International Journal of Cardiology 199 (2015) 25–30
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Risk and prediction of dementia in patients with atrial fibrillation — A nationwide population-based cohort study Jo-Nan Liao a,b,1, Tze-Fan Chao a,b,1, Chia-Jen Liu c,d, Kang-Ling Wang a,b, Su-Jung Chen d,e, Ta-Chuan Tuan a,b, Yenn-Jiang Lin a,b, Shih-Lin Chang a,b, Li-Wei Lo a,b, Yu-Feng Hu a,b, Fa-Po Chung a,b, Hsuan-Ming Tsao b,f,⁎, Tzeng-Ji Chen g, Gregory Y.H. Lip h,2, Shih-Ann Chen a,b,⁎⁎,2 a
Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan Institute of Clinical Medicine, and Cardiovascular Research Center, National Yang-Ming University, Taipei, Taiwan c Division of Hematology and Oncology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan d Institute of Public Health and School of Medicine, National Yang-Ming University, Taipei, Taiwan e Division of Infectious Diseases, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan f National Yang-Ming University Hospital, Taiwan g Department of Family Medicine, Taipei Veterans General Hospital, Taipei, Taiwan h University of Birmingham Centre for Cardiovascular Sciences, City Hospital, Birmingham, United Kingdom b
a r t i c l e
i n f o
Article history: Received 2 February 2015 Received in revised form 20 June 2015 Accepted 27 June 2015 Available online 6 July 2015 Keywords: Atrial fibrillation Dementia CHADS2 score CHA2DS2-VASc score
a b s t r a c t Background: Atrial fibrillation (AF) is associated with an increased risk of cognitive impairment and functional decline, and may contribute to development of dementia. Objectives: Data from a nationwide large-scale population-based cohort study are lacking. Besides, how best to predict the occurrence of incident dementia among AF subjects remains uncertain. Methods: A total of 332,665 AF subjects without dementia were identified as the study group from the “National Health Insurance Research Database” in Taiwan. For each study patient, one age- and sex-matched subject without AF and dementia was selected as the control group. The study end point was occurrence of dementia, and the usefulness of CHADS2 and CHA2DS2-VASc scores in predicting dementia was analyzed. Results: During the follow-up, 29,012 AF patients experienced dementia with an annual incidence of 2.12%, higher than non-AF subjects (1.50%). Patients with AF possessed a higher risk of dementia with a hazard ratio (HR) of 1.420 after adjustments for age, gender, baseline differences and medication use. Among AF patients, the CHADS2 and CHA2DS2-VASc scores were significant predictors of dementia with an adjusted HR of 1.520 and 1.497 per 1 increment of the CHADS2 and CHA2DS2-VASc scores, respectively. The c-index for CHA2DS2-VASc in predicting dementia (0.611, 95% confidence interval [CI] = 0.608–0.614) was significantly higher than the CHADS2 score (0.589, 95% CI = 0.586–0.592) (DeLong test p b 0.001). Conclusions: In this nationwide cohort study, AF was independently associated with a higher risk of dementia. The CHA2DS2-VASc score can be used to estimate the risk of dementia in AF patients. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia in clinical practice [1], and carries a five-fold risk of stroke, a three-fold incidence of heart failure and a higher mortality [2].
⁎ Correspondence to: HM Tsao, National Yang-Ming University Hospital, No 152, Xin-Min road, Yi-Lan, Taiwan. ⁎⁎ Correspondence to: SA Chen, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, No. 201, Sec. 2, Shih-Pai Road, Taipei, Taiwan. E-mail addresses: [email protected] (H.-M. Tsao), [email protected] (S.-A. Chen). 1 Dr. Jo-Nan Liao and Dr. Tze-Fan Chao contributed equally to this study. 2 Joint senior authors.
http://dx.doi.org/10.1016/j.ijcard.2015.06.170 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.
AF prevalence increases with age and affects up to 9% of the population older than 80 years old [3]. Aging of population also leads to increased risk of dementia, a disorder which is characteristic of memory and cognitive impairment. Various studies suggested that AF is associated with an increased risk of cognitive impairment and functional decline and may contribute to development of dementia [4–7]. In a systematic review, Udompanich et al. [8] found that among cross-sectional studies, patients with AF had a 1.7 (95% CI = 1.2–2.5) to 3.3 (95% CI = 1.6–6.5) greater risk of cognitive impairment, and a 2.3-fold (95% CI = 1.4–3.7) increased risk of dementia, compared to patients in sinus rhythm; however, there was marked heterogeneity in the design, size and quality of studies and reporting of the data which precluded formal meta-analysis. Similarly, Santangeli et al. performed a meta-analysis incorporating
26
J.-N. Liao et al. / International Journal of Cardiology 199 (2015) 25–30
8 studies, a total of 77,668 patients, and suggested that AF increased the risk of incident dementia by 42% [7]. However, 3 studies enrolled in the meta-analysis did not show a significant association between AF and dementia in the original investigations. Importantly, Asians were not included in most studies. Besides, the risk factors of dementia and how to predict its occurrence among AF patients were not clear. The CHADS2 and CHA2DS2-VASc scores were commonly used to guide anti-thrombotic therapy for stroke prevention in AF. Several components of these 2 scoring schemes, such as an old age, hypertension, and diabetes, were reported to be important risk factors of dementia [9–11]. Furthermore, the occurrence of ischemic stroke, whose risk could be estimated using the CHADS2 and CHA2DS2-VASc scores, would significantly predispose patients to a higher risk of subsequent dementia. However, the relationships between the CHADS 2 and CHA 2DS 2-VASc scores and dementia in AF patients have not been previously studied. In the present study, we aim to investigate whether AF is associated with a higher risk of dementia in Asians using a nationwide database which would have less selection biases and could more accurately represent the real-world condition. We will also test the hypothesis that the CHADS2 and CHA2DS2-VASc scores could be used to predict the occurrence of dementia in AF.
definition in NHIRD has been validated previously [13,14]. The diagnosis of dementia was based on the ICD-9-CM codes (290.0–290.4, 331.0) registered by the physicians responsible for the treatments of patients [15]. For each study patient, one age- and sexmatched subject without past history of AF and dementia was selected as the control group. The flow chart of the enrollment of study population was shown in Fig. 1. Information about important comorbid conditions of each individual was retrieved from the medical claims based on the ICD-9-CM codes. We defined patients with a certain disease only when it was a discharge diagnosis or repeatedly confirmed more than twice in outpatient department. The diagnostic accuracies of important comorbidities in NHIRD, such as hypertension, diabetes mellitus, heart failure, myocardial infarction, hyperlipidemia and chronic obstructive pulmonary disease, have been validated before [16,17]. The Charlson index was assessed as the overall severity of comorbidities [18]. 2.3. Calculations of CHADS2 and CHA2DS2-VASc scores and definition of income level and clinical end points
2. Methods
The CHADS2 score was calculated for each patient by assigning 1 point each for age ≥ 75 years, hypertension, diabetes mellitus, and heart failure, and 2 points each for a previous stroke or transient ischemic attack (TIA) [19]. The CHA2DS2-VASc score was calculated for each patient by assigning 1 point each for age between 65 and 74 years, history of hypertension, diabetes, recent cardiac failure, vascular disease (myocardial infarction or peripheral artery disease), and female gender, and 2 points each for a history of a stroke, TIA, or age ≥ 75 years [20]. Insurance premiums, calculated according to the beneficiary's total income, were used to estimate monthly income. Monthly income was grouped into low income (monthly income b 20,000 New Taiwan Dollar [NTD]), medium income (20,000 NTD ≤ monthly income b 40,000 NTD), and high income (monthly income ≥ 40,000 NTD) [21]. The study end point was the occurrence of dementia (ICD-9-CM codes: 290.0–290.4, 331.0) during the follow-up period [15].
2.1. Database
2.4. Statistical analysis
This study used the “National Health Insurance Research Database” (NHIRD) released by the Taiwan National Health Research Institutes (NHRI). The National Health Insurance (NHI) system is a mandatory universal health insurance program that offers comprehensive medical care coverage to all Taiwanese residents. NHIRD contains detailed health care data from N23 million enrollees, representing N99% of Taiwan's population. In this cohort dataset, the patients' original identification numbers were encrypted to protect privacy, and the encrypting procedure was consistent so that a linkage of the claims belonging to each patient was feasible within the NHI database, and could be followed continuously. Numerous scientific research papers have been published using data from NHIRD (http://nhird.nhri.org.tw/en/Research.html). The large sample size of this database provided a good opportunity to study the association between AF and dementia.
Data were presented as the mean value and standard deviation for normally distributed continuous variables and proportions for categorical variables. Comparisons of the continuous variables were performed using an unpaired two-tailed t-test for normally distributed variables and Mann–Whitney rank-sum test for skewed variables. Nominal variables were compared by chi-square analysis with a Yates correction or Fisher's exact test. The risk of dementia occurrence was assessed using the Cox regression analysis. The cumulative incidence curve of dementia was plotted via the Kaplan–Meier method, with statistical significance examined by the log-rank test. All statistical significances were set at a p b 0.05 using the SPSS version 17.0 statistical software (SPSS, Chicago, IL, USA).
2.2. Study population From January 1, 1996 to December 31, 2011, a total of 332,665 patients with AF with age ≥ 20 years and no history of dementia for at least 1 year preceding study enrollment were retrieved from the NHIRD since the record of NHIRD can be traced back to 1995. AF was diagnosed using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes (427.31). To ensure the accuracy of diagnosis, we defined patients with AF only when it was listed as a discharge diagnosis or confirmed more than twice in the outpatient department [12]. The diagnostic accuracy of AF using this
3. Results 3.1. Baseline characteristics of patients The mean age of the study population was 70.3 ± 13.0 years, and 55.9% of patients were men. The baseline characteristics of AF and non-AF groups were shown in Table 1. Hypertension was the most prevalent comorbidity, which was present in 67.4% of AF patients.
Fig. 1. Flow diagram of study enrollment. From 1996 to 2011, a total of 332,665 patients with AF and no history of dementia were identified from the NHIRD. For each AF patient, one age- and gender-matched subject without past history of AF and dementia was selected from the NHIRD to constitute the control group. AF = atrial fibrillation; NHIRD = National Health Insurance Research Database.
J.-N. Liao et al. / International Journal of Cardiology 199 (2015) 25–30
27
Table 1 Baseline characteristics of subjects with and without AF. Variables
AF patients (n = 332,665)
Non-AF patients (n = 332,665)
p value
Age, years Age ≥ 65 years, n (%) Male gender, n (%)
70.3 ± 13.0 240,198 (72.2%) 186,000 (55.9%)
70.3 ± 13.0 240,198 (72.2%) 186,000 (55.9%)
0.994 1.000 1.000
Underlying diseases, n (%) Hypertension Diabetes mellitus Heart failure Vascular disease Dyslipidemia CVA ESRD COPD Malignancy Autoimmune disease Charlson index
224,169 (67.4%) 92,547 (27.8%) 133,802 (40.2%) 29,308 (8.8%) 82,321 (24.7%) 107,580 (32.3%) 24,232 (7.3%) 113,110 (34%) 34,659 (10.4%) 17,863 (5.4%) 2.5 ± 2.0
153,275 (46.1%) 65,802 (19.8%) 30,705 (9.2%) 9082 (2.7%) 67,376 (20.3%) 57,059 (17.2%) 10,801 (3.2%) 72,387 (21.8%) 25,238 (7.6%) 14,751 (4.4%) 1.5 ± 1.7
b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001
Medication use, n (%) Aspirin Clopidogrel Warfarin ACEIs/ARBs Statin
105,496 (31.7%) 13,939 (4.2%) 33,328 (10.0%) 83,985 (25.2%) 4299 (1.3%)
58,816 (17.7%) 6193 (1.9%) 1919 (0.6%) 69,982 (21%) 26,992 (8.1%)
b0.001 b0.001 b0.001 b0.001 b0.001
Income level Low Medium High
177,680 (53.4%) 111,147 (33.4%) 43,838 (13.2%)
228,641 (68.7%) 63,942 (19.2%) 40,082 (12.0%)
b0.001 b0.001 b0.001
ACEI: angiotensin-converting enzyme inhibitor; AF: atrial fibrillation; ARB: angiotensin II-receptor blocker; COPD: chronic obstructive pulmonary disease; CVA: cerebral vascular accident; ESRD: end-stage renal disease.
AF group exhibited more comorbidities, including hypertension, diabetes mellitus, heart failure, previous vascular diseases, dyslipidemia, cerebral vascular accident (CVA), end-stage renal disease (ESRD), chronic obstructive pulmonary disease (COPD), malignancy, autoimmune diseases, and a higher Charlson index than non-AF group. More patients with AF were taking aspirin, clopidogrel, warfarin, angiotensinconverting enzyme inhibitors/angiotensin II-receptor blockers (ACEIs/ARBs) and statin as compared to patients in non-AF group. 3.2. Atrial fibrillation and the risk of dementia During the follow-up period, 29,012 AF patients (8.7%) experienced dementia. The annual incidence of dementia was higher in AF than in non-AF groups (2.12% versus 1.50%) (Table 2). The Kaplan–Meier survival analysis showed that AF was associated with a higher risk of dementia during the follow-up period (Log-rank p b 0.001) (Fig. 2). The presence of AF was significantly associated with an increased risk of dementia, with a hazard ratio (HR) of 1.426 (95% confidence interval [CI] = 1.403–1.450; p b 0.001) in the univariate Cox regression analysis; and 1.420 (95% CI = 1.394–1.448; p b 0.001) after adjustments for age, gender, baseline characteristics and medication use (Table 3). In the subgroup analysis, AF was consistently associated with a higher risk of dementia (Fig. 3). Even in subjects (n = 158,963) without any underlying disease listed in Table 1 (hypertension, diabetes, heart failure, vascular disease, dyslipidemia, CVA, ESRD, COPD, malignancy
Table 2 Incidence (per 100 person-years) of dementia in patients with and without AF. Groups
Number of events
Number of patients
Person-years
Incidence*
With AF Without AF Total
29,012 27,889 56,901
332,665 332,665 665,330
1,368,674 1,864,040 3,232,714
2.12 1.50 1.76
AF: atrial fibrillation. ⁎ Number of new-onset dementia per 100 person-years of follow-up.
Fig. 2. Cumulative event rate of dementia in patients with or without AF. Log rank test demonstrated significantly different risk of dementia between two groups. As the cumulative incidence curve showed, AF patients had a higher risk of dementia compared to non-AF subjects. AF = atrial fibrillation.
and autoimmune disease), AF was still significantly associated with more occurrence of dementia with an adjusted HR of 1.314 (95% CI = 1.249–1.381; p b 0.001). 3.3. Risk factors of incident dementia in patients with atrial fibrillation Among patients with AF, those who had occurrence of dementia were older, mostly female and had more underlying diseases. The CHADS2 and CHA2DS2-VASc scores were higher in those with dementia than without dementia. More patients experiencing occurrence of dementia were taking aspirin while fewer patients were using warfarin. Low income level was also associated with more incident dementia (Table 4). Similar to the prediction of ischemic stroke, patients with a CHADS2 score of 0 were not truly low risk for dementia with an annual incidence of 0.63%. These patients can be further stratified by the CHA2DS2-VASc score and the risk of dementia can be as high as 1.72% for those with a CHA 2 DS 2-VASc score of 3. On the contrary, patients with a CHA2 DS 2-VASc score of 0 had a truly low risk of dementia during the follow-up (0.19% per year). In both univariate and multivariate Cox regression models, the CHADS2 and CHA2DS2-VASc scores were significant predictors of dementia with an adjusted hazard ratios of 1.520 (95% CI = 1.505–1.535; p b 0.001) and 1.497 (95% CI = 1.485–1.508; p b 0.001) per 1 increment of the CHADS2 and CHA2DS2-VASc scores, respectively (Table 5). With increase of every point of CHADS2 and CHA2DS2-VASc scores, the incidence augmented gradually, up to 2.12% per year for CHADS2 score of 6 and 5.74% per year for CHA2DS2-VASc score of 9 (Table 6). The c-index for CHA2DS2-VASc in predicting dementia (0.611, 95% CI = 0.608–0.614) was significantly higher than the CHADS2 score (0.589, 95% CI = 0.586–0.592) (DeLong test p b 0.001). For patients with a CHADS2 score of 0 (n = 40,179), the patients' CHA2DS2-VASc scores ranged from 0 to 3 and the annual risk of dementia were 0.19%, 0.75%, 1.54% and 1.72% for patients with a CHA2DS2-VASc score of 0, 1, 2 and 3, respectively. 4. Discussion 4.1. Main findings In this nationwide cohort study, we compared the risk of dementia between AF and non-AF subjects. Our principal findings were as follows: (1) The annual incidence of dementia was 2.12% for AF patients, and the risk of dementia was 42% higher than non-AF subjects after
28
J.-N. Liao et al. / International Journal of Cardiology 199 (2015) 25–30
Table 3 Cox regression models on the relationship between the presence of atrial fibrillation and risk of dementia. ACEI: angiotensin-converting enzyme inhibitor; ARB: angiotensin II-receptor blocker; CI: confidence interval; COPD: chronic obstructive pulmonary disease; CVA: cerebral vascular accident; HR: hazard ratio. Models
HR (AF versus non-AF patients)
95% CI
p value
Model 1: unadjusted regression analysis Model 2: adjusted for age, gender Model 3: adjusted for age, gender, hypertension, diabetes mellitus, heart failure, vascular diseases, dyslipidemia, CVA, end-stage renal disease, COPD, malignancy, autoimmune diseases and Charlson index Model 4: adjusted for all variables in model 3, and the use of aspirin, clopidogrel, warfarin, ACEIs/ARBs and statin Model 5: adjusted for all variables in model 4, and income level
1.426 1.681 1.469
1.403–1.450 1.653–1.709 1.443–1.496
b0.001 b0.001 b0.001
1.416 1.420
1.389–1.443 1.394–1.448
b0.001 b0.001
adjusting for age, gender and baseline characteristics; (2) both CHADS2 and CHA2DS2-VASc scores were significant predictors of dementia among AF patients, and the CHA2DS2-VASc score performed better than CHADS2 score. 4.2. Atrial fibrillation and dementia With an increasingly aging population, the disease burden of dementia and AF is greater since both conditions mostly affect elderly patients. Many studies proposed significant correlation between AF and dementia [4–7]. Marzona et al. conducted a post-hoc analysis of two randomized controlled trials (“ONTARGET” and “TRANSCEND” trials) incorporating 31,506 patients which included 1016 participants
with AF at the baseline and an additional 2052 participants exhibiting AF during the follow-up [6]. The study demonstrated that the presence of AF was associated with cognitive and functional decline and increased the risk of dementia by 30% [6]. Forti et al. reported that AF was not significantly associated with the occurrence of dementia in patients with normal cognitive functions at baseline [22]. The present study enrolled a large number of patients and to the best of our knowledge, represents the largest analysis examining AF, cognitive function and dementia. We demonstrated that AF was significantly associated with the occurrence of dementia with a HR of 1.420, consistent with the risk reported by Santangeli et al. [7]. Although AF patients enrolled in our study possessed more comorbidities, AF was still an important risk factor of dementia even in subjects who did not
Fig. 3. AF and risk of dementia in different subgroups of patients. In the subgroup analysis, the presence of AF was consistently associated with a higher risk of dementia. *Adjusted for age, gender, hypertension, diabetes mellitus, heart failure, vascular diseases, dyslipidemia, CVA, ESRD, COPD, malignancy, autoimmune diseases, the use of aspirin, clopidogrel, warfarin, ACEI/ARB, statin, Charlson index, income level, and systemic diseases. +Systemic diseases included hypertension, diabetes, heart failure, vascular disease, dyslipidemia, CVA, ESRD, COPD, malignancy and autoimmune disease. ACEI = angiotensin-converting enzyme inhibitor; AF = atrial fibrillation; ARB = angiotensin II-receptor blocker; CI = confidence interval; COPD = chronic obstructive pulmonary disease; CVA = cerebral vascular accident; ESRD = end-stage renal disease; HR = hazard ratio.
J.-N. Liao et al. / International Journal of Cardiology 199 (2015) 25–30 Table 4 Baseline characteristics of AF patients with and without development of dementia. Variables
With dementia (n = 29,012)
Without dementia (n = 303,653)
p value
Age, years Age ≥ 65 years, n (%) Male gender, n (%) Underlying diseases, n (%) Hypertension Diabetes mellitus Heart failure Vascular disease Dyslipidemia CVA ESRD COPD Malignancy Autoimmune disease Charlson index CHADS2 score CHA2DS2-VASc score Medication use, n (%) Aspirin Clopidogrel Warfarin ACEIs/ARBs Statin Income level Low Medium High
76.2 ± 8.3 26,707 (92.1%) 14,576 (50.2%)
69.7 ± 13.2 213,491 (70.3%) 171,424 (56.5%)
b0.001 b0.001 b0.001
21,211 (73.1%) 8365 (28.8%) 11,543 (39.8%) 2260 (7.8%) 6432 (22.2%) 12,473 (43.0%) 1770 (6.1%) 10,419 (35.9%) 2500 (8.6%) 1564 (5.4%) 2.6 ± 2.0 2.9 ± 1.6 4.4 ± 1.8
202,958 (66.8%) 84,142 (27.7%) 122,259 (40.3%) 27,048 (8.9%) 75,889 (25.0%) 95,107 (31.3%) 22,462 (7.4%) 102,691 (33.8%) 32,159 (10.6%) 16,299 (5.4%) 2.5 ± 2.0 2.4 ± 1.6 3.6 ± 2.0
b0.001 b0.001 0.116 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 0.861 b0.001 b0.001 b0.001
10,256 (35.4%) 1167 (4.0%) 2732 (9.4%) 7341 (25.3%) 362 (1.2%)
95,240 (31.4%) 12,772 (4.2%) 30,596 (10.1%) 76,644 (25.2%) 3937 (1.3%)
b0.001 0.136 b0.001 0.816 0.495
16,332 (58.3%) 8891 (30.6%) 3789 (13.1%)
161,348 (53.1%) 102,256 (33.7%) 40,049 (13.2%)
b0.001 b0.001 b0.001
ACEI: angiotensin-converting enzyme inhibitor; AF: atrial fibrillation; ARB: angiotensin II-receptor blocker; COPD: chronic obstructive pulmonary disease; CVA: cerebral vascular accident; ESRD: end-stage renal disease.
have any underlying disease. Thus, the higher risk of dementia in AF observed in the present study cannot be explained by the baseline differences between AF and non-AF patients. 4.3. Pathogenesis of dementia in atrial fibrillation What is the possible explanation for the link between AF and dementia? This close relationship may be in part due to similar clinical risk factors for both conditions, such as age, hypertension, diabetes, smoking and systemic inflammation [9,23]. An increased systemic inflammatory status, a phenomenon proposed to be present in AF [24], has also been suggested to be associated with accelerated cognitive decline and increased risk for dementia [25,26]. In addition to inflammation, several studies showed that atherosclerosis was also involved in the pathogenesis of dementia by promoting both vascular and plaque/tangle pathologies [27,28].
Table 5 Cox regression analysis using CHADS2 and CHA2DS2-VASc scores in predicting dementia.
CHADS2 score CHA2DS2-VASc score
Hazard ratio (95% CI)
p value
Adjusted hazard ratio (95% CI)
p value
1.466 (1.455–1.476) 1.425 (1.417–1.434)
b0.001 b0.001
1.520 (1.505–1.535)⁎ 1.497 (1.485–1.508)+
b0.001 b0.001
ACEI: angiotensin-converting enzyme inhibitor; ARB: angiotensin II-receptor blocker; CI: confidence interval. ⁎ The CHADS2 score was adjusted for the gender and co-morbidities other than the components of the CHADS2 scoring system (previous autoimmune disease, chronic obstructive pulmonary disease, dyslipidemia, end-stage renal disease, malignancy, vascular diseases, Charlson index, use of aspirin, clopidogrel, warfarin, ACEIs/ARBs, statin and income level) in the multivariate regression analysis. + The CHA2DS2-VASc score was adjusted for co-morbidities other than the components of the CHA 2 DS 2 -VASc scoring system (autoimmune disease, chronic obstructive pulmonary disease, dyslipidemia, end-stage renal disease, malignancy, Charlson index, use of aspirin, clopidogrel, warfarin, ACEIs/ARBs, statin and income level) in the multivariate regression analysis.
29
Table 6 Incidence (per 100-person-years) of dementia in patients with different CHADS2 and CHA2DS2-VASc scores. (a) CHADS2 score
Number of patients
Number of dementia
Person-years Incidence⁎
0 1 2 3 4 5 6 Total
40,179 68,841 72,576 63,053 46,012 31,395 10,609 332,665
1634 4501 6196 6327 5309 3795 1250 29,012
260,030 359,135 303,614 216,557 136,293 73,714 19,331 1,368,674
(b) CHA2DS2-VASc score
Number of patients
Number of dementia
Person-years Incidence⁎
0 1 2 3 4 5 6 7 8 9 Total
15,867 36,276 49,057 58,008 58,246 49,160 35,728 21,268 7937 1118 332,665
192 1237 3114 4946 5896 5480 4443 2683 927 94 29,012
103,006 222,642 257,399 256,646 217,188 153,469 94,437 47,474 14,775 1638 1,368,674
0.63 1.25 2.04 2.92 3.90 5.15 6.47 2.12
0.19 0.56 1.21 1.93 2.71 3.57 4.70 5.65 6.27 5.74 2.12
⁎ Number of ischemic strokes per 100 person-years of follow-up.
Since systemic atherosclerosis, as indicated by increased carotid intima-media thickness, is an important risk factor of AF [29–31], this may be responsible for the relationship between AF and dementia. Importantly, AF is a well-known risk factor of clinical or silent cerebral infarctions, which may predispose patients to an increased risk of dementia [32]. For example, Dublin et al. performed a study enrolling 328 participants undergoing autopsy, where 134 participants having AF were more likely to have gross infarcts while some infarcts were not clinically recognized before death. In pathological analysis, AF increased neuritic plaques and neurofibrillary tangles by 47% and 40%, respectively [33]. Indeed, multiple episodes of small subclinical stroke may underlie the early cognitive decline in the long term [4]. Nevertheless, in the post-hoc analysis of the “ONTARGET” and “TRANSCEND” trials, AF-related risk of dementia was consistent among patients with and without stroke, which suggested that cerebral infarction caused by AF-related thromboembolism may not be the only explanation for this observation [6]. Global brain hypoperfusion due to irregular heartbeat, either low or high ventricular rate, may also account for the increased risk of dementia in AF patients [34].
4.4. Predictors of dementia in AF Since AF patients had a higher risk of dementia, how to predict its occurrence in the AF population is an important issue. At present, a simple easy way to predict incident dementia in AF is lacking. In the present study using the Taiwan AF cohort, both CHADS2 and CHA2DS2-VASc scores were significant predictors of dementia in AF subjects. The CHA2DS2-VASc score performed better than CHADS2 in predicting dementia assessed by the C-statistic. We also reported the incidence of dementia for AF patients with each CHADS2 and CHA2DS2-VASc scores. This information is useful when caring AF patients since the risk of ischemic stroke and dementia could be estimated simultaneously using the same scheme, the CHA2DS2-VASc score. Based on the findings of the present study, physicians should be watchful to clinical manifestations implying any cognitive decline and functional impairment for AF patients, especially those with a high CHA2DS2-VASc score.
30
J.-N. Liao et al. / International Journal of Cardiology 199 (2015) 25–30
4.5. Study limitations Our study is the first population-based investigation focusing on the association between AF and dementia, which enrolled a large sample of AF and non-AF subjects, enabling a long-term follow-up of incident dementia. However, there are still some limitations of our study. First, the diagnosis of dementia was based on the diagnostic codes registered by the physicians responsible for the treatments of patients, and data about the degree and severity of cognitive impairment were lacking. Besides, information about the subtypes of AF (paroxysmal or non-paroxysmal) was not available, and it was also a common limitation in previous studies [4,5]. Second, multiple baseline characteristics were different between AF and non-AF groups since the present study was not a prospective trial. However, AF still remained as an independent risk factor of dementia even after the adjustment for these confounders in the multivariate Cox analysis. Even when we only focused on patients without important comorbidities which were different in distribution between AF and non-AF groups, the risk of dementia remained higher among AF patients. Therefore, the increased risk of dementia in AF group may not be fully explained by these baseline differences. Last, although we demonstrated that the risk of dementia was higher in AF patients with a high CHA2DS2-VASc score, it remains unknown whether aggressive managements of underlying diseases can decrease the risk of dementia. Large-scale prospective randomized trials are necessary to study this issue. 5. Conclusion In this nationwide cohort study, we demonstrated that AF was independently associated with a higher risk of dementia. The CHA2DS2-VASc score was useful in predicting dementia in AF patients, and superior to CHADS2 score. Conflict of interest statement No conflict of interest was declared. Acknowledgments This work was supported in part by grants from the National Science Council (NSC98-2410-H-010-003-MY2), and Taipei Veterans General Hospital (V99C1-140, V99A-153, V100D-002-3, V101D-001-2, and V102B-025). References [1] C.T. January, L.S. Wann, J.S. Alpert, et al., AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the Heart Rhythm Society, J. Am. Coll. Cardiol. 64 (2014) 2246–2280. [2] A.J. Camm, G.Y. Lip, R. De Caterina, et al., 2012 focused update of the ESC Guidelines for the management of atrial fibrillation: an update of the 2010 ESC Guidelines for the management of atrial fibrillation. Developed with the special contribution of the European Heart Rhythm Association, Eur. Heart J. 33 (2012) 2719–2747. [3] G.V. Naccarelli, H. Varker, J. Lin, K.L. Schulman, Increasing prevalence of atrial fibrillation and flutter in the United States, Am. J. Cardiol. 104 (2009) 1534–1539. [4] T.J. Bunch, J.P. Weiss, B.G. Crandall, et al., Atrial fibrillation is independently associated with senile, vascular, and Alzheimer's dementia, Heart Rhythm. 7 (2010) 433–437. [5] S. Dublin, M.L. Anderson, S.J. Haneuse, et al., Atrial fibrillation and risk of dementia: a prospective cohort study, J. Am. Geriatr. Soc. 59 (2011) 1369–1375. [6] I. Marzona, M. O'Donnell, K. Teo, et al., Increased risk of cognitive and functional decline in patients with atrial fibrillation: results of the ONTARGET and TRANSCEND studies, CMAJ 184 (2012) E329–E336.
[7] P. Santangeli, L. Di Biase, R. Bai, et al., Atrial fibrillation and the risk of incident dementia: a meta-analysis, Heart Rhythm. 9 (2012) 1761–1768. [8] S. Udompanich, G.Y. Lip, S. Apostolakis, D.A. Lane, Atrial fibrillation as a risk factor for cognitive impairment: a semi-systematic review, QJM 106 (2013) 795–802. [9] R.A. Whitmer, S. Sidney, J. Selby, S.C. Johnston, K. Yaffe, Midlife cardiovascular risk factors and risk of dementia in late life, Neurology 64 (2005) 277–281. [10] FP Lu, KP Lin, HK Kuo, Diabetes and the risk of multi-system aging phenotypes: a systematic review and meta-analysis, PLoS One 4 (2009) e4144. [11] S.I. Sharp, D. Aarsland, S. Day, H. Sonnesyn, C. Ballard, Hypertension is a potential risk factor for vascular dementia: systematic review, Int. J. Geriatr. Psychiatry 26 (2011) 661–669. [12] S.J. Chen, C.J. Liu, T.F. Chao, et al., Dental scaling and atrial fibrillation: a nationwide cohort study, Int. J. Cardiol. 168 (2013) 2300–2303. [13] L.J. Lin, M.H. Cheng, C.H. Lee, D.C. Wung, C.L. Cheng, Y.H. Kao Yang, Compliance with antithrombotic prescribing guidelines for patients with atrial fibrillation—a nationwide descriptive study in Taiwan, Clin. Ther. 30 (2008) 1726–1736. [14] C.H. Chang, Y.C. Lee, C.T. Tsai, et al., Continuation of statin therapy and a decreased risk of atrial fibrillation/flutter in patients with and without chronic kidney disease, Atherosclerosis 232 (2014) 224–230. [15] P.L. Chen, W.J. Lee, W.Z. Sun, Y.J. Oyang, J.L. Fuh, Risk of dementia in patients with insomnia and long-term use of hypnotics: a population-based retrospective cohort study, PLoS One 7 (2012) e49113. [16] C.L. Cheng, Y.H. Kao, S.J. Lin, C.H. Lee, M.L. Lai, Validation of the National Health Insurance Research Database with ischemic stroke cases in Taiwan, Pharmacoepidemiol. Drug Saf. 20 (2011) 236–242. [17] C.C. Lin, M.S. Lai, C.Y. Syu, S.C. Chang, F.Y. Tseng, Accuracy of diabetes diagnosis in health insurance claims data in Taiwan, J. Formos. Med. Assoc. 104 (2005) 157–163. [18] V. de Groot, H. Beckerman, G.J. Lankhorst, L.M. Bouter, How to measure comorbidity. A critical review of available methods, J. Clin. Epidemiol. 56 (2003) 221–229. [19] B.F. Gage, A.D. Waterman, W. Shannon, M. Boechler, M.W. Rich, M.J. Radford, Validation of clinical classification schemes for predicting stroke: results from the National Registry of Atrial Fibrillation, JAMA 285 (2001) 2864–2870. [20] G.Y. Lip, R. Nieuwlaat, R. Pisters, D.A. Lane, H.J. Crijns, 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 137 (2010) 263–272. [21] C.L. Perng, C.C. Shen, L.Y. Hu, et al., Risk of depressive disorder following non-alcoholic cirrhosis: a nationwide population-based study, PLoS One 9 (2014) e88721. [22] P. Forti, F. Maioli, N. Pisacane, E. Rietti, F. Montesi, G. Ravaglia, Atrial fibrillation and risk of dementia in non-demented elderly subjects with and without mild cognitive impairment, Neurol. Res. 28 (2006) 625–629. [23] M.T. Schram, S.M. Euser, A.J. de Craen, et al., Systemic markers of inflammation and cognitive decline in old age, J. Am. Geriatr. Soc. 55 (2007) 708–716. [24] M.A. Crandall, B.D. Horne, J.D. Day, et al., Atrial fibrillation and CHADS2 risk factors are associated with highly sensitive C-reactive protein incrementally and independently, Pacing Clin. Electrophysiol. 32 (2009) 648–652. [25] G. Xu, Z. Zhou, W. Zhu, X. Fan, X. Liu, Plasma C-reactive protein is related to cognitive deterioration and dementia in patients with mild cognitive impairment, J. Neurol. Sci. 284 (2009) 77–80. [26] H. Wersching, T. Duning, H. Lohmann, et al., Serum C-reactive protein is linked to cerebral microstructural integrity and cognitive function, Neurology 74 (2010) 1022–1029. [27] A.E. Roher, C. Esh, T.A. Kokjohn, et al., Circle of willis atherosclerosis is a risk factor for sporadic Alzheimer's disease, Arterioscler. Thromb. Vasc. Biol. 23 (2003) 2055–2062. [28] M. Yarchoan, S.X. Xie, M.A. Kling, et al., Cerebrovascular atherosclerosis correlates with Alzheimer pathology in neurodegenerative dementias, Brain 135 (2012) 3749–3756. [29] J. Heeringa, D.A. van der Kuip, A. Hofman, et al., Subclinical atherosclerosis and risk of atrial fibrillation: the Rotterdam study, Arch. Intern. Med. 167 (2007) 382–387. [30] L.Y. Chen, D.C. Foo, R.C. Wong, et al., Increased carotid intima-media thickness and arterial stiffness are associated with lone atrial fibrillation, Int. J. Cardiol. 168 (2013) 3132–3134. [31] TF Chao, YC Huang, CJ Liu, et al., Acute myocardial infarction in patients with atrial fibrillation with a CHA2DS2-VASc score of 0 or 1: a nationwide cohort study, Heart Rhythm. 11 (2014) 1941–1947. [32] M.D. Ezekowitz, K.E. James, S.M. Nazarian, et al., Silent cerebral infarction in patients with nonrheumatic atrial fibrillation. The veterans affairs stroke prevention in nonrheumatic atrial fibrillation investigators, Circulation 92 (1995) 2178–2182. [33] S. Dublin, M.L. Anderson, S.R. Heckbert, et al., Neuropathologic changes associated with atrial fibrillation in a population-based autopsy cohort, J. Gerontol. A Biol. Sci. Med. Sci. 69 (2014) 609–615. [34] F. Cacciatore, G. Testa, A. Langellotto, et al., Role of ventricular rate response on dementia in cognitively impaired elderly subjects with atrial fibrillation: a 10-year study, Dement. Geriatr. Cogn. Disord. 34 (2012) 143–148.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Risk and prediction of dementia in patients with atrial fibrillation — A nationwide population-based cohort study Jo-Nan Liao a,b,1, Tze-Fan Chao a,b,1, Chia-Jen Liu c,d, Kang-Ling Wang a,b, Su-Jung Chen d,e, Ta-Chuan Tuan a,b, Yenn-Jiang Lin a,b, Shih-Lin Chang a,b, Li-Wei Lo a,b, Yu-Feng Hu a,b, Fa-Po Chung a,b, Hsuan-Ming Tsao b,f,⁎, Tzeng-Ji Chen g, Gregory Y.H. Lip h,2, Shih-Ann Chen a,b,⁎⁎,2 a
Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan Institute of Clinical Medicine, and Cardiovascular Research Center, National Yang-Ming University, Taipei, Taiwan c Division of Hematology and Oncology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan d Institute of Public Health and School of Medicine, National Yang-Ming University, Taipei, Taiwan e Division of Infectious Diseases, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan f National Yang-Ming University Hospital, Taiwan g Department of Family Medicine, Taipei Veterans General Hospital, Taipei, Taiwan h University of Birmingham Centre for Cardiovascular Sciences, City Hospital, Birmingham, United Kingdom b
a r t i c l e
i n f o
Article history: Received 2 February 2015 Received in revised form 20 June 2015 Accepted 27 June 2015 Available online 6 July 2015 Keywords: Atrial fibrillation Dementia CHADS2 score CHA2DS2-VASc score
a b s t r a c t Background: Atrial fibrillation (AF) is associated with an increased risk of cognitive impairment and functional decline, and may contribute to development of dementia. Objectives: Data from a nationwide large-scale population-based cohort study are lacking. Besides, how best to predict the occurrence of incident dementia among AF subjects remains uncertain. Methods: A total of 332,665 AF subjects without dementia were identified as the study group from the “National Health Insurance Research Database” in Taiwan. For each study patient, one age- and sex-matched subject without AF and dementia was selected as the control group. The study end point was occurrence of dementia, and the usefulness of CHADS2 and CHA2DS2-VASc scores in predicting dementia was analyzed. Results: During the follow-up, 29,012 AF patients experienced dementia with an annual incidence of 2.12%, higher than non-AF subjects (1.50%). Patients with AF possessed a higher risk of dementia with a hazard ratio (HR) of 1.420 after adjustments for age, gender, baseline differences and medication use. Among AF patients, the CHADS2 and CHA2DS2-VASc scores were significant predictors of dementia with an adjusted HR of 1.520 and 1.497 per 1 increment of the CHADS2 and CHA2DS2-VASc scores, respectively. The c-index for CHA2DS2-VASc in predicting dementia (0.611, 95% confidence interval [CI] = 0.608–0.614) was significantly higher than the CHADS2 score (0.589, 95% CI = 0.586–0.592) (DeLong test p b 0.001). Conclusions: In this nationwide cohort study, AF was independently associated with a higher risk of dementia. The CHA2DS2-VASc score can be used to estimate the risk of dementia in AF patients. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia in clinical practice [1], and carries a five-fold risk of stroke, a three-fold incidence of heart failure and a higher mortality [2].
⁎ Correspondence to: HM Tsao, National Yang-Ming University Hospital, No 152, Xin-Min road, Yi-Lan, Taiwan. ⁎⁎ Correspondence to: SA Chen, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, No. 201, Sec. 2, Shih-Pai Road, Taipei, Taiwan. E-mail addresses: [email protected] (H.-M. Tsao), [email protected] (S.-A. Chen). 1 Dr. Jo-Nan Liao and Dr. Tze-Fan Chao contributed equally to this study. 2 Joint senior authors.
http://dx.doi.org/10.1016/j.ijcard.2015.06.170 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.
AF prevalence increases with age and affects up to 9% of the population older than 80 years old [3]. Aging of population also leads to increased risk of dementia, a disorder which is characteristic of memory and cognitive impairment. Various studies suggested that AF is associated with an increased risk of cognitive impairment and functional decline and may contribute to development of dementia [4–7]. In a systematic review, Udompanich et al. [8] found that among cross-sectional studies, patients with AF had a 1.7 (95% CI = 1.2–2.5) to 3.3 (95% CI = 1.6–6.5) greater risk of cognitive impairment, and a 2.3-fold (95% CI = 1.4–3.7) increased risk of dementia, compared to patients in sinus rhythm; however, there was marked heterogeneity in the design, size and quality of studies and reporting of the data which precluded formal meta-analysis. Similarly, Santangeli et al. performed a meta-analysis incorporating
26
J.-N. Liao et al. / International Journal of Cardiology 199 (2015) 25–30
8 studies, a total of 77,668 patients, and suggested that AF increased the risk of incident dementia by 42% [7]. However, 3 studies enrolled in the meta-analysis did not show a significant association between AF and dementia in the original investigations. Importantly, Asians were not included in most studies. Besides, the risk factors of dementia and how to predict its occurrence among AF patients were not clear. The CHADS2 and CHA2DS2-VASc scores were commonly used to guide anti-thrombotic therapy for stroke prevention in AF. Several components of these 2 scoring schemes, such as an old age, hypertension, and diabetes, were reported to be important risk factors of dementia [9–11]. Furthermore, the occurrence of ischemic stroke, whose risk could be estimated using the CHADS2 and CHA2DS2-VASc scores, would significantly predispose patients to a higher risk of subsequent dementia. However, the relationships between the CHADS 2 and CHA 2DS 2-VASc scores and dementia in AF patients have not been previously studied. In the present study, we aim to investigate whether AF is associated with a higher risk of dementia in Asians using a nationwide database which would have less selection biases and could more accurately represent the real-world condition. We will also test the hypothesis that the CHADS2 and CHA2DS2-VASc scores could be used to predict the occurrence of dementia in AF.
definition in NHIRD has been validated previously [13,14]. The diagnosis of dementia was based on the ICD-9-CM codes (290.0–290.4, 331.0) registered by the physicians responsible for the treatments of patients [15]. For each study patient, one age- and sexmatched subject without past history of AF and dementia was selected as the control group. The flow chart of the enrollment of study population was shown in Fig. 1. Information about important comorbid conditions of each individual was retrieved from the medical claims based on the ICD-9-CM codes. We defined patients with a certain disease only when it was a discharge diagnosis or repeatedly confirmed more than twice in outpatient department. The diagnostic accuracies of important comorbidities in NHIRD, such as hypertension, diabetes mellitus, heart failure, myocardial infarction, hyperlipidemia and chronic obstructive pulmonary disease, have been validated before [16,17]. The Charlson index was assessed as the overall severity of comorbidities [18]. 2.3. Calculations of CHADS2 and CHA2DS2-VASc scores and definition of income level and clinical end points
2. Methods
The CHADS2 score was calculated for each patient by assigning 1 point each for age ≥ 75 years, hypertension, diabetes mellitus, and heart failure, and 2 points each for a previous stroke or transient ischemic attack (TIA) [19]. The CHA2DS2-VASc score was calculated for each patient by assigning 1 point each for age between 65 and 74 years, history of hypertension, diabetes, recent cardiac failure, vascular disease (myocardial infarction or peripheral artery disease), and female gender, and 2 points each for a history of a stroke, TIA, or age ≥ 75 years [20]. Insurance premiums, calculated according to the beneficiary's total income, were used to estimate monthly income. Monthly income was grouped into low income (monthly income b 20,000 New Taiwan Dollar [NTD]), medium income (20,000 NTD ≤ monthly income b 40,000 NTD), and high income (monthly income ≥ 40,000 NTD) [21]. The study end point was the occurrence of dementia (ICD-9-CM codes: 290.0–290.4, 331.0) during the follow-up period [15].
2.1. Database
2.4. Statistical analysis
This study used the “National Health Insurance Research Database” (NHIRD) released by the Taiwan National Health Research Institutes (NHRI). The National Health Insurance (NHI) system is a mandatory universal health insurance program that offers comprehensive medical care coverage to all Taiwanese residents. NHIRD contains detailed health care data from N23 million enrollees, representing N99% of Taiwan's population. In this cohort dataset, the patients' original identification numbers were encrypted to protect privacy, and the encrypting procedure was consistent so that a linkage of the claims belonging to each patient was feasible within the NHI database, and could be followed continuously. Numerous scientific research papers have been published using data from NHIRD (http://nhird.nhri.org.tw/en/Research.html). The large sample size of this database provided a good opportunity to study the association between AF and dementia.
Data were presented as the mean value and standard deviation for normally distributed continuous variables and proportions for categorical variables. Comparisons of the continuous variables were performed using an unpaired two-tailed t-test for normally distributed variables and Mann–Whitney rank-sum test for skewed variables. Nominal variables were compared by chi-square analysis with a Yates correction or Fisher's exact test. The risk of dementia occurrence was assessed using the Cox regression analysis. The cumulative incidence curve of dementia was plotted via the Kaplan–Meier method, with statistical significance examined by the log-rank test. All statistical significances were set at a p b 0.05 using the SPSS version 17.0 statistical software (SPSS, Chicago, IL, USA).
2.2. Study population From January 1, 1996 to December 31, 2011, a total of 332,665 patients with AF with age ≥ 20 years and no history of dementia for at least 1 year preceding study enrollment were retrieved from the NHIRD since the record of NHIRD can be traced back to 1995. AF was diagnosed using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes (427.31). To ensure the accuracy of diagnosis, we defined patients with AF only when it was listed as a discharge diagnosis or confirmed more than twice in the outpatient department [12]. The diagnostic accuracy of AF using this
3. Results 3.1. Baseline characteristics of patients The mean age of the study population was 70.3 ± 13.0 years, and 55.9% of patients were men. The baseline characteristics of AF and non-AF groups were shown in Table 1. Hypertension was the most prevalent comorbidity, which was present in 67.4% of AF patients.
Fig. 1. Flow diagram of study enrollment. From 1996 to 2011, a total of 332,665 patients with AF and no history of dementia were identified from the NHIRD. For each AF patient, one age- and gender-matched subject without past history of AF and dementia was selected from the NHIRD to constitute the control group. AF = atrial fibrillation; NHIRD = National Health Insurance Research Database.
J.-N. Liao et al. / International Journal of Cardiology 199 (2015) 25–30
27
Table 1 Baseline characteristics of subjects with and without AF. Variables
AF patients (n = 332,665)
Non-AF patients (n = 332,665)
p value
Age, years Age ≥ 65 years, n (%) Male gender, n (%)
70.3 ± 13.0 240,198 (72.2%) 186,000 (55.9%)
70.3 ± 13.0 240,198 (72.2%) 186,000 (55.9%)
0.994 1.000 1.000
Underlying diseases, n (%) Hypertension Diabetes mellitus Heart failure Vascular disease Dyslipidemia CVA ESRD COPD Malignancy Autoimmune disease Charlson index
224,169 (67.4%) 92,547 (27.8%) 133,802 (40.2%) 29,308 (8.8%) 82,321 (24.7%) 107,580 (32.3%) 24,232 (7.3%) 113,110 (34%) 34,659 (10.4%) 17,863 (5.4%) 2.5 ± 2.0
153,275 (46.1%) 65,802 (19.8%) 30,705 (9.2%) 9082 (2.7%) 67,376 (20.3%) 57,059 (17.2%) 10,801 (3.2%) 72,387 (21.8%) 25,238 (7.6%) 14,751 (4.4%) 1.5 ± 1.7
b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001
Medication use, n (%) Aspirin Clopidogrel Warfarin ACEIs/ARBs Statin
105,496 (31.7%) 13,939 (4.2%) 33,328 (10.0%) 83,985 (25.2%) 4299 (1.3%)
58,816 (17.7%) 6193 (1.9%) 1919 (0.6%) 69,982 (21%) 26,992 (8.1%)
b0.001 b0.001 b0.001 b0.001 b0.001
Income level Low Medium High
177,680 (53.4%) 111,147 (33.4%) 43,838 (13.2%)
228,641 (68.7%) 63,942 (19.2%) 40,082 (12.0%)
b0.001 b0.001 b0.001
ACEI: angiotensin-converting enzyme inhibitor; AF: atrial fibrillation; ARB: angiotensin II-receptor blocker; COPD: chronic obstructive pulmonary disease; CVA: cerebral vascular accident; ESRD: end-stage renal disease.
AF group exhibited more comorbidities, including hypertension, diabetes mellitus, heart failure, previous vascular diseases, dyslipidemia, cerebral vascular accident (CVA), end-stage renal disease (ESRD), chronic obstructive pulmonary disease (COPD), malignancy, autoimmune diseases, and a higher Charlson index than non-AF group. More patients with AF were taking aspirin, clopidogrel, warfarin, angiotensinconverting enzyme inhibitors/angiotensin II-receptor blockers (ACEIs/ARBs) and statin as compared to patients in non-AF group. 3.2. Atrial fibrillation and the risk of dementia During the follow-up period, 29,012 AF patients (8.7%) experienced dementia. The annual incidence of dementia was higher in AF than in non-AF groups (2.12% versus 1.50%) (Table 2). The Kaplan–Meier survival analysis showed that AF was associated with a higher risk of dementia during the follow-up period (Log-rank p b 0.001) (Fig. 2). The presence of AF was significantly associated with an increased risk of dementia, with a hazard ratio (HR) of 1.426 (95% confidence interval [CI] = 1.403–1.450; p b 0.001) in the univariate Cox regression analysis; and 1.420 (95% CI = 1.394–1.448; p b 0.001) after adjustments for age, gender, baseline characteristics and medication use (Table 3). In the subgroup analysis, AF was consistently associated with a higher risk of dementia (Fig. 3). Even in subjects (n = 158,963) without any underlying disease listed in Table 1 (hypertension, diabetes, heart failure, vascular disease, dyslipidemia, CVA, ESRD, COPD, malignancy
Table 2 Incidence (per 100 person-years) of dementia in patients with and without AF. Groups
Number of events
Number of patients
Person-years
Incidence*
With AF Without AF Total
29,012 27,889 56,901
332,665 332,665 665,330
1,368,674 1,864,040 3,232,714
2.12 1.50 1.76
AF: atrial fibrillation. ⁎ Number of new-onset dementia per 100 person-years of follow-up.
Fig. 2. Cumulative event rate of dementia in patients with or without AF. Log rank test demonstrated significantly different risk of dementia between two groups. As the cumulative incidence curve showed, AF patients had a higher risk of dementia compared to non-AF subjects. AF = atrial fibrillation.
and autoimmune disease), AF was still significantly associated with more occurrence of dementia with an adjusted HR of 1.314 (95% CI = 1.249–1.381; p b 0.001). 3.3. Risk factors of incident dementia in patients with atrial fibrillation Among patients with AF, those who had occurrence of dementia were older, mostly female and had more underlying diseases. The CHADS2 and CHA2DS2-VASc scores were higher in those with dementia than without dementia. More patients experiencing occurrence of dementia were taking aspirin while fewer patients were using warfarin. Low income level was also associated with more incident dementia (Table 4). Similar to the prediction of ischemic stroke, patients with a CHADS2 score of 0 were not truly low risk for dementia with an annual incidence of 0.63%. These patients can be further stratified by the CHA2DS2-VASc score and the risk of dementia can be as high as 1.72% for those with a CHA 2 DS 2-VASc score of 3. On the contrary, patients with a CHA2 DS 2-VASc score of 0 had a truly low risk of dementia during the follow-up (0.19% per year). In both univariate and multivariate Cox regression models, the CHADS2 and CHA2DS2-VASc scores were significant predictors of dementia with an adjusted hazard ratios of 1.520 (95% CI = 1.505–1.535; p b 0.001) and 1.497 (95% CI = 1.485–1.508; p b 0.001) per 1 increment of the CHADS2 and CHA2DS2-VASc scores, respectively (Table 5). With increase of every point of CHADS2 and CHA2DS2-VASc scores, the incidence augmented gradually, up to 2.12% per year for CHADS2 score of 6 and 5.74% per year for CHA2DS2-VASc score of 9 (Table 6). The c-index for CHA2DS2-VASc in predicting dementia (0.611, 95% CI = 0.608–0.614) was significantly higher than the CHADS2 score (0.589, 95% CI = 0.586–0.592) (DeLong test p b 0.001). For patients with a CHADS2 score of 0 (n = 40,179), the patients' CHA2DS2-VASc scores ranged from 0 to 3 and the annual risk of dementia were 0.19%, 0.75%, 1.54% and 1.72% for patients with a CHA2DS2-VASc score of 0, 1, 2 and 3, respectively. 4. Discussion 4.1. Main findings In this nationwide cohort study, we compared the risk of dementia between AF and non-AF subjects. Our principal findings were as follows: (1) The annual incidence of dementia was 2.12% for AF patients, and the risk of dementia was 42% higher than non-AF subjects after
28
J.-N. Liao et al. / International Journal of Cardiology 199 (2015) 25–30
Table 3 Cox regression models on the relationship between the presence of atrial fibrillation and risk of dementia. ACEI: angiotensin-converting enzyme inhibitor; ARB: angiotensin II-receptor blocker; CI: confidence interval; COPD: chronic obstructive pulmonary disease; CVA: cerebral vascular accident; HR: hazard ratio. Models
HR (AF versus non-AF patients)
95% CI
p value
Model 1: unadjusted regression analysis Model 2: adjusted for age, gender Model 3: adjusted for age, gender, hypertension, diabetes mellitus, heart failure, vascular diseases, dyslipidemia, CVA, end-stage renal disease, COPD, malignancy, autoimmune diseases and Charlson index Model 4: adjusted for all variables in model 3, and the use of aspirin, clopidogrel, warfarin, ACEIs/ARBs and statin Model 5: adjusted for all variables in model 4, and income level
1.426 1.681 1.469
1.403–1.450 1.653–1.709 1.443–1.496
b0.001 b0.001 b0.001
1.416 1.420
1.389–1.443 1.394–1.448
b0.001 b0.001
adjusting for age, gender and baseline characteristics; (2) both CHADS2 and CHA2DS2-VASc scores were significant predictors of dementia among AF patients, and the CHA2DS2-VASc score performed better than CHADS2 score. 4.2. Atrial fibrillation and dementia With an increasingly aging population, the disease burden of dementia and AF is greater since both conditions mostly affect elderly patients. Many studies proposed significant correlation between AF and dementia [4–7]. Marzona et al. conducted a post-hoc analysis of two randomized controlled trials (“ONTARGET” and “TRANSCEND” trials) incorporating 31,506 patients which included 1016 participants
with AF at the baseline and an additional 2052 participants exhibiting AF during the follow-up [6]. The study demonstrated that the presence of AF was associated with cognitive and functional decline and increased the risk of dementia by 30% [6]. Forti et al. reported that AF was not significantly associated with the occurrence of dementia in patients with normal cognitive functions at baseline [22]. The present study enrolled a large number of patients and to the best of our knowledge, represents the largest analysis examining AF, cognitive function and dementia. We demonstrated that AF was significantly associated with the occurrence of dementia with a HR of 1.420, consistent with the risk reported by Santangeli et al. [7]. Although AF patients enrolled in our study possessed more comorbidities, AF was still an important risk factor of dementia even in subjects who did not
Fig. 3. AF and risk of dementia in different subgroups of patients. In the subgroup analysis, the presence of AF was consistently associated with a higher risk of dementia. *Adjusted for age, gender, hypertension, diabetes mellitus, heart failure, vascular diseases, dyslipidemia, CVA, ESRD, COPD, malignancy, autoimmune diseases, the use of aspirin, clopidogrel, warfarin, ACEI/ARB, statin, Charlson index, income level, and systemic diseases. +Systemic diseases included hypertension, diabetes, heart failure, vascular disease, dyslipidemia, CVA, ESRD, COPD, malignancy and autoimmune disease. ACEI = angiotensin-converting enzyme inhibitor; AF = atrial fibrillation; ARB = angiotensin II-receptor blocker; CI = confidence interval; COPD = chronic obstructive pulmonary disease; CVA = cerebral vascular accident; ESRD = end-stage renal disease; HR = hazard ratio.
J.-N. Liao et al. / International Journal of Cardiology 199 (2015) 25–30 Table 4 Baseline characteristics of AF patients with and without development of dementia. Variables
With dementia (n = 29,012)
Without dementia (n = 303,653)
p value
Age, years Age ≥ 65 years, n (%) Male gender, n (%) Underlying diseases, n (%) Hypertension Diabetes mellitus Heart failure Vascular disease Dyslipidemia CVA ESRD COPD Malignancy Autoimmune disease Charlson index CHADS2 score CHA2DS2-VASc score Medication use, n (%) Aspirin Clopidogrel Warfarin ACEIs/ARBs Statin Income level Low Medium High
76.2 ± 8.3 26,707 (92.1%) 14,576 (50.2%)
69.7 ± 13.2 213,491 (70.3%) 171,424 (56.5%)
b0.001 b0.001 b0.001
21,211 (73.1%) 8365 (28.8%) 11,543 (39.8%) 2260 (7.8%) 6432 (22.2%) 12,473 (43.0%) 1770 (6.1%) 10,419 (35.9%) 2500 (8.6%) 1564 (5.4%) 2.6 ± 2.0 2.9 ± 1.6 4.4 ± 1.8
202,958 (66.8%) 84,142 (27.7%) 122,259 (40.3%) 27,048 (8.9%) 75,889 (25.0%) 95,107 (31.3%) 22,462 (7.4%) 102,691 (33.8%) 32,159 (10.6%) 16,299 (5.4%) 2.5 ± 2.0 2.4 ± 1.6 3.6 ± 2.0
b0.001 b0.001 0.116 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 0.861 b0.001 b0.001 b0.001
10,256 (35.4%) 1167 (4.0%) 2732 (9.4%) 7341 (25.3%) 362 (1.2%)
95,240 (31.4%) 12,772 (4.2%) 30,596 (10.1%) 76,644 (25.2%) 3937 (1.3%)
b0.001 0.136 b0.001 0.816 0.495
16,332 (58.3%) 8891 (30.6%) 3789 (13.1%)
161,348 (53.1%) 102,256 (33.7%) 40,049 (13.2%)
b0.001 b0.001 b0.001
ACEI: angiotensin-converting enzyme inhibitor; AF: atrial fibrillation; ARB: angiotensin II-receptor blocker; COPD: chronic obstructive pulmonary disease; CVA: cerebral vascular accident; ESRD: end-stage renal disease.
have any underlying disease. Thus, the higher risk of dementia in AF observed in the present study cannot be explained by the baseline differences between AF and non-AF patients. 4.3. Pathogenesis of dementia in atrial fibrillation What is the possible explanation for the link between AF and dementia? This close relationship may be in part due to similar clinical risk factors for both conditions, such as age, hypertension, diabetes, smoking and systemic inflammation [9,23]. An increased systemic inflammatory status, a phenomenon proposed to be present in AF [24], has also been suggested to be associated with accelerated cognitive decline and increased risk for dementia [25,26]. In addition to inflammation, several studies showed that atherosclerosis was also involved in the pathogenesis of dementia by promoting both vascular and plaque/tangle pathologies [27,28].
Table 5 Cox regression analysis using CHADS2 and CHA2DS2-VASc scores in predicting dementia.
CHADS2 score CHA2DS2-VASc score
Hazard ratio (95% CI)
p value
Adjusted hazard ratio (95% CI)
p value
1.466 (1.455–1.476) 1.425 (1.417–1.434)
b0.001 b0.001
1.520 (1.505–1.535)⁎ 1.497 (1.485–1.508)+
b0.001 b0.001
ACEI: angiotensin-converting enzyme inhibitor; ARB: angiotensin II-receptor blocker; CI: confidence interval. ⁎ The CHADS2 score was adjusted for the gender and co-morbidities other than the components of the CHADS2 scoring system (previous autoimmune disease, chronic obstructive pulmonary disease, dyslipidemia, end-stage renal disease, malignancy, vascular diseases, Charlson index, use of aspirin, clopidogrel, warfarin, ACEIs/ARBs, statin and income level) in the multivariate regression analysis. + The CHA2DS2-VASc score was adjusted for co-morbidities other than the components of the CHA 2 DS 2 -VASc scoring system (autoimmune disease, chronic obstructive pulmonary disease, dyslipidemia, end-stage renal disease, malignancy, Charlson index, use of aspirin, clopidogrel, warfarin, ACEIs/ARBs, statin and income level) in the multivariate regression analysis.
29
Table 6 Incidence (per 100-person-years) of dementia in patients with different CHADS2 and CHA2DS2-VASc scores. (a) CHADS2 score
Number of patients
Number of dementia
Person-years Incidence⁎
0 1 2 3 4 5 6 Total
40,179 68,841 72,576 63,053 46,012 31,395 10,609 332,665
1634 4501 6196 6327 5309 3795 1250 29,012
260,030 359,135 303,614 216,557 136,293 73,714 19,331 1,368,674
(b) CHA2DS2-VASc score
Number of patients
Number of dementia
Person-years Incidence⁎
0 1 2 3 4 5 6 7 8 9 Total
15,867 36,276 49,057 58,008 58,246 49,160 35,728 21,268 7937 1118 332,665
192 1237 3114 4946 5896 5480 4443 2683 927 94 29,012
103,006 222,642 257,399 256,646 217,188 153,469 94,437 47,474 14,775 1638 1,368,674
0.63 1.25 2.04 2.92 3.90 5.15 6.47 2.12
0.19 0.56 1.21 1.93 2.71 3.57 4.70 5.65 6.27 5.74 2.12
⁎ Number of ischemic strokes per 100 person-years of follow-up.
Since systemic atherosclerosis, as indicated by increased carotid intima-media thickness, is an important risk factor of AF [29–31], this may be responsible for the relationship between AF and dementia. Importantly, AF is a well-known risk factor of clinical or silent cerebral infarctions, which may predispose patients to an increased risk of dementia [32]. For example, Dublin et al. performed a study enrolling 328 participants undergoing autopsy, where 134 participants having AF were more likely to have gross infarcts while some infarcts were not clinically recognized before death. In pathological analysis, AF increased neuritic plaques and neurofibrillary tangles by 47% and 40%, respectively [33]. Indeed, multiple episodes of small subclinical stroke may underlie the early cognitive decline in the long term [4]. Nevertheless, in the post-hoc analysis of the “ONTARGET” and “TRANSCEND” trials, AF-related risk of dementia was consistent among patients with and without stroke, which suggested that cerebral infarction caused by AF-related thromboembolism may not be the only explanation for this observation [6]. Global brain hypoperfusion due to irregular heartbeat, either low or high ventricular rate, may also account for the increased risk of dementia in AF patients [34].
4.4. Predictors of dementia in AF Since AF patients had a higher risk of dementia, how to predict its occurrence in the AF population is an important issue. At present, a simple easy way to predict incident dementia in AF is lacking. In the present study using the Taiwan AF cohort, both CHADS2 and CHA2DS2-VASc scores were significant predictors of dementia in AF subjects. The CHA2DS2-VASc score performed better than CHADS2 in predicting dementia assessed by the C-statistic. We also reported the incidence of dementia for AF patients with each CHADS2 and CHA2DS2-VASc scores. This information is useful when caring AF patients since the risk of ischemic stroke and dementia could be estimated simultaneously using the same scheme, the CHA2DS2-VASc score. Based on the findings of the present study, physicians should be watchful to clinical manifestations implying any cognitive decline and functional impairment for AF patients, especially those with a high CHA2DS2-VASc score.
30
J.-N. Liao et al. / International Journal of Cardiology 199 (2015) 25–30
4.5. Study limitations Our study is the first population-based investigation focusing on the association between AF and dementia, which enrolled a large sample of AF and non-AF subjects, enabling a long-term follow-up of incident dementia. However, there are still some limitations of our study. First, the diagnosis of dementia was based on the diagnostic codes registered by the physicians responsible for the treatments of patients, and data about the degree and severity of cognitive impairment were lacking. Besides, information about the subtypes of AF (paroxysmal or non-paroxysmal) was not available, and it was also a common limitation in previous studies [4,5]. Second, multiple baseline characteristics were different between AF and non-AF groups since the present study was not a prospective trial. However, AF still remained as an independent risk factor of dementia even after the adjustment for these confounders in the multivariate Cox analysis. Even when we only focused on patients without important comorbidities which were different in distribution between AF and non-AF groups, the risk of dementia remained higher among AF patients. Therefore, the increased risk of dementia in AF group may not be fully explained by these baseline differences. Last, although we demonstrated that the risk of dementia was higher in AF patients with a high CHA2DS2-VASc score, it remains unknown whether aggressive managements of underlying diseases can decrease the risk of dementia. Large-scale prospective randomized trials are necessary to study this issue. 5. Conclusion In this nationwide cohort study, we demonstrated that AF was independently associated with a higher risk of dementia. The CHA2DS2-VASc score was useful in predicting dementia in AF patients, and superior to CHADS2 score. Conflict of interest statement No conflict of interest was declared. Acknowledgments This work was supported in part by grants from the National Science Council (NSC98-2410-H-010-003-MY2), and Taipei Veterans General Hospital (V99C1-140, V99A-153, V100D-002-3, V101D-001-2, and V102B-025). References [1] C.T. January, L.S. Wann, J.S. Alpert, et al., AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the Heart Rhythm Society, J. Am. Coll. Cardiol. 64 (2014) 2246–2280. [2] A.J. Camm, G.Y. Lip, R. De Caterina, et al., 2012 focused update of the ESC Guidelines for the management of atrial fibrillation: an update of the 2010 ESC Guidelines for the management of atrial fibrillation. Developed with the special contribution of the European Heart Rhythm Association, Eur. Heart J. 33 (2012) 2719–2747. [3] G.V. Naccarelli, H. Varker, J. Lin, K.L. Schulman, Increasing prevalence of atrial fibrillation and flutter in the United States, Am. J. Cardiol. 104 (2009) 1534–1539. [4] T.J. Bunch, J.P. Weiss, B.G. Crandall, et al., Atrial fibrillation is independently associated with senile, vascular, and Alzheimer's dementia, Heart Rhythm. 7 (2010) 433–437. [5] S. Dublin, M.L. Anderson, S.J. Haneuse, et al., Atrial fibrillation and risk of dementia: a prospective cohort study, J. Am. Geriatr. Soc. 59 (2011) 1369–1375. [6] I. Marzona, M. O'Donnell, K. Teo, et al., Increased risk of cognitive and functional decline in patients with atrial fibrillation: results of the ONTARGET and TRANSCEND studies, CMAJ 184 (2012) E329–E336.
[7] P. Santangeli, L. Di Biase, R. Bai, et al., Atrial fibrillation and the risk of incident dementia: a meta-analysis, Heart Rhythm. 9 (2012) 1761–1768. [8] S. Udompanich, G.Y. Lip, S. Apostolakis, D.A. Lane, Atrial fibrillation as a risk factor for cognitive impairment: a semi-systematic review, QJM 106 (2013) 795–802. [9] R.A. Whitmer, S. Sidney, J. Selby, S.C. Johnston, K. Yaffe, Midlife cardiovascular risk factors and risk of dementia in late life, Neurology 64 (2005) 277–281. [10] FP Lu, KP Lin, HK Kuo, Diabetes and the risk of multi-system aging phenotypes: a systematic review and meta-analysis, PLoS One 4 (2009) e4144. [11] S.I. Sharp, D. Aarsland, S. Day, H. Sonnesyn, C. Ballard, Hypertension is a potential risk factor for vascular dementia: systematic review, Int. J. Geriatr. Psychiatry 26 (2011) 661–669. [12] S.J. Chen, C.J. Liu, T.F. Chao, et al., Dental scaling and atrial fibrillation: a nationwide cohort study, Int. J. Cardiol. 168 (2013) 2300–2303. [13] L.J. Lin, M.H. Cheng, C.H. Lee, D.C. Wung, C.L. Cheng, Y.H. Kao Yang, Compliance with antithrombotic prescribing guidelines for patients with atrial fibrillation—a nationwide descriptive study in Taiwan, Clin. Ther. 30 (2008) 1726–1736. [14] C.H. Chang, Y.C. Lee, C.T. Tsai, et al., Continuation of statin therapy and a decreased risk of atrial fibrillation/flutter in patients with and without chronic kidney disease, Atherosclerosis 232 (2014) 224–230. [15] P.L. Chen, W.J. Lee, W.Z. Sun, Y.J. Oyang, J.L. Fuh, Risk of dementia in patients with insomnia and long-term use of hypnotics: a population-based retrospective cohort study, PLoS One 7 (2012) e49113. [16] C.L. Cheng, Y.H. Kao, S.J. Lin, C.H. Lee, M.L. Lai, Validation of the National Health Insurance Research Database with ischemic stroke cases in Taiwan, Pharmacoepidemiol. Drug Saf. 20 (2011) 236–242. [17] C.C. Lin, M.S. Lai, C.Y. Syu, S.C. Chang, F.Y. Tseng, Accuracy of diabetes diagnosis in health insurance claims data in Taiwan, J. Formos. Med. Assoc. 104 (2005) 157–163. [18] V. de Groot, H. Beckerman, G.J. Lankhorst, L.M. Bouter, How to measure comorbidity. A critical review of available methods, J. Clin. Epidemiol. 56 (2003) 221–229. [19] B.F. Gage, A.D. Waterman, W. Shannon, M. Boechler, M.W. Rich, M.J. Radford, Validation of clinical classification schemes for predicting stroke: results from the National Registry of Atrial Fibrillation, JAMA 285 (2001) 2864–2870. [20] G.Y. Lip, R. Nieuwlaat, R. Pisters, D.A. Lane, H.J. Crijns, 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 137 (2010) 263–272. [21] C.L. Perng, C.C. Shen, L.Y. Hu, et al., Risk of depressive disorder following non-alcoholic cirrhosis: a nationwide population-based study, PLoS One 9 (2014) e88721. [22] P. Forti, F. Maioli, N. Pisacane, E. Rietti, F. Montesi, G. Ravaglia, Atrial fibrillation and risk of dementia in non-demented elderly subjects with and without mild cognitive impairment, Neurol. Res. 28 (2006) 625–629. [23] M.T. Schram, S.M. Euser, A.J. de Craen, et al., Systemic markers of inflammation and cognitive decline in old age, J. Am. Geriatr. Soc. 55 (2007) 708–716. [24] M.A. Crandall, B.D. Horne, J.D. Day, et al., Atrial fibrillation and CHADS2 risk factors are associated with highly sensitive C-reactive protein incrementally and independently, Pacing Clin. Electrophysiol. 32 (2009) 648–652. [25] G. Xu, Z. Zhou, W. Zhu, X. Fan, X. Liu, Plasma C-reactive protein is related to cognitive deterioration and dementia in patients with mild cognitive impairment, J. Neurol. Sci. 284 (2009) 77–80. [26] H. Wersching, T. Duning, H. Lohmann, et al., Serum C-reactive protein is linked to cerebral microstructural integrity and cognitive function, Neurology 74 (2010) 1022–1029. [27] A.E. Roher, C. Esh, T.A. Kokjohn, et al., Circle of willis atherosclerosis is a risk factor for sporadic Alzheimer's disease, Arterioscler. Thromb. Vasc. Biol. 23 (2003) 2055–2062. [28] M. Yarchoan, S.X. Xie, M.A. Kling, et al., Cerebrovascular atherosclerosis correlates with Alzheimer pathology in neurodegenerative dementias, Brain 135 (2012) 3749–3756. [29] J. Heeringa, D.A. van der Kuip, A. Hofman, et al., Subclinical atherosclerosis and risk of atrial fibrillation: the Rotterdam study, Arch. Intern. Med. 167 (2007) 382–387. [30] L.Y. Chen, D.C. Foo, R.C. Wong, et al., Increased carotid intima-media thickness and arterial stiffness are associated with lone atrial fibrillation, Int. J. Cardiol. 168 (2013) 3132–3134. [31] TF Chao, YC Huang, CJ Liu, et al., Acute myocardial infarction in patients with atrial fibrillation with a CHA2DS2-VASc score of 0 or 1: a nationwide cohort study, Heart Rhythm. 11 (2014) 1941–1947. [32] M.D. Ezekowitz, K.E. James, S.M. Nazarian, et al., Silent cerebral infarction in patients with nonrheumatic atrial fibrillation. The veterans affairs stroke prevention in nonrheumatic atrial fibrillation investigators, Circulation 92 (1995) 2178–2182. [33] S. Dublin, M.L. Anderson, S.R. Heckbert, et al., Neuropathologic changes associated with atrial fibrillation in a population-based autopsy cohort, J. Gerontol. A Biol. Sci. Med. Sci. 69 (2014) 609–615. [34] F. Cacciatore, G. Testa, A. Langellotto, et al., Role of ventricular rate response on dementia in cognitively impaired elderly subjects with atrial fibrillation: a 10-year study, Dement. Geriatr. Cogn. Disord. 34 (2012) 143–148.
