 80% and most likely not associated with smoking habits obtained at baseline or at follow-up. Information bias is, thus, not a likely explanation of the study results.

Ideally, the analyses would have been conducted on a cohort not receiving anticoagulant treatment, given the fact that smokers are more likely to be treated with VKA than nonsmokers. However, this was not the case in this study. In the first analyses, only VKA treatment was taken into account. The time axis was the time from the diagnosis of AF and, thus, closely controlled for. We did not differentiate between subtypes of AF and, therefore, are not able to determine if the subtype of AF differed among smoking subgroups, just as some atrial flutter cases may have been included due to coding similarities. Nor did we differentiate between subtypes of ischemic strokes and, therefore, are not able to conclude if smoking has a stronger association with a certain subtype of ischemic stroke. As the primary study outcome was thromboembolism or death, we focused on the possible added value of smoking after the use of risk factors recommended by guidelines,12,13 and we controlled for the components of CHA2DS2-VASc score. Smoking may also be associated with other lifestyle risk factors for thromboembolism or death. These risk factors would have been potential confounders if the aim had been to specify the independent association for smoking, but not in this study, which concerned a possible added risk to the already known risk factors. Confounding by other known stroke-risk factors is, thus, not an issue of concern. Indeed, we have chosen a composite end point including the two outcomes significantly affected by intervention with VKA, since it reduces both stroke and death in patients with AF.22 Importantly, some deaths in our study were likely due to undiagnosed ischemic stroke, due to the register-based nature of the dataset. This might partly explain the stronger association of smoking and death rather than smoking and thromboembolism.

Clinical Implications

What are the potential clinical implications? Smoking status could possibly be used to flag the patient with AF who is at high risk of thromboembolism or death even after controlling for established risk factors. The impact of smoking requires validation in cohorts with more heavy smokers and older patients with AF. Furthermore, the results from the DCH cohort need confirmation in other ethnic groups. As part of the holistic management of AF, attention should also be directed toward smoking cessation.

Smoking is associated with a higher risk of thromboembolism or death in patients with AF even after adjusting for well-recognized risk factors used in stroke-risk stratification schemes. The associations were possibly modified by sex, as we found the strongest associations among women.

Author contributions: Ms Albertsen had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Ms Albertsen: contributed to the study concept and design, data analysis and interpretation, drafting or revising of the manuscript, and approval of the final manuscript and served as the principal author.

Dr Rasmussen: contributed to data analysis and interpretation, drafting or revising of the manuscript, and approval of the final manuscript.

Dr Lane: contributed to data analysis and interpretation, drafting or revising of the manuscript, and approval of the final manuscript.

Mr Overvad: contributed to the study concept and design, data analysis and interpretation, drafting or revising of the manuscript, and approval of the final manuscript.

Dr Skjøth: contributed to statistical analyses, data analysis and interpretation, drafting or revising of the manuscript, and approval of the final manuscript.

Dr Overvad: contributed to data analysis and interpretation, drafting or revising of the manuscript, and approval of the final manuscript.

Dr Lip: contributed to data analysis and interpretation, drafting or revising of the manuscript, and approval of the final manuscript.

Dr Larsen: contributed to the study concept and design, data analysis and interpretation, drafting or revising of the manuscript, approval of the final manuscript.

Financial/nonfinancial disclosures: The authors have reported to CHEST the following conflicts of interest: Drs Rasmussen and Larsen have served as speakers for Bristol-Myers-Squibb Co/Pfizer Inc and Boehringer Ingelheim GmBH. Dr Lane has received funding for research for Bayer Healthcare (Bayer AG) and Boehringer Ingelheim GmBH, and has served as a speaker for Bayer AG, Boehringer Ingelheim GmBH, and Bristol-Myers-Squibb Co/Pfizer Inc. Dr Lip has served as a consultant for Bayer AG, Astellas Pharma Inc, Merck & Co Inc, Sanofi SA, Bristol-Myers-Squibb Co/Pfizer Inc, Daiichi-Sankyo Co Ltd, Biotronik SE & Co, Portola Pharmaceuticals Inc, and Boehringer Ingelheim GmBH, and has served as a speaker for Bayer Pharmaceuticals Corp, Bristol-Myers-Squibb Co/Pfizer Inc, Boehringer Ingelheim GmBH, and Sanofi Aventis LLC. Ms Albertsen, Mr Overvad, and Drs Skjøth and Overvad have reported that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.

Role of sponsors: The sponsor had no role in the design of the study, the collection and analysis of the data, or the preparation of the manuscript.

Additional information: The e-Tables and e-Figure can be found in the “Supplemental Materials” area of the online article.

AF

atrial fibrillation

CHA2DS2-VASc

congestive heart failure, hypertension, age ≥ 75 years, diabetes mellitus, previous stroke/transient ischemic attack, vascular disease, age 65 to 74 years, and sex category

DCH

Diet, Cancer, and Health

HR

hazard ratio

ICD

International Classification of Diseases

IDI

integrated discrimination improvement

NRI

net reclassification improvement

VKA

vitamin K antagonist

The Gallup Organisation, Hungary. Flash Eurobarometer No 253. Survey on Tobacco. Analytical report. 2009. European Commission website. http://ec.europa.eu/public_opinion/flash/fl_253_en.pdf. Accessed July 26, 2013.
 
Kenfield SA, Stampfer MJ, Rosner BA, Colditz GA. Smoking and smoking cessation in relation to mortality in women. JAMA. 2008;299(17):2037-2047. [CrossRef] [PubMed]
 
Kawachi I, Colditz GA, Stampfer MJ, et al. Smoking cessation and decreased risk of stroke in women. JAMA. 1993;269(2):232-236. [CrossRef] [PubMed]
 
Shinton R, Beevers G. Meta-analysis of relation between cigarette smoking and stroke. BMJ. 1989;298(6676):789-794. [CrossRef] [PubMed]
 
Wolf PA, D’Agostino RB, Kannel WB, Bonita R, Belanger AJ. Cigarette smoking as a risk factor for stroke. The Framingham Study. JAMA. 1988;259(7):1025-1029. [CrossRef] [PubMed]
 
Chamberlain AM, Agarwal SK, Folsom AR, et al. Smoking and incidence of atrial fibrillation: results from the Atherosclerosis Risk in Communities (ARIC) study. Heart Rhythm. 2011;8(8):1160-1166. [CrossRef] [PubMed]
 
Heeringa J, Kors JA, Hofman A, van Rooij FJA, Witteman JCM. Cigarette smoking and risk of atrial fibrillation: the Rotterdam Study. Am Heart J. 2008;156(6):1163-1169. [CrossRef] [PubMed]
 
Krahn AD, Manfreda J, Tate RB, Mathewson FAL, Cuddy TE. The natural history of atrial fibrillation: incidence, risk factors, and prognosis in the Manitoba Follow-Up Study. Am J Med. 1995;98(5):476-484. [CrossRef] [PubMed]
 
Benjamin EJ, Levy D, Vaziri SM, D’Agostino RB, Belanger AJ, Wolf PA. Independent risk factors for atrial fibrillation in a population-based cohort. The Framingham Heart Study. JAMA. 1994;271(11):840-844. [CrossRef] [PubMed]
 
Lloyd-Jones DM, Wang TJ, Leip EP, et al. Lifetime risk for development of atrial fibrillation: the Framingham Heart Study. Circulation. 2004;110(9):1042-1046. [CrossRef] [PubMed]
 
Benjamin EJ, Wolf PA, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation. 1998;98(10):946-952. [CrossRef] [PubMed]
 
You JJ, Singer DE, Howard PA, et al. Antithrombotic therapy for atrial fibrillation: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2012;141(2_suppl):e531S-e575S.
 
Camm AJ, Kirchhof P, Lip GY, et al; European Heart Rhythm Association; European Association for Cardio-Thoracic Surgery. Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). Eur Heart J. 2010;31(19):2369-2429. [CrossRef] [PubMed]
 
Lip GYH, Nieuwlaat R, Pisters R, Lane DA, Crijns HJGM. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based approach: the Euro Heart Survey on atrial fibrillation. Chest. 2010;137(2):263-272. [CrossRef] [PubMed]
 
Lip GYH, Frison L, Halperin JL, Lane DA. Identifying patients at high risk for stroke despite anticoagulation: a comparison of contemporary stroke risk stratification schemes in an anticoagulated atrial fibrillation cohort. Stroke. 2010;41(12):2731-2738. [CrossRef] [PubMed]
 
Atrial Fibrillation Investigators. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation. Analysis of pooled data from five randomized controlled trials. Arch Intern Med. 1994;154(13):1449-1457. [CrossRef] [PubMed]
 
Ezekowitz MD, Bridgers SL, James KE, et al; Veterans Affairs Stroke Prevention in Nonrheumatic Atrial Fibrillation Investigators. Warfarin in the prevention of stroke associated with nonrheumatic atrial fibrillation. N Engl J Med. 1992;327(20):1406-1412. [CrossRef] [PubMed]
 
Stroke Prevention in Atrial Fibrillation Investigators. Stroke prevention in atrial fibrillation study. Final results. Circulation. 1991;84(2):527-539. [CrossRef] [PubMed]
 
The Boston Area Anticoagulation Trial for Atrial Fibrillation Investigators. The effect of low-dose warfarin on the risk of stroke in patients with nonrheumatic atrial fibrillation. N Engl J Med. 1990;323(22):1505-1511. [CrossRef] [PubMed]
 
Tjønneland A, Olsen A, Boll K, et al. Study design, exposure variables, and socioeconomic determinants of participation in Diet, Cancer and Health: a population-based prospective cohort study of 57,053 men and women in Denmark. Scand J Public Health. 2007;35(4):432-441. [CrossRef] [PubMed]
 
Thygesen LC, Daasnes C, Thaulow I, Brønnum-Hansen H. Introduction to Danish (nationwide) registers on health and social issues: structure, access, legislation, and archiving. Scand J Public Health. 2011;39(suppl 7):12-16. [CrossRef] [PubMed]
 
Hart RG, Pearce LA, Aguilar MI. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med. 2007;146(12):857-867. [CrossRef] [PubMed]
 
Hosmer DW, Lemeshow S, May S. Applied Survival Analysis: Regression Modeling of Time-to-Event Data. Hoboken, NJ: John Wiley & Sons, Inc; 1999.
 
Harrell FE. Regression Modeling Strategies.1st ed. New York, NY: Springer-Verlag; 2001.
 
Heagerty PJ, Lumley T, Pepe MS. Time-dependent ROC curves for censored survival data and a diagnostic marker. Biometrics. 2000;56(2):337-344. [CrossRef] [PubMed]
 
Chambless LE, Cummiskey CP, Cui G. Several methods to assess improvement in risk prediction models: extension to survival analysis. Stat Med. 2011;30(1):22-38. [CrossRef] [PubMed]
 
Novello P, Ajmar G, Bianchini D, et al. Ischemic stroke and atrial fibrillation. A clinical study. Ital J Neurol Sci. 1993;14(7):571-576. [CrossRef] [PubMed]
 
Friberg L, Benson L, Rosenqvist M, Lip GYH. Assessment of female sex as a risk factor in atrial fibrillation in Sweden: nationwide retrospective cohort study. BMJ. 2012;344:e3522. [CrossRef] [PubMed]
 
Avgil Tsadok M, Jackevicius CA, Rahme E, Humphries KH, Behlouli H, Pilote L. Sex differences in stroke risk among older patients with recently diagnosed atrial fibrillation. JAMA. 2012;307(18):1952-1958. [CrossRef] [PubMed]
 
Friberg J, Scharling H, Gadsbøll N, Truelsen T, Jensen GB; Copenhagen City Heart Study. Comparison of the impact of atrial fibrillation on the risk of stroke and cardiovascular death in women versus men (The Copenhagen City Heart Study). Am J Cardiol. 2004;94(7):889-894. [CrossRef] [PubMed]
 
Rix TA, Riahi S, Overvad K, Lundbye-Christensen S, Schmidt EB, Joensen AM. Validity of the diagnoses atrial fibrillation and atrial flutter in a Danish patient registry. Scand Cardiovasc J. 2012;46(3):149-153. [CrossRef] [PubMed]
 

Figures

Figure Jump LinkFigure 1. Flowchart showing people included in the final study population. AF* = atrial fibrillation; CPR = Central Person Register.Grahic Jump Location
Figure Jump LinkFigure 2. Continuous analyses of the amount of tobacco and risk of thromboembolism and death. Data are from multivariate analyses on 2,032 men and 1,129 women, adjusted for vitamin K antagonist-treatment, congestive heart failure, hypertension, diabetes, prior stroke/thromboembolism/transient ischemic attack, vascular disease, and age. Dots represent previous smokers with a 95% CI and the horizontal dashed lines indicates hazard ratio = 1.Grahic Jump Location

Tables

Table Graphic Jump Location
Table 1 —Clinical Characteristics of 3,161 Incident Atrial Fibrillation Cases According to Smoking Status

Data are given as median (10%-90%) or % unless otherwise indicated. VKA = vitamin K antagonist.

a 

Congestive heart failure covers both heart failure and left ventricular dysfunction.

b 

VKA treatment at time of atrial fibrillation diagnosis.

Table Graphic Jump Location
Table 2 —Crude Incidence per 100 Person-Years (95% CI) for Thromboembolisma and Death After Incident Atrial Fibrillationb
a 

Thromboembolism comprising ischemic stroke and arterial thromboembolism.

b 

Whole cohort and men and women separately.

Table Graphic Jump Location
Table 3 —Hazard Ratios (95% CI) for Thromboembolisma or Death Among 3,161 Women and Men After Incident Atrial Fibrillation

See Table 1 legend for expansion of abbreviation.

a 

Thromboembolism comprising ischemic stroke and arterial thromboembolism.

b 

VKA-adjusted: with time from atrial fibrillation diagnosis as time axis and adjusted for VKA treatment.

c 

Multivariate analysis adjusted for VKA treatment, congestive heart failure, hypertension, diabetes, prior stroke/thromboembolism/transient ischemic attack, vascular disease and age.

Table Graphic Jump Location
Table 4 —Predictive Value (95% CI) of Smoking Over the CHA2DS2-VASc Score Using C-Statistics, NRI, and IDI

CHA2DS2-VASc = congestive heart failure, hypertension, age 75 years, diabetes mellitus, previous stroke/transient ischemic attack, vascular disease, age 65 to 74 years, and sex category; IDI = integrated discrimination improvement; NRI = net reclassification improvement.

a 

Comprising thromboembolism or death.

References

The Gallup Organisation, Hungary. Flash Eurobarometer No 253. Survey on Tobacco. Analytical report. 2009. European Commission website. http://ec.europa.eu/public_opinion/flash/fl_253_en.pdf. Accessed July 26, 2013.
 
Kenfield SA, Stampfer MJ, Rosner BA, Colditz GA. Smoking and smoking cessation in relation to mortality in women. JAMA. 2008;299(17):2037-2047. [CrossRef] [PubMed]
 
Kawachi I, Colditz GA, Stampfer MJ, et al. Smoking cessation and decreased risk of stroke in women. JAMA. 1993;269(2):232-236. [CrossRef] [PubMed]
 
Shinton R, Beevers G. Meta-analysis of relation between cigarette smoking and stroke. BMJ. 1989;298(6676):789-794. [CrossRef] [PubMed]
 
Wolf PA, D’Agostino RB, Kannel WB, Bonita R, Belanger AJ. Cigarette smoking as a risk factor for stroke. The Framingham Study. JAMA. 1988;259(7):1025-1029. [CrossRef] [PubMed]
 
Chamberlain AM, Agarwal SK, Folsom AR, et al. Smoking and incidence of atrial fibrillation: results from the Atherosclerosis Risk in Communities (ARIC) study. Heart Rhythm. 2011;8(8):1160-1166. [CrossRef] [PubMed]
 
Heeringa J, Kors JA, Hofman A, van Rooij FJA, Witteman JCM. Cigarette smoking and risk of atrial fibrillation: the Rotterdam Study. Am Heart J. 2008;156(6):1163-1169. [CrossRef] [PubMed]
 
Krahn AD, Manfreda J, Tate RB, Mathewson FAL, Cuddy TE. The natural history of atrial fibrillation: incidence, risk factors, and prognosis in the Manitoba Follow-Up Study. Am J Med. 1995;98(5):476-484. [CrossRef] [PubMed]
 
Benjamin EJ, Levy D, Vaziri SM, D’Agostino RB, Belanger AJ, Wolf PA. Independent risk factors for atrial fibrillation in a population-based cohort. The Framingham Heart Study. JAMA. 1994;271(11):840-844. [CrossRef] [PubMed]
 
Lloyd-Jones DM, Wang TJ, Leip EP, et al. Lifetime risk for development of atrial fibrillation: the Framingham Heart Study. Circulation. 2004;110(9):1042-1046. [CrossRef] [PubMed]
 
Benjamin EJ, Wolf PA, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation. 1998;98(10):946-952. [CrossRef] [PubMed]
 
You JJ, Singer DE, Howard PA, et al. Antithrombotic therapy for atrial fibrillation: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2012;141(2_suppl):e531S-e575S.
 
Camm AJ, Kirchhof P, Lip GY, et al; European Heart Rhythm Association; European Association for Cardio-Thoracic Surgery. Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). Eur Heart J. 2010;31(19):2369-2429. [CrossRef] [PubMed]
 
Lip GYH, Nieuwlaat R, Pisters R, Lane DA, Crijns HJGM. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based approach: the Euro Heart Survey on atrial fibrillation. Chest. 2010;137(2):263-272. [CrossRef] [PubMed]
 
Lip GYH, Frison L, Halperin JL, Lane DA. Identifying patients at high risk for stroke despite anticoagulation: a comparison of contemporary stroke risk stratification schemes in an anticoagulated atrial fibrillation cohort. Stroke. 2010;41(12):2731-2738. [CrossRef] [PubMed]
 
Atrial Fibrillation Investigators. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation. Analysis of pooled data from five randomized controlled trials. Arch Intern Med. 1994;154(13):1449-1457. [CrossRef] [PubMed]
 
Ezekowitz MD, Bridgers SL, James KE, et al; Veterans Affairs Stroke Prevention in Nonrheumatic Atrial Fibrillation Investigators. Warfarin in the prevention of stroke associated with nonrheumatic atrial fibrillation. N Engl J Med. 1992;327(20):1406-1412. [CrossRef] [PubMed]
 
Stroke Prevention in Atrial Fibrillation Investigators. Stroke prevention in atrial fibrillation study. Final results. Circulation. 1991;84(2):527-539. [CrossRef] [PubMed]
 
The Boston Area Anticoagulation Trial for Atrial Fibrillation Investigators. The effect of low-dose warfarin on the risk of stroke in patients with nonrheumatic atrial fibrillation. N Engl J Med. 1990;323(22):1505-1511. [CrossRef] [PubMed]
 
Tjønneland A, Olsen A, Boll K, et al. Study design, exposure variables, and socioeconomic determinants of participation in Diet, Cancer and Health: a population-based prospective cohort study of 57,053 men and women in Denmark. Scand J Public Health. 2007;35(4):432-441. [CrossRef] [PubMed]
 
Thygesen LC, Daasnes C, Thaulow I, Brønnum-Hansen H. Introduction to Danish (nationwide) registers on health and social issues: structure, access, legislation, and archiving. Scand J Public Health. 2011;39(suppl 7):12-16. [CrossRef] [PubMed]
 
Hart RG, Pearce LA, Aguilar MI. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med. 2007;146(12):857-867. [CrossRef] [PubMed]
 
Hosmer DW, Lemeshow S, May S. Applied Survival Analysis: Regression Modeling of Time-to-Event Data. Hoboken, NJ: John Wiley & Sons, Inc; 1999.
 
Harrell FE. Regression Modeling Strategies.1st ed. New York, NY: Springer-Verlag; 2001.
 
Heagerty PJ, Lumley T, Pepe MS. Time-dependent ROC curves for censored survival data and a diagnostic marker. Biometrics. 2000;56(2):337-344. [CrossRef] [PubMed]
 
Chambless LE, Cummiskey CP, Cui G. Several methods to assess improvement in risk prediction models: extension to survival analysis. Stat Med. 2011;30(1):22-38. [CrossRef] [PubMed]
 
Novello P, Ajmar G, Bianchini D, et al. Ischemic stroke and atrial fibrillation. A clinical study. Ital J Neurol Sci. 1993;14(7):571-576. [CrossRef] [PubMed]
 
Friberg L, Benson L, Rosenqvist M, Lip GYH. Assessment of female sex as a risk factor in atrial fibrillation in Sweden: nationwide retrospective cohort study. BMJ. 2012;344:e3522. [CrossRef] [PubMed]
 
Avgil Tsadok M, Jackevicius CA, Rahme E, Humphries KH, Behlouli H, Pilote L. Sex differences in stroke risk among older patients with recently diagnosed atrial fibrillation. JAMA. 2012;307(18):1952-1958. [CrossRef] [PubMed]
 
Friberg J, Scharling H, Gadsbøll N, Truelsen T, Jensen GB; Copenhagen City Heart Study. Comparison of the impact of atrial fibrillation on the risk of stroke and cardiovascular death in women versus men (The Copenhagen City Heart Study). Am J Cardiol. 2004;94(7):889-894. [CrossRef] [PubMed]
 
Rix TA, Riahi S, Overvad K, Lundbye-Christensen S, Schmidt EB, Joensen AM. Validity of the diagnoses atrial fibrillation and atrial flutter in a Danish patient registry. Scand Cardiovasc J. 2012;46(3):149-153. [CrossRef] [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).
Supporting Data

Online Supplement

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543