ORIGINAL ARTICLE
Increased risk of ischaemic stroke among patients with multiple sclerosis C.-H. Tsenga,b, W.-S. Huanga,b, C.-L. Linc and Y.-J. Changd a
Department of Neurology, China Medical University Hospital, Taichung; bSchool of Medicine, China Medical University College of Medicine, Taichung; cManagement Office for Health Data, China Medical University Hospital, Taichung; and dDepartment of Health Promotion and Health Education, National Taiwan Normal University, Taipei, Taiwan
EUROPEAN JOURNAL OF NEUROLOGY
Keywords:
autoimmune disease, ischaemic stroke, multiple sclerosis Received 19 April 2014 Accepted 19 September 2014 European Journal of Neurology 2015, 22: 500–506 doi:10.1111/ene.12598
Background and purpose: Inflammatory processes including autoimmune diseases which ignite endothelial dysfunction and atherosclerosis may promote development of cardiovascular diseases including ischaemic stroke. This study aimed to evaluate whether multiple sclerosis (MS) increases stroke risk. Methods: A national insurance claim data set of 22 million enrollees in Taiwan was used to identify 1174 patients with MS and 4696 randomly selected age- and gender-matched controls from 1 January 1997 to 31 December 2010. Both cohorts were followed up until the occurrence of stroke or censor. Relevant covariates, such as age, gender, hypertension, diabetes, hyperlipidaemia, coronary artery disease, congestive heart failure and pregnancy, were included for further survey. The hazard ratio (HR) of stroke was assessed using a Cox proportional hazards regression model. Results: After adjusting for the relevant covariates, the MS cohort had an increased risk of stroke (adjusted HR = 12.1 for 1 year; adjusted HR = 4.69 for 2–5 years) compared with the control cohort within 5 years of follow-up. Amongst participants without comorbidities, the MS cohort was still at a greater stroke risk than the control cohort [HR 4.93, 95% confidence interval (CI) 2.85–8.55]. Moreover, in the population aged ≤40, MS was associated with a significantly increased risk of stroke (HR 12.7, 95% CI 3.44–46.7). Conclusions: Multiple sclerosis is declared to be associated with an increased risk in developing stroke, which requires closer attention to this group of patients for stroke prevention, especially in the younger population.
Introduction Stroke, especially ischaemic stroke, is one of the leading causes of death and adult disability in the world [1–4]. For stroke prevention which could reduce stroke-related healthcare and economic burdens, early identification of stroke risk factors is the major priority. However, despite extensive exploration, in about 25%–40% of stroke patients there are still no manifest risk factors identified [5,6], especially in the young population [7,8]. Besides the well-established stroke Correspondence: Y.-J. Chang, Department of Health Promotion and Health Education, National Taiwan Normal University, 162 Heping East Road Section 1, Taipei 10610, Taiwan (tel.: +886 4 2205 2121; fax: +886 4 2234 4055; e-mail: [email protected]). Drs Tseng and Huang contributed equally to this work.
500
risk factors including hypertension [1–4,8], diabetes [1–4,7], hyperlipidaemia [1–4,7], coronary artery disease (CAD) [3,4,7], congestive heart failure (CHF) [3,4] and pregnancy [9,10], it is necessary to investigate unfamiliar risk factors more comprehensively. Multiple sclerosis (MS), a disorder often seen in young females [11,12], is generally believed to be an immune-mediated chronic inflammatory disease that could damage the fatty myelin sheaths around the axons of the central nervous system, including brain and spinal cord, and lead to demyelination, remyelination, axonal loss, gliosis and neurodegeneration of the central nervous system [13,14]; it could also increase the risk of arterial atherosclerosis [15,16]. Such a condition of chronic inflammation caused by the immuneassociated process in MS may last for months to years
© 2014 EAN
MULTIPLE SCLEROSIS AND ISCHAEMIC STROKE
[13,14]. As shown in some chronic inflammatory diseases, such as rheumatoid arthritis [17,18], psoriasis [19,20] and periodontitis [21,22], chronic inflammation has been proved to cause endothelial dysfunction and accelerate atherosclerosis, a clarified cause of the advancement of cardiovascular diseases [23–25] including ischaemic stroke [23,24]. It is not exactly clear how diseases with chronic inflammation enhance the pathogenesis of ischaemic stroke through the possible mechanism of inflammation-related endothelial dysfunction and subsequent atherosclerotic change in the arteries. To the best of our knowledge, only a few recent studies have connected MS with the development of ischaemic stroke [26–28]. However, these studies were all aimed at the relationship between MS and ischaemic stroke in Caucasian people living at higher latitudes, not in the non-Caucasian population dwelling at lower latitudes. In the present population-based study, a National Health Insurance (NHI) claim database in Taiwan was used to investigate the association between MS and ischaemic stroke in a cohort of more than 22 million enrollees for a follow-up period of 14 years from 1 January 1997 to 31 December 2010.
Methods Data source
A population-based retrospective cohort study was designed using the beneficiary files of the NHI inpatient claim database in Taiwan. The Taiwan NHI system consists of over 98% of the entire population in Taiwan. The inpatient claim database used in this study was derived from the NHI Research Databases (NHIRDs) [29–31]. The identification information of patients in the NHI were scrambled prior to release from the NHIRDs; thus, Institutional Review Board (or Ethics Committee) approval was exempt. The residents’ identification was used to link three data files, including the registry of beneficiaries, the catastrophic illness programme (CIP) and inpatients’ claims [29]. The codes of the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9CM) and the International Classification of Diseases, 9th Revision, Procedures (ICD-9-PCS) were utilized for coding the diseases and procedures of interest in the present study. Study subjects
For both inpatients and outpatients to be registered in the CIP, they required the approval of the insurance authority. In the CIP, 1174 patients who were newly
© 2014 EAN
501
diagnosed to have MS (ICD-9-CM code 340) from 1 January 1997 to 31 December 2010 but who had no previous medical history of stroke (ICD-9-CM codes 430–438) before MS diagnosis were identified. The entry date was defined as the date of the diagnosis of MS. The frequency-matched control cohort (n = 4696) was formed by the random selection of age- and gender-matched people without MS and stroke with matched entry date from the NHIRDs. Outcome and relevant variables
Ischaemic stroke (ICD-9-CM codes 433–437) occurring during the study period (1997–2010) was the endpoint in this study. Relevant covariates included age, gender and baseline comorbidities diagnosed before the index entry date, including hypertension (ICD-9CM codes 401–405), diabetes (ICD-9-CM codes 250), hyperlipidaemia (ICD-9-CM codes 272.0–272.4), CAD (ICD-9-CM codes 410–414), CHF (ICD-9-CM codes 398.91, 404.01, 404.03, 404.10, 404.11, 404.13, 404.91, 404.93, 428) and pregnancy (ICD-9-CM codes 640.x1– 676.x1, 640.x2–676.x2, 650–659 and ICD-9-PCS codes 72–74). Statistical analysis
The chi-squared test and t test were used to estimate the differences in discrete and continuous variables between the MS cohort and the control cohort. Person-years were computed from the entry date to the first date of the occurrence of ischaemic stroke, withdrawal from the insurance programme, death or the end of 2010. The gender-, age- and comorbidityspecific incidence rates (per 1000 person-years) of ischaemic stroke were compared between cohorts. Using the Cox proportional hazards regression model, hazard ratios (HRs) were estimated to compare the risk of developing ischaemic stroke between the two cohorts. To measure the severity of MS, an MS severity index was adopted which was defined as the division of total length of hospital stay due to MS during the follow-up duration by the length of the entire follow-up duration. MS severity was further classified into mild (the first tertile in MS severity), moderate (the second tertile in MS severity) and severe (the third tertile in MS severity) by tertile [30,31]. To assess the validity of stroke risk evaluation with this MS severity index, another method of MS severity stratification with hospitalization due to MS as a time-varying covariate was designed to evaluate the stroke risk in MS patients with or without hospitalization. The risk of stroke was also evaluated with separation of follow-up periods, ≤1 year, 2–5 years and
C.-H. TSENG ET AL.
502
>5 years. The stroke-free rates were plotted using the Kaplan–Meier model and the difference between cohorts was assessed by the log-rank test. A twotailed P value 55 Mean SDa Comorbidityb Diabetes Hypertension Hyperlipidaemia Coronary artery disease Congestive heart failure Pregnancy a
b
No (N = 4696) n (%)
Yes (N = 1174) n (%)
3668 (78.1) 1028 (21.9)
917 (78.1) 257 (21.9)
0.99
2852 (60.7) 1340 (25.8) 504 (10.7) 37.1 13.9
713 (60.7) 335 (28.5) 126 (10.7) 37.1 13.8
0.99
167 72 22 36
(3.56) (1.53) (0.47) (0.77)
61 74 51 15
(5.20) (6.30) (4.34) (1.28)
P
0.92 0.009
Increased risk of ischaemic stroke among patients with multiple sclerosis C.-H. Tsenga,b, W.-S. Huanga,b, C.-L. Linc and Y.-J. Changd a
Department of Neurology, China Medical University Hospital, Taichung; bSchool of Medicine, China Medical University College of Medicine, Taichung; cManagement Office for Health Data, China Medical University Hospital, Taichung; and dDepartment of Health Promotion and Health Education, National Taiwan Normal University, Taipei, Taiwan
EUROPEAN JOURNAL OF NEUROLOGY
Keywords:
autoimmune disease, ischaemic stroke, multiple sclerosis Received 19 April 2014 Accepted 19 September 2014 European Journal of Neurology 2015, 22: 500–506 doi:10.1111/ene.12598
Background and purpose: Inflammatory processes including autoimmune diseases which ignite endothelial dysfunction and atherosclerosis may promote development of cardiovascular diseases including ischaemic stroke. This study aimed to evaluate whether multiple sclerosis (MS) increases stroke risk. Methods: A national insurance claim data set of 22 million enrollees in Taiwan was used to identify 1174 patients with MS and 4696 randomly selected age- and gender-matched controls from 1 January 1997 to 31 December 2010. Both cohorts were followed up until the occurrence of stroke or censor. Relevant covariates, such as age, gender, hypertension, diabetes, hyperlipidaemia, coronary artery disease, congestive heart failure and pregnancy, were included for further survey. The hazard ratio (HR) of stroke was assessed using a Cox proportional hazards regression model. Results: After adjusting for the relevant covariates, the MS cohort had an increased risk of stroke (adjusted HR = 12.1 for 1 year; adjusted HR = 4.69 for 2–5 years) compared with the control cohort within 5 years of follow-up. Amongst participants without comorbidities, the MS cohort was still at a greater stroke risk than the control cohort [HR 4.93, 95% confidence interval (CI) 2.85–8.55]. Moreover, in the population aged ≤40, MS was associated with a significantly increased risk of stroke (HR 12.7, 95% CI 3.44–46.7). Conclusions: Multiple sclerosis is declared to be associated with an increased risk in developing stroke, which requires closer attention to this group of patients for stroke prevention, especially in the younger population.
Introduction Stroke, especially ischaemic stroke, is one of the leading causes of death and adult disability in the world [1–4]. For stroke prevention which could reduce stroke-related healthcare and economic burdens, early identification of stroke risk factors is the major priority. However, despite extensive exploration, in about 25%–40% of stroke patients there are still no manifest risk factors identified [5,6], especially in the young population [7,8]. Besides the well-established stroke Correspondence: Y.-J. Chang, Department of Health Promotion and Health Education, National Taiwan Normal University, 162 Heping East Road Section 1, Taipei 10610, Taiwan (tel.: +886 4 2205 2121; fax: +886 4 2234 4055; e-mail: [email protected]). Drs Tseng and Huang contributed equally to this work.
500
risk factors including hypertension [1–4,8], diabetes [1–4,7], hyperlipidaemia [1–4,7], coronary artery disease (CAD) [3,4,7], congestive heart failure (CHF) [3,4] and pregnancy [9,10], it is necessary to investigate unfamiliar risk factors more comprehensively. Multiple sclerosis (MS), a disorder often seen in young females [11,12], is generally believed to be an immune-mediated chronic inflammatory disease that could damage the fatty myelin sheaths around the axons of the central nervous system, including brain and spinal cord, and lead to demyelination, remyelination, axonal loss, gliosis and neurodegeneration of the central nervous system [13,14]; it could also increase the risk of arterial atherosclerosis [15,16]. Such a condition of chronic inflammation caused by the immuneassociated process in MS may last for months to years
© 2014 EAN
MULTIPLE SCLEROSIS AND ISCHAEMIC STROKE
[13,14]. As shown in some chronic inflammatory diseases, such as rheumatoid arthritis [17,18], psoriasis [19,20] and periodontitis [21,22], chronic inflammation has been proved to cause endothelial dysfunction and accelerate atherosclerosis, a clarified cause of the advancement of cardiovascular diseases [23–25] including ischaemic stroke [23,24]. It is not exactly clear how diseases with chronic inflammation enhance the pathogenesis of ischaemic stroke through the possible mechanism of inflammation-related endothelial dysfunction and subsequent atherosclerotic change in the arteries. To the best of our knowledge, only a few recent studies have connected MS with the development of ischaemic stroke [26–28]. However, these studies were all aimed at the relationship between MS and ischaemic stroke in Caucasian people living at higher latitudes, not in the non-Caucasian population dwelling at lower latitudes. In the present population-based study, a National Health Insurance (NHI) claim database in Taiwan was used to investigate the association between MS and ischaemic stroke in a cohort of more than 22 million enrollees for a follow-up period of 14 years from 1 January 1997 to 31 December 2010.
Methods Data source
A population-based retrospective cohort study was designed using the beneficiary files of the NHI inpatient claim database in Taiwan. The Taiwan NHI system consists of over 98% of the entire population in Taiwan. The inpatient claim database used in this study was derived from the NHI Research Databases (NHIRDs) [29–31]. The identification information of patients in the NHI were scrambled prior to release from the NHIRDs; thus, Institutional Review Board (or Ethics Committee) approval was exempt. The residents’ identification was used to link three data files, including the registry of beneficiaries, the catastrophic illness programme (CIP) and inpatients’ claims [29]. The codes of the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9CM) and the International Classification of Diseases, 9th Revision, Procedures (ICD-9-PCS) were utilized for coding the diseases and procedures of interest in the present study. Study subjects
For both inpatients and outpatients to be registered in the CIP, they required the approval of the insurance authority. In the CIP, 1174 patients who were newly
© 2014 EAN
501
diagnosed to have MS (ICD-9-CM code 340) from 1 January 1997 to 31 December 2010 but who had no previous medical history of stroke (ICD-9-CM codes 430–438) before MS diagnosis were identified. The entry date was defined as the date of the diagnosis of MS. The frequency-matched control cohort (n = 4696) was formed by the random selection of age- and gender-matched people without MS and stroke with matched entry date from the NHIRDs. Outcome and relevant variables
Ischaemic stroke (ICD-9-CM codes 433–437) occurring during the study period (1997–2010) was the endpoint in this study. Relevant covariates included age, gender and baseline comorbidities diagnosed before the index entry date, including hypertension (ICD-9CM codes 401–405), diabetes (ICD-9-CM codes 250), hyperlipidaemia (ICD-9-CM codes 272.0–272.4), CAD (ICD-9-CM codes 410–414), CHF (ICD-9-CM codes 398.91, 404.01, 404.03, 404.10, 404.11, 404.13, 404.91, 404.93, 428) and pregnancy (ICD-9-CM codes 640.x1– 676.x1, 640.x2–676.x2, 650–659 and ICD-9-PCS codes 72–74). Statistical analysis
The chi-squared test and t test were used to estimate the differences in discrete and continuous variables between the MS cohort and the control cohort. Person-years were computed from the entry date to the first date of the occurrence of ischaemic stroke, withdrawal from the insurance programme, death or the end of 2010. The gender-, age- and comorbidityspecific incidence rates (per 1000 person-years) of ischaemic stroke were compared between cohorts. Using the Cox proportional hazards regression model, hazard ratios (HRs) were estimated to compare the risk of developing ischaemic stroke between the two cohorts. To measure the severity of MS, an MS severity index was adopted which was defined as the division of total length of hospital stay due to MS during the follow-up duration by the length of the entire follow-up duration. MS severity was further classified into mild (the first tertile in MS severity), moderate (the second tertile in MS severity) and severe (the third tertile in MS severity) by tertile [30,31]. To assess the validity of stroke risk evaluation with this MS severity index, another method of MS severity stratification with hospitalization due to MS as a time-varying covariate was designed to evaluate the stroke risk in MS patients with or without hospitalization. The risk of stroke was also evaluated with separation of follow-up periods, ≤1 year, 2–5 years and
C.-H. TSENG ET AL.
502
>5 years. The stroke-free rates were plotted using the Kaplan–Meier model and the difference between cohorts was assessed by the log-rank test. A twotailed P value 55 Mean SDa Comorbidityb Diabetes Hypertension Hyperlipidaemia Coronary artery disease Congestive heart failure Pregnancy a
b
No (N = 4696) n (%)
Yes (N = 1174) n (%)
3668 (78.1) 1028 (21.9)
917 (78.1) 257 (21.9)
0.99
2852 (60.7) 1340 (25.8) 504 (10.7) 37.1 13.9
713 (60.7) 335 (28.5) 126 (10.7) 37.1 13.8
0.99
167 72 22 36
(3.56) (1.53) (0.47) (0.77)
61 74 51 15
(5.20) (6.30) (4.34) (1.28)
P
0.92 0.009
