Increased stroke risk among adult
asthmatic patients
Wei-Sheng Chung
*,†, Cheng-Li Lin
‡,§, Yung-Fu Chen
†, Feng-Ming Ho
¶, Wu-Huei Hsu
**and Chia-Hung Kao
**,††Introduction
Stroke is referred to as a cerebrovascular accident (CVA) and can be classified into ischaemic stroke and haemorrhagic stroke. The burden of stroke increased considerably based on the absolute number of ischaemic and haemorrhagic stoke events that occurred worldwide in the past decades [1]. Stroke causes death and impairs gross physical function. Stroke is the leading cause of adult disability and the third leading cause of death in Taiwan [2,3].
Asthma is a common chronic inflammatory disease of the airways, which is characterized by reversible airway obstruction and bronchospasm. An estimated 300 million people of all ages and all racial backgrounds suffer from asthma, which is increasingly affecting healthcare systems, families and patients worldwide [4]. The purpose of current asthma treatments, whether preventative or medicinal, is to control rather than cure the disease.
Inflammatory processes are crucial in the pathophysiology of atherosclerotic disease and hypertension [5–7]. In recent studies, elevated concentrations of thrombin
have been detected in the sputum of asthmatic patients [8,9]. Additional studies have reported that asthma is associated with prothrombotic factors and endothelial dysfunction in the development of atherothrombosis [10–12]. In addition, several epidemiological studies have reported that asthma is associated with the risk of pulmonary thromboembolism and cardiovascular diseases [13–15]. However, asthma is not generally considered as a risk factor for stroke, although asthmatic patients
may exhibit an elevated concentration of coagulation factors [8,11]. Therefore, we investigated whether asthma increases the
risk of stroke using a retrospective nationwide populationbased cohort study.
Methods
Data source
A single-payer compulsory National Health Insurance (NHI) programme was implemented in Taiwan in 1995, covering nearly 99% of 23_74 million residents in Taiwan by the end of 2009 [16]. The National Health Research Institute (NHRI), Ministry of Health and Welfare, established and manages the National Health Insurance Research Database (NHIRD). To protect patient privacy, the NHRI had encrypted in advance all personal identification numbers into unique numbers before releasing the data files to users. With approval from the NHRI, we were able to use the scrambled patient identification to link files, including the registry of medical facilities,
details of inpatient orders, ambulatory cares, dental services and prescriptions. In this study, we obtained a subdata set for one million people randomly selected from the entire
registered population. This subdata set is similar to the entire population in the distributions of sex, age and average payroll-related insurance payment. The disease criteria were
defined according to the International Classifications of Diseases, 9th Revision, Clinical Modification (ICD-9-CM). To
ensure the accuracy of the claims data, the Bureau of National Health Insurance (BNHI) would invite experts to review
claims on a random sample basis for every one hundred outpatient claims and of every fifteen inpatient claims. False claims or unnecessary treatment would result in severe penalty from BNHI. This study was exempted from full ethical
review by China Medical University (IRB Permit Number: CMU-REC-101-012).
Sampled patients
Figure 1 shows the selection process of the participants in the two study cohorts. We enrolled patients who were 18 years of age and older with newly diagnosed asthma (ICD-9 code 493) based on claims data between 1 January 2000 and 31 December 2010. To ensure the accuracy of the asthma diagnosis, we selected patients who had received treatment involving inhaled
corticosteroids, inhaled short-acting beta-2 agonists (SABA) or inhaled long-acting beta-2 agonists (LABA) as the asthmatic cohort. The date of diagnosis of asthma was defined as the index date for initiating follow-up person-years measurement. For each asthmatic patient, four reference individuals were selected from the people without a diagnosis of asthma as the nonasthmatic comparison cohort and were frequency matched according to sex, age (every 5-years span) and index year of diagnosed asthma. We excluded the patients with a history of stroke (ICD-9 codes 430-438) and those with incomplete age or sex information at the baseline. The baseline comorbidity history was determined for each patient; these comorbidities
included atrial fibrillation (ICD-9 code 427.31), hypertension (ICD-9 codes 401-405), diabetes (ICD-9 code 250), hyperlipidemia (ICD-9 code 272), heart failure (ICD-9 code 428), alcoholism (ICD-9 codes 303, 305.0, A215 and V113), obesity (ICD-9 code 278), chronic obstructive pulmonary disease (COPD; ICD-9 codes 4ICD-91, 4ICD-92, 4ICD-96), deep vein thrombosis (DVT; ICD-ICD-9 code 453.8) and coronary artery disease (CAD; ICD-9 codes 413, 414.0, 414.3, 414.8, 414.9).
Main outcome
The primary outcome was newly diagnosed stroke reported in hospitalization records, including haemorrhagic stroke (ICD-9 codes 430-432) and ischaemic stroke (ICD-9 codes 433-438). The follow-up person-years were determined by calculating the interval between the index date and the date on which any of the following occurred first: the date of stroke diagnosis, the date of withdrawal from the NHI programme, the date of death or 31 December 2011.
Statistical analysis
The proportionate distributions of demographic status and comorbidity between the asthmatic and nonasthmatic cohorts were compared and examined using the chi-squared test. The mean ages and standard deviations (SDs) were measured and examined using a t-test. Similarly, the incidence density rates based on demographic variables and comorbidity were calculated for each cohort and the subgroups. Univariable and
multivariable Cox proportional hazard regression models were used to estimate the hazard ratio (HR) and 95% confidence
interval (CI) to assess the effects of asthma on the risk of stroke. The multivariable models were simultaneously adjusted for demographic characteristics and comorbidities. Further analysis was performed to assess the dose response of asthmatic exacerbation on the risk of stroke according to the number of emergency room (ER) visits and hospitalizations for asthma. We divided the asthmatic cohort into three mutually exclusive groups, according to drug treatment status, as follows: (i) treatment with inhaled corticosteroids, (ii) treatment with inhaled
SABA or LABA and (iii) both inhaled corticosteroids and inhaled SABA or LABA. To estimate the cumulative incidence of stroke risk in the asthmatic and nonasthmatic cohorts, we performed survival analysis using the Kaplan–Meier method, and significance was determined using the log-rank test. The entire matching procedure and all statistical analyses were performed using
SAS 9.2 software (SAS Institute, Cary, NC, USA). A two-tailed
P-value of < 0.05 was considered statistically significant.
Results
Demographic characteristics and comorbidity
between the asthmatic patients and nonasthmatic
controls
We analysed the data collected from a total of 58 069 people in the nonasthmatic cohort and 14 518 patients in the asthmatic cohort. The average follow-up duration was 7_21 years for the asthmatic cohort and 7_35 years for the nonasthmatic cohort. In the asthmatic cohort, 54_3% of the patients were female. The mean age of the study patients was 52_6 _ 17_6 years for the asthmatic cohort and 52_0 _ 17_6 years for the nonasthmatic cohort. Compared with those of the nonasthmatic cohort, the comorbidities of atrial fibrillation (0_97% vs. 0_43%), hypertension (38_1% vs. 26_9%), hyperlipidemia (20_4% vs. 15_1%), heart failure (4_90% vs. 1_33%), alcoholism (1_10% vs. 0_56%), obesity (1_60% vs. 0_82%), COPD (28_9% vs. 5_12%), DVT (0_14% vs. 0_11%) and CAD (21_1% vs. 11_9%) were more prevalent in the asthmatic cohort (Table 1).
Comparison of the incidence and hazard ratio of
stroke stratified by sex, age and comorbidity
between the asthmatic patients and nonasthmatic
controls
The overall incidence of stroke was greater in the asthmatic cohort than in the nonasthmatic cohort (10_9 and 7_13 per 1000
person-years, crude HR = 1_53, 95% CI = 1_47–1_60), with an adjusted HR of 1_37 (95% CI = 1_27–1_48) when adjusting for
age, sex and comorbidities (Table 2). Furthermore, patients with asthma were 1_38-fold more likely to develop ischaemic stroke (95% CI = 1_27–1_49) and were 1_31-fold more likely to develop
haemorrhagic stroke (95% CI = 1_09–1_65) than were the nonasthmatic controls after adjusting for age, sex and comorbidities.
The stroke incidence was greater in men than in women in both cohorts. However, the adjusted HR of stroke for the asthmatic cohort compared with that of the nonasthmatic cohort was greater for women than for men. The stroke incidence increased
with age in both cohorts. However, the age-specific asthma-tononasthma relative risk was the greatest for the oldest group
aged 75+ years (adjusted HR = 1_48; 95% CI = 1_29–1_70). The comorbidity-specific asthmatic-cohort-to-nonasthmatic-cohort adjusted HR of stroke was significant for patients both without comorbidity (adjusted HR = 1_37, 95% CI = 1_13–1_67) and with comorbidity (adjusted HR = 1_33, 95% CI = 1_23–1_43).
Hazard ratios of stroke associated with the number
of annual asthmatic exacerbations
Compared with the nonasthmatic cohort, patients in the asthmatic cohort visiting the ER visits more than three times were
associated with a significantly higher risk of stroke (adjusted HR = 3_05, 95% CI = 2_75–3_38) (Table 3). Furthermore, the risk of stroke increased as the number of hospitalizations for asthma increased. The patients who were admitted for asthma more than five times exhibited the highest risk of stroke with an adjusted HR of 60_8 (95% CI = 42_8–86_3).
Hazard ratios of stroke associated with asthma
treatment
Table 4 shows the analysis of treatment associated with stroke risk among the asthmatic patients. Compared with asthmatic patients who received inhaled corticosteroid treatment, the patients who received inhaled SABA or LABA treatment had a significantly increased risk of stroke (adjusted HR = 1_93, 95%
CI = 1_63–2_27), followed by those who had received both inhaled corticosteroids and inhaled SABA or LABA treatment (adjusted HR = 1_33, 95% CI = 1_13–1_56).
Comparing the probabilities of stroke between the
asthmatic and nonasthmatic cohorts using Kaplan–
Meier analysis
The Kaplan–Meier analysis revealed that the risk of stroke increased during the follow-up period in both cohorts, and the cumulative incidences of stroke were significantly higher in the asthmatic cohort than in the nonasthmatic cohort (log-rank P < 0.0001) (Fig. 2).
Discussion
This is the first study to investigate whether asthma increases the risk of developing stroke by applying a longitudinal population-based cohort study to an Asian population. This study
demonstrated that asthmatic patients exhibit a 1_38-fold greater risk of stroke development than that of the general population, after adjusting for age, sex and comorbidities. Onufrak et al.
[15] demonstrated that women with adult-onset asthma experienced a twofold increase in the risk of stroke in four communities
in the United States. Iribarren et al. [17] indicated that
asthma is associated with a 1_2-fold hazard of cerebrovascular disease for adults in Northern Carolina. These two studies reported the stroke risk exhibited by asthmatic patients in local
communities in the United States. Although the asthmatic patients in this study presented a
higher prevalence of comorbidities and coexisting conditions associated with the development of stroke than did the comparison cohort, asthma remained an independent risk factor for
developing stroke after adjusting for covariates.
The mechanisms underlying the epidemiologic association between asthma and stroke remain unclear. A possible theory is that chronic airway inflammation may contribute to systemic inflammation and to the vulnerability to vascular disease [18,19]. Allergens are responsible for an inflammatory response in the lungs of asthmatic patients, in whom the response is aggravated by the proinflammatory effects of platelets and decreased cytoprotective effects [11]. Coagulation is activated
in the airways of asthmatic patients by the leaking of clotting and tissue factors. Fibrin deposition is further facilitated by
decreased activity of the anticoagulant protein C and the inhibition of fibrinolysis [11]. In addition to chronic inflammation,
other contributing thrombovascular risk factors and effects of asthma medications may also play vital roles in stroke development. Vijayakumar et al. [20] indicated that asthma is associated
with increased arterial inflammation. Furthermore,
Onufrak et al. reported that adult-onset asthma is associated with an increased risk of carotid atherosclerosis and stroke among women [12,15].
In this study, more than half of the patients in the asthmatic cohort were women. The finding of this study is consistent with that described in a previous report, indicating that women have a higher prevalence rate than that of men [21]. Onufrak et al. [15] indicated that adult-onset asthma may be a significant risk
factor for stroke in women, but not in men. However, the current study demonstrated that the asthmatic patients exhibited
a significantly higher risk of stroke than that of the comparison cohort in patients of both sexes, after adjusting for
covariates. A possible reason for the stroke risk differences among asthmatic patients between Western countries and Taiwan may be related to racial differences [22].
The stroke incidence increased with age in both cohorts. The age-specific crude relative risk indicated that stroke risk was significantly higher in the asthmatic cohort than in the nonasthmatic cohort. However, the risk of stroke was significantly higher in the asthmatic cohort than in the nonasthmatic cohort for those aged 50 years or older, after adjusting for sex and comorbidities. As people age, activity levels decrease and cardiopulmonary systems deteriorate. Previous studies have reported that cerebral blood flow
decreases during normal ageing [23,24]. Asthma may exacerbate cerebral vascular thrombosis and increase the vulnerability of the cerebral vessels, thereby contributing to stroke risk in elderly patients. The asthmatic patients exhibited an increased adjusted HR of stroke compared with the comparison cohort, irrespective of whether they had any
comorbidity. The Global Initiative for Asthma recommended that asthma care be based on the clinical control status of uncontrolled,
partly controlled and adequately controlled, instead of asthma severity [25]. No study has investigated the relationship
between the number of asthmatic exacerbations and hospitalizations and the risk of stroke development. The present study
demonstrated that the risk of developing stroke increased when the number of asthmatic exacerbations required for emergency room visits was > 3 times a year, and when the number of hospitalizations increased. This result suggests that poor control is a critical factor for stroke in asthmatic patients. The
mechanism by which the number of asthmatic exacerbations, emergency room visits and hospitalizations may predispose stroke seems complex. Inflammation may alter the balance between procoagulant and fibrinolytic activities because inflammation and coagulation stimulate each other [26]. Moreover, the relationship between the number of asthmatic exacerbations necessary for hospitalization and the risk of stroke development exhibited a dose–response effect. Considering the effect of asthma treatment on the risk of stroke development, the patients receiving SABA or LABA exhibited a significantly higher risk of stroke compared with the patients receiving only inhaled corticosteroids, after adjusting for covariates. Although beta-2 agonists are typically inhaled with low systemic absorption, increased plasma levels have been reported [27]. Beta-2 agonists are considered as a significant risk factor for incident atrial fibrillation [28–30].
Subclinical atrial tachyarrhythmias were associated with a significantly increased risk of ischaemic stroke [31].
The strength of this study is that it provides a nationwide population-based longitudinal cohort study on the risk of stroke development in people with asthma. These findings can be generalized to the general population. However, several limitations must be considered when interpreting these findings. The NHIRD does not provide detailed lifestyle information, such as smoking, body mass index, physical activity
levels, socioeconomic status and family history, which are all potential confounding factors for this study. Second, the study
cases were physician-diagnosed asthmatic patients, thus potentially causing a misclassification bias despite the ability of the auditing mechanism of the Bureau of National Health
Insurance to facilitate the minimization of diagnostic uncertainty and misclassification. The lack of drug data, such as
those on hormone replacement therapy and the use of contraceptive, anticoagulant, and antiplatelet drugs, may also have
influenced the primary outcomes of this study.
This nationwide study of 14 518 asthmatic patients with 104 697 person-year follow-ups indicates that asthmatic
patients have a 38% increased risk of developing stroke compared with the general population. These findings emphasize
the importance of clinician awareness of potential stroke development among asthmatic patients.
