Summary
Lebanon is a developing country where the prevalence of stroke subtypes and their correlation with risk factors have not been studied, even though stroke management is an ongoing major national healthcare challenge. In a retrospective study conducted in two university hospitals, data were collected on all stroke cases admitted in 2012 and 2013. Ischemic strokes were then classified according to a modified TOAST classification. A total of 254 inpatients (mean age 68.41 years ±13.34, 55.1% males) was included in the study; of these, 15% had had a hemorrhagic stroke and was therefore excluded. Conversely to findings from studies in other Arab and Asian countries, where small vessel disease is the most frequent subtype, our study showed a predominance of large artery atherosclerosis (53.5%) which, in comparison with other subtypes, was found to be associated with dyslipidemia (OR= 3.82, 95% CI= [1.76–8.28]; p=0.001). Cardioembolic stroke and small vessel disease were found to be positively associated with aging and living in Beirut, respectively. Larger studies are needed to explain these findings.
Keywords: Beirut, Lebanon, Risk factors, Stroke subtypes, TOAST classification
Introduction
Stroke is a heterogeneous cerebrovascular disease with a wide variety of phenotypes and underlying etiologies (Amarenco et al., 2009). For this reason, its classification into subtypes is necessary prior to any intervention (Adams et al., 1993). Ischemic stroke is the most common type, accounting for more than 80% of all strokes (Pendlebury et al., 2009). However, its subtyping requires extensive assessment, from the patient’s first evaluation to late prognosis (Kim and Kim, 2014). Moreover, studying the association between cerebrovascular risk factors and ischemic subtypes is crucial for implementing an efficacious preventive strategy (Rahim et al., 2014).
Epidemiological studies have shown a large variation between countries in the prevalence of stroke types and ischemic subtypes (Schulz and Rothwell, 2003; Turin et al., 2010; Kim and Kim, 2014). This variation may be related to differences in lifestyles, associated risk factors and preventive measures; differences in study designs and stroke classification tools may also have contributed to discrepancies (Schulz and Rothwell, 2003; Kim and Kim, 2014). Moreover, studies have shown different associations between ischemic subtypes and stroke risk factors (Ihle-Hansen et al., 2012). However, hypertension and diabetes were more linked to small vessel disease, whereas large artery atherosclerosis was more associated with hyperlipidemia and smoking (Khan et al., 2007; Ueshima et al., 2008).
Lebanon, an upper-middle-income country in the Eastern Mediterranean Region (EMR) (World Health Organization WHO, 2015), suffers from a paucity of data on the prevalence of stroke subtypes and their associations with risk factors, even though stroke was classified as the second leading cause of mortality in 2012 (9.4% of all deaths), after ischemic heart disease (31.1%) (WHO, 2015). In fact, over the past two to three decades, as a result of lifestyle modifications, economic growth and urbanization, this country and the region as a whole have seen an epidemiological transition (Rahim et al., 2014). Life expectancy has risen, from 67 years in 1990 to 80 years in 2013, and diseases related to aging are expected to further increase, notably stroke (WHO, 2015). Behavioral risk factors such as smoking (cigarettes 38.5% and waterpipe 22.4%) (Sibai and Hwalla, 2008), insufficient physical activity (38%) (WHO, 2015) and obesity (28%) (Nasreddine et al., 2012) are alarmingly high among Lebanese adults and available data show a high prevalence among adolescents as well (El-Roueiheb et al., 2008; Chacar and Salameh, 2011). Related metabolic diseases, especially diabetes and hypertension, have their highest prevalence rates in the EMR (13 and 27% respectively) (WHO, 2015).
The current situation in Lebanon warrants urgent epidemiological studies designed to assess stroke prevalence by types and subtypes and the related risk factor associations, thereby making it possible to implement subsequent preventive strategies in an unhealthy aging population. The present study thus analyzed the prevalence of all hemorrhagic stroke and ischemic stroke subtypes in a Lebanese hospital-based inpatient population. Ischemic stroke subtypes were further categorized according to a modified (risk factor-free) TOAST (Trial of Org 10172 in Acute Stroke Treatment) classification (Adams et al., 1993) with the aim of evaluating their correlations with validated stroke risk factors and with sociodemographic characteristics of the sample.
Materials and methods
This study was retrospectively conducted in two university hospitals in Beirut: the Lebanese University Hospital - Geitaoui (LUH) and the Rafic Hariri University Hospital (RHUH). All medical records of inpatients admitted between January 2012 and December 2013, with a discharge diagnosis of stroke, were examined. Sociodemographic characteristics, individual vascular risk factors, medical history, clinical assessment and neuroimaging data of each patient were retrieved for analysis. After consecutive file numbering, patients with hemorrhagic strokes were identified and excluded, whereas ischemic stroke patients were divided into subgroups and analyzed. A stroke was considered ischemic when brain imaging revealed acute infarction or showed no evidence of hemorrhage.
Risk factor recording
Patients were recorded as:
hypertensive if they were reported as taking any antihypertensive medication prior to admission or if they had a persistent systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg (James et al., 2014).
hyperlipidemic if they were reported as taking a statin or any other lipid lowering agent, or were diagnosed hyperlipidemic according to the 2013 American College of Cardiology/American Heart Association report (Stone et al., 2014).
diabetic if they were reported as taking a current medication for diabetes mellitus or had at least two of fasting glucose measurements ≥ 126 mg/dL (7.0 mmol/L) (American Diabetes Association, 2014).
Moreover, any personal history of transient ischemic attack (TIA), stroke, ischemic heart disease (HD), atrial fibrillation (AF), peripheral artery disease (PAD), migraine, and family history of stroke or heart disease was recorded. HD included any history of angina pectoris or myocardial infarction. ECGs and available echocardiograms were checked to confirm medical history. Smoking and alcohol consumption profiles were also recorded, as were levels of physical activity. Current smokers were compared with the group of non-smokers and former smokers, the latter being defined as those who have not smoked for a minimum of 6 months. Daily alcohol consumption was defined as a regular consumption as opposed to occasional, rare or no consumption. Patients doing regular physical activity (daily or weekly, moderate or intense) were considered physically active. Body mass index and metabolic syndrome were not included in the analysis since a large number of medical records lacked waist, weight and height measurements.
Ischemic stroke classification
Ischemic strokes were retrospectively classified according to the modified TOAST classification tool into five etiological categories (Adams et al., 1993):
Large artery atherosclerosis (LA): patients with this subtype would have, as clinical findings, a cortical impairment (e.g. aphasia, neglect, focal motor deficit) or brain stem or cerebellar dysfunction (e.g. dysarthria, dizziness, ataxia, diplopia). Paraclinical evidence of a stenosis > 50% of an appropriate intracranial or extracranial artery is also needed, supported by either a brain CT scan or MRI showing the presence of an infarct >1.5 cm in diameter in the subcortical hemispheric area, in the cerebrum or cerebellum, or in the brain stem. This category is usually linked to atherosclerosis or embolism. Potential sources of cardiogenic embolism should be excluded.
Cardio-embolism (CE): at least one cardiac source of an embolus must be identified for a possible or probable diagnosis (a valvular heart disease, AF, a history of acute myocardial infarction, or infective endocarditis). Clinical and brain imaging findings are similar to those described for the LA subtype. A decreased level of consciousness and neurological symptoms of abrupt onset with maximal severity at onset help distinguish cardiogenic embolism from other mechanisms of cerebral ischemia.
Small vessel disease (SV): this subtype corresponds to small lacunar infarcts; the clinical symptoms are usually those of one of the classical lacunar syndromes (pure motor hemiparesis, pure sensory stroke, sensorimotor stroke, ataxic hemiparesis) without symptoms related to cerebral cortical dysfunction, and they are supported by small infarcts of less than 1.5 cm in diameter on brain MRI or brain CT scan (which occasionally may appear normal). Potential cardiac sources of embolism should be absent, and evaluation of the large extracranial arteries should not demonstrate a stenosis >50% in an ipsilateral artery.
Stroke of other determined etiology (Other): this category includes patients with rare causes of stroke, such as non-atherosclerotic vasculopathies, hypercoagulable states, or hematological disorders.
-
Stroke of undetermined etiology (Undetermined): this category includes:
patients with no evidence of any clear etiology despite an extensive evaluation;
patients with two or more potential causes of stroke with no clear final diagnosis;
patients who developed an ischemic stroke but have a negative or incomplete evaluation.
All the study patients were assessed with a brain CT scan and an ECG on admission. Blood samples were systematically drawn to investigate lipid and glycemic profiles. Moreover, 49.6% of the patients had a cerebral MRI.
Stroke risk factors, notably hypertension, hyperlipidemia, diabetes and smoking were not included in the classification system used (modified risk factor-free TOAST) so as not to introduce bias into the multivariable analysis of associations of stroke subtypes with risk factors (Jackson and Sudlow, 2005). However, AF and HD were considered in the CE classification and were consequently excluded from the regression models.
Statistical analysis
First, the sample baseline characteristics and prevalence of risk factors among stroke patients were described. We then compared distributions of patient characteristics, medical history and vascular risk factors among the ischemic stroke subtypes. Comparison of means was performed using analysis of variance, with Bonferroni correction on post-hoc tests. Non-parametric tests were used in the case of small subgroups (Kruskal-Wallis). Comparison of proportions was performed using Pearson’s Chi-squared test, with further standardized residuals’ two-by-two comparison. Fisher’s exact test was used for small proportions. In the multivariable analysis, each ischemic stroke subtype was compared with the others for associations with baseline characteristics and risk factors. We estimated the odds of the presence of a risk factor in a particular subtype to the odds of its presence in all the other subtypes. We selected variables that were associated with the outcome with a p-value below 0.2 in bivariate analyses. These were then submitted to a binary logistic regression by forward stepwise selection. After ensuring noncollinearity and sampling adequacy by the Hosmer-Lemeshow test, we reported adjusted odds ratios (aORs). A p-value <0.05 was considered significant. Furthermore, we analyzed the characteristics of patients living in Beirut compared with other patients using a logistic regression. Statistical analysis was performed using SPSS Software for Windows version 18.0.
Results
During the years 2012 and 2013, 254 patients were admitted to LUH and RHUH with a final diagnosis of stroke. They were aged between 31 and 94 years, with a mean age of 68.41 years (±13.34). Male sex was slightly predominant (55.1%) and 51.6% of the patients was living in Beirut. On the basis of clinical and imaging examination, 163 (64.2%) patients were affected by an ischemic stroke, whereas 38 (15.0%) had a hemorrhagic stroke and 53 (20.9%) a TIA. 64 (25.2%) patients had suffered a stroke previously (57 ischemic and 7 hemorrhagic) and were excluded from the subtype classification using the modified TOAST criteria. The most prevalent risk factors among all the patients were hypertension (75.2%), absence of physical activity (89.4%), and diabetes mellitus (40.9%). Further details are presented in Table I.
Table I.
Baseline characteristics of patients (n= 254).
| Sociodemographics | N (%) |
|---|---|
| Male sex | 140 (55.1) |
| Mean age, years (SD) | 68.41 (13.34) |
| Residents of Beirut | 131 (51.6) |
| Covered by an insurance | 206 (81.1) |
| Stroke type | |
| TIA | 53 (20.9) |
| Ischemic | 163 (64.2) |
| Hemorrhagic | 38 (15.0) |
| Risk factors | |
| Previous stroke | 64 (25.2) |
| Previous TIA | 13 (5.1) |
| Family stroke history | 9 (3.5) |
| Family heart disease history | 13 (5.1) |
| HD | 71 (28.0) |
| AF | 47 (18.5) |
| PAD | 2 (0.8) |
| Migraine | 2 (0.8) |
| Hypertension | 191 (75.2) |
| Hyperlipidemia | 71 (28.0) |
| Diabetes mellitus | 104 (40.9) |
| Current smoking | 90 (35.4) |
| Daily alcohol consumption | 12 (4.7) |
| No physical activity | 227 (89.4) |
Abbreviations:SD=standard deviation; TIA=transient ischemic attack; HD=ischemic heart disease, AF=atrial fibrillation; PAD=peripheral artery disease.
159 patients (163 ischemic stroke + 53 TIA – 57 recurrent stroke) were thus eligible for subtype classification according to the modified TOAST criteria. They had a mean age of 67.8 years (±13.56) and 84 (52.8%) were males. LA was the most prevalent ischemic stroke subtype (53.5%) followed by CE (16.4%), SV (13.2%), “Undetermined” (15.1%) and “Other” (1.9%). Due to their small number (only 3 patients), strokes of “Other” determined etiology were combined with those of “Undetermined” etiology in the bivariate and multivariable analyses. The prevalence of risk factors in the different TOAST subtypes is presented in Table II. Significant differences between groups were observed for age (patients with CE were older, p=0.004), sex (CE patients were mainly females, unlike other groups, p=0.012), Beirut residency (Beirut residents were highly predominant among SV patients, p=0.005), hypertension (predominant among SV patients, p=0.035), and hyperlipidemia (predominant among LA patients, p=0.005). Relationships between each ischemic stroke subtype and patients’ characteristics and risk factors are displayed in Tables III, IV and V. Hyperlipidemia was strongly associated with the LA subtype versus other subtypes (OR=3.82, 95%CI= [1.76–8.28]; p=0.001) (Table III). CE was negatively associated with male gender (aOR=0.32, 95%CI= [0.12–0.83]; p=0.020) and hyperlipidemia (aOR=0.26, 95%CI= [0.07–0.96]; p=0.043), and positively associated with age (aOR=1.04, 95%CI= [1.00–1.08]; p=0.047) (Table IV). When comparing the SV subtype with the other groups, living in Beirut was strongly associated with the model (OR=3.25, 95%CI= [1.19–8.88]; p=0.021) (Table V). For the last group (ischemic strokes of Other or Undetermined etiology), age (aOR= 0.97, 95%CI= [0.93–0.99]; p=0.034) and Beirut residency (aOR= 0.346, 95% CI= [0.13–0.95]; p=0.040) were significantly and negatively correlated with the model (data not shown in a Table). As shown in Table VI, stroke patients living in Beirut were found to be significantly older (aOR=1.05, 95%CI= [1.02–1.08]; p=0.001) and more hypertensive (aOR=2.22, 95%CI= [0.97–5.09]; p=0.059) than those living outside the capital when adjusting for other characteristics and risk factors.
Table II.
Characteristics of patients classified using the modified TOAST criteria.
| Total n (%) |
LA n (%) |
CE n (%) |
SV n (%) |
Other or Undetermined n (%) |
p-value* | |
|---|---|---|---|---|---|---|
| Patients | 159 (100) | 85 (53.5) | 26 (16.4) | 21 (13.2) | 27 (17.0) | |
| Mean Age (y) (SD) | 67.80(13.56) | 68.35 (13.60) | 73.96 (11.27) | 66.90 (14.43) | 60.81 (12.17) | 0.004 |
|
| ||||||
| ≤ 44 y | 9 (5.7) | 5 (5.9) | 0 | 2 (9.5) | 2 (7.4) | 0.005 |
|
| ||||||
| [45–54] y | 19 (11.9) | 8 (9.4) | 2 (7.7) | 1 (4.8) | 8 (29.6) | |
|
| ||||||
| [55–64] y | 33 (20.8) | 14 (16.5) | 2 (7.7) | 8 (38.1) | 9 (33.3) | |
|
| ||||||
| [65–74] y | 46 (28.9) | 29 (34.1) | 9 (34.6) | 4 (19) | 4 (14.8) | |
|
| ||||||
| [75–84] y | 32 (20.1) | 18 (21.2) | 7 (26.9) | 3 (14.3) | 4 (14.8) | |
|
| ||||||
| ≥85 y | 20 (12.6) | 11 (12.9) | 6 (23.1) | 3 (14.3) | 0 | |
|
| ||||||
| Male sex | 84 (52.8) | 48 (56.5) | 7 (26.9) | 10 (47.6) | 19 (70.4) | 0.012 |
|
| ||||||
| Beirut residents | 75 (47.2) | 39 (45.9) | 15 (57.7) | 15 (71.4) | 6 (22.2) | 0.005 |
|
| ||||||
| Previous TIA | 8 (5.0) | 3 (3.5) | 2 (7.7) | 1 (4.8) | 2 (7.4) | 0.786 |
|
| ||||||
| Family stroke history | 3 (1.9) | 1 (1.2) | 0 | 1 (4.8) | 1 (3.7) | 0.513 |
|
| ||||||
| Family HD history | 7 (4.4) | 2 (2.4) | 2 (7.7) | 1 (4.8) | 2 (7.4) | 0.561 |
|
| ||||||
| Hypertension | 120 (75.5) | 66 (77.6) | 20 (76.9) | 19 (90.5) | 15 (55.6) | 0.035 |
|
| ||||||
| Hyper-lipidemia | 45 (28.3) | 34 (40) | 3 (11.5) | 4 (19) | 4 (14.8) | 0.005 |
|
| ||||||
| Diabetes mellitus | 66 (41.5) | 36 (42.4) | 12 (46.2) | 10 (47.6) | 8 (29.6) | 0.542 |
|
| ||||||
| Current smoking | 62 (39.0) | 36 (42.4) | 4 (15.4) | 7 (33.3) | 15 (55.6) | 0.050 |
|
| ||||||
| Daily alcohol consumption | 8 (5.0) | 6 (7.1) | 0 | 0 | 2 (7.4) | 0.154 |
|
| ||||||
| No physical activity | 136 (85.5) | 72 (84.7) | 24 (92.3) | 17 (81) | 23 (85.2) | 0.680 |
Abbreviations: LA=large artery atherosclerosis; CE=cardio-embolism; SV=small vessel disease; SD=standard deviation; TIA=transient ischemic attack; HD=ischemic heart disease.
Pearson Chi-square for comparison of proportions, Analysis of variance for comparison of means.
Table III.
Associations between large artery atherosclerosis (LA) and potential risk factors (reference: CE, SV, Other and Undetermined).
| Explanatory variable | Bivariate analysis | Multivariable analysis† | ||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p-value | OR | 95% CI | p-value | |
|
|
||||||
| Age (y) | 1.01 | 0.98–1.03 | 0.580 | |||
|
| ||||||
| Male sex | 1.37 | 0.73–2.56 | 0.325 | |||
|
| ||||||
| Beirut residents | 0.90 | 0.48–1.67 | 0.727 | |||
|
| ||||||
| Previous TIA | 0.51 | 0.12–2.19 | 0.361 | |||
|
| ||||||
| Family stroke history | 0.43 | 0.04–4.83 | 0.493 | |||
|
| ||||||
| Family HD history* | 0.33 | 0.06–1.77 | 0.196 | |||
|
| ||||||
| Hypertension | 1.29 | 0.62–2.65 | 0.495 | |||
|
| ||||||
| Hyperlipidemia* | 3.82 | 1.76–8.28 | 0.001 | 3.82 | 1.76–8.28 | 0.001 |
|
| ||||||
| Diabetes mellitus | 1.08 | 0.57–2.03 | 0.817 | |||
|
| ||||||
| Current smoking* | 2.42 | 0.64–9.14 | 0.191 | |||
|
| ||||||
| Daily alcohol consumption | 4.50 | 0.41–49.6 | 0.219 | |||
|
| ||||||
| No physical activity | 0.87 | 0.36–2.11 | 0.750 | |||
Abbreviations: OR=odds ratio; CI=confidence interval; TIA=transient ischemic attack; HD=ischemic heart disease
Forward Stepwise Selection
Variables included in the multivariable analysis
Table IV.
Associations between cardio-embolism (CE) and potential risk factors (reference: LA, SV, Other and Undetermined).
| Explanatory variable | Bivariate analysis | Multivariable analysis† | ||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p-value | aOR | 95% CI | p-value | |
| Age (y)* | 1.05 | 1.01–1.08 | 0.013 | 1.04 | 1.00–1.08 | 0.047 |
| Male sex* | 0.27 | 0.11–0.68 | 0.006 | 0.32 | 0.12–0.83 | 0.020 |
| Beirut residents | 1.66 | 0.71–3.88 | 0.243 | |||
| Previous TIA | 1.76 | 0.34–9.27 | 0.502 | |||
| Family HD history | 2.13 | 0.39–11.6 | 0.381 | |||
| Hypertension | 1.1 | 0.41–2.97 | 0.851 | |||
| Hyperlipidemia* | 0.28 | 0.08–0.99 | 0.049 | 0.26 | 0.07–0.96 | 0.043 |
| Diabetes mellitus | 1.25 | 0.54–2.92 | 0.600 | |||
| Current smoking* | 0.18 | 0.04–0.98 | 0.047 | |||
| No physical activity | 2.25 | 0.49–10.3 | 0.294 | |||
Abbreviations: OR=odds ratio; CI=confidence interval; aOR=adjusted odds ratio; TIA=transient ischemic attack; HD=ischemic heart disease
Forward Stepwise Selection
Variables included in the multivariable analysis
Table V.
Associations between small vessel disease (SV) and potential risk factors (reference: LA, CE, Other and Undetermined).
| Explanatory variable | Bivariate analysis | Multivariable analysis† | ||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p-value | OR | 95% CI | p-value | |
| Age (y) | 0.99 | 0.96–1.03 | 0.746 | |||
| Male sex | 0.79 | 0.31–1.97 | 0.608 | |||
| Beirut residents* | 3.25 | 1.19–8.88 | 0.021 | 3.25 | 1.19–8.88 | 0.021 |
| Previous TIA | 0.94 | 0.11–8.01 | 0.952 | |||
| Family stroke history | 3.40 | 0.30–39.2 | 0.327 | |||
| Family HD history | 1.10 | 0.13–9.62 | 0.931 | |||
| Hypertension* | 3.48 | 0.77–15.7 | 0.104 | |||
| Hyperlipidemia | 0.56 | 0.18–1.76 | 0.318 | |||
| Diabetes mellitus | 1.33 | 0.53–3.35 | 0.543 | |||
| Current smoking* | 0.34 | 0.07–1.59 | 0.170 | |||
| No physical activity | 0.68 | 0.21–2.24 | 0.524 |
Abbreviations: OR=odds ratio; CI=confidence interval; TIA=transient ischemic attack; HD=ischemic heart disease
Forward Stepwise Selection
Variables included in the multivariable analysis
Table VI.
Characteristics of Beirut residents compared to other residents.
| Explanatory variable | Bivariate analysis | Multivariable analysis† | ||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p-value | aOR | 95% CI | p-value | |
| Age* | 1.05 | 1.02–1.08 | 0.000 | 1.05 | 1.02–1.08 | 0.001 |
| Male sex* | 0.56 | 0.30–1.06 | 0.075 | |||
| AF* | 1.95 | 0.85–4.48 | 0.118 | |||
| HD* | 1.70 | 0.85–3.39 | 0.133 | |||
| Hypertension* | 2.91 | 1.33–6.37 | 0.008 | 2.22 | 0.97–5.09 | 0.059 |
| Current smoking* | 0.34 | 0.09–1.28 | 0.110 |
Abbreviations: OR=odds ratio; CI=confidence interval; aOR=adjusted odds ratio; AF=atrial fibrillation; HD=ischemic heart disease
Forward Stepwise Selection
Variables included in the multivariable analysis
Discussion
Consulting the existing literature, the male predominance observed in our patients is also found in most studies (whether hospital or population based) for ischemic as well as for hemorrhagic stroke (Schulz and Rothwell, 2003; Sallam et al., 2009; Ihle-Hansen et al., 2012). An interesting finding concerned the mean age of our sample, which was older than that reported in studies from Arab countries (EMR) (Bahou et al., 2004; Sallam et al., 2009; Al-Asadi and Habib, 2014), but still younger than the mean age in developed countries (Hajat et al., 2001; Ihle-Hansen et al., 2012; Kolominsky-Rabas et al., 2015). Stroke patients in population-based studies are usually shown to have a higher mean age than those in hospital-based studies (Schulz and Rothwell, 2003).
Hypertension, diabetes mellitus, current smoking and physical inactivity were the most prevalent risk factors for stroke among our population. Hypertension has been found to be the most prevalent risk factor among stroke patients in all countries (Turin et al., 2010; Aljefree and Ahmed, 2015) and its prevalence does not usually differ between hospital- and population-based studies.
However, dyslipidemia, found in 28% of our patients, showed higher rates in many other studies: 55% in Norway (Ihle-Hansen et al., 2012), 57% in Kuwait (Ashkanani et al., 2013) and 78% in low-to-middle-income Arab countries (Abboud et al., 2013).
Diabetes mellitus rates are higher in Lebanon, as in other Arab countries, than in developed countries: 19% in Japan (Turin et al., 2010) and 12% in Norway (Ihle-Hansen et al., 2012) versus 56% in Kuwait (Ashkanani et al., 2013) and 46% in the study by Abboud et al. (2013). Population-based studies show lower frequencies of diabetes among stroke patients (Bejot et al., 2008).
The high smoking rate obtained was unsurprising, having been shown in previous studies (Sibai and Hwalla, 2008) and found in many Arab countries (Aljefree and Ahmed, 2015). Lower rates are found in developed countries: 23% in Norway and 26% in Japan (Turin et al., 2010; Ihle-Hansen et al., 2012). Moreover, the frequency of AF (18.5%) among the stroke patients in our sample is high compared with values recorded in other Arab countries (El-Sayed et al., 1999; Hamad et al., 2001; Ashkanani et al., 2013; Al-Asadi and Habib, 2014), but lower than the frequency in Western countries, where the population is older; AF is indeed known to increase with age (Ihle-Hansen et al., 2012).
With regard to the ischemic stroke subtypes defined according to the modified TOAST classification criteria, LA was the most common, accounting for 53.5% of all ischemic strokes, a rate higher than in published data. The high prevalence of smoking may be linked to this finding as might the finding of a higher rate of untreated cases of dyslipidemia, but the association needs further investigation. However, epidemiological findings from French, Korean and some Arab publications also report a higher prevalence of LA, reaching 35.8% (Bejot et al., 2008), 36% (Kim and Kim, 2014) and 38.1% (Deleu et al., 2011) of all ischemic subtypes, respectively. Conversely, the CE subtype is more frequently reported in Western countries (Schulz and Rothwell, 2003; Ueshima et al., 2008; Ihle-Hansen et al., 2012; Kim and Kim, 2014), whereas the SV subtype is reported with high prevalence in Asian and Arab countries (Al-Shammri et al., 2003; Bahou et al., 2004; Khan et al., 2008; Turin et al., 2010).
Furthermore, the present study shows a significant association between ischemic stroke subtypes and specific risk factors. In line with data from the literature, lipid abnormalities were found to be strong predictors of LA and negative predictors of CE (Schulz and Rothwell, 2003; Khan et al., 2007). However, the study by Schulz and Rothwell (2003) showed a lower aOR for the association (aOR=2.10, 95%CI= [1.55–2.83], p<0.0001). Moreover, we found that CE was negatively associated with male gender, a finding also reported by Schulz and Rothwell (aOR=0.7, 95%CI= [0.6–0.9], p=0.01). In line with existing data (Schulz and Rothwell, 2003; Bejot et al., 2008; Turin et al., 2010), we also found that aging was a predictor of CE, as the incidence of AF increases with age. The SV subtype was found to be correlated only with Beirut residency after adjusting for other variables. Hypertension, although higher among SV patients (p=0.035), was not found to be significantly associated with this subtype in the multivariable regression, contrary to the findings of others (Khan et al., 2007). This could be linked to the relatively small sample size in the present study especially given that Beirut residents are found to be more hypertensive than other residents in the multivariable model. Furthermore, a recent study (Gebeily et al., 2014) found hypertension and age to be associated with cerebral white matter hyperintensities found on brain imaging in Lebanese patients, and these hyperintensities are known to be more commonly seen among SV patients (Chen et al., 2009). Thus, the association between Beirut residency SV could be mediated by these two factors: people living in Beirut were older and more hypertensive. From another perspective, some studies have suggested an association between air pollution and SV (Andersen et al., 2010; O’Donnell et al., 2011; Corea et al., 2012). In fact, air pollution may trigger various mechanisms related to the development of SV, such as systemic inflammation, activation of thrombosis pathways, arteriolar narrowing and impaired vasodilation, and accelerated progression of atherosclerosis (Brook, 2008). However, the capital, Beirut, is the most populated city in Lebanon and also highly polluted (Afif et al., 2009; Borgie et al., 2014), and this may further explain the correlation between Beirut residency and the SV subtype in our study. Another hypothesis that should be tested in future studies is the relationship between stroke subtypes and obesity, a potential residual confounding risk factor in the multivariable regressions.
The modified risk factor-free method (modified TOAST criteria) used to classify ischemic stroke cases into subtypes strengthens the validity of our results. However, there are some limitations to be underlined. First, the data collection was conducted retrospectively, which may introduce an information/classification bias. Although the TOAST classification has previously been applied retrospectively with accurate and reproducible results (Goldstein et al., 2001), it would have been preferable to prospectively subtype the ischemic stroke cases. It would also have been more appropriate to have the medical records analyzed separately by two neurologists, then measuring the degree of agreement between the two classifications. Second, an inclusion bias may exist due to the hospital-based nature of the study. In fact, some differences in the prevalence of ischemic stroke subtypes, and their associations with risk factors, have been reported between hospital- and population-based studies (Schulz and Rothwell, 2003). Hospital-based studies may include more severe stroke cases, cases occurring in a younger population, or cases with different frequencies of risk factors, as shown previously (Giroud et al., 1997). Moreover, the fact that both recruiting centers were in Beirut may have biased the results and the prevalence of the LA subtype, which may not be representative of the whole Lebanese population. Thus, larger population-based studies should be conducted in the future to confirm our findings.
This is the first epidemiological study to assess stroke subtypes in Lebanon using the modified TOAST classification. The LA ischemic subtype may be predominant among hospitalized patients, and associated with a high percentage of uncontrolled cases of dyslipidemia. Interestingly, the SV subtype may be more frequent among Beirut residents, an association that may be partly explained by the high level of pollution in the capital. Thus, a larger population-based survey is needed to further confirm these findings and lead towards better risk factor identification and control. The implementation of a comprehensive stroke prevention strategy at national level is, furthermore, urgent in an unhealthy aging population.
Finally, this study was conducted for observational purposes with no traceability of participants. Thus, no ethical approval was found to be necessary by the Lebanese University ethics committee and by the hospitals’ ethics committees.
References
- Abboud H, Labreuche J, Arauz A, et al. Demographics, socio-economic characteristics, and risk factor prevalence in patients with non-cardioembolic ischaemic stroke in low- and middle-income countries: the OPTIC registry. Int J Stroke. 2013;8(Suppl A100):4–13. doi: 10.1111/j.1747-4949.2012.00893.x. [DOI] [PubMed] [Google Scholar]
- Adams HP, Bendixen BH, Kappelle LJ, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke. 1993;24(1):35–41. doi: 10.1161/01.str.24.1.35. [DOI] [PubMed] [Google Scholar]
- Afif C, Dutot AL, Jambert C, et al. Statistical approach for the characterization of NO2 concentrations in Beirut. Air Qual Atmosphere Health. 2009;2(2):57–67. [Google Scholar]
- Al-Asadi JN, Habib HA. Risk factors and 30-day case fatality of first-ever stroke in Basrah, Iraq. Niger Med J J Niger Med Assoc. 2014;55(3):209–213. doi: 10.4103/0300-1652.132041. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Aljefree N, Ahmed F. Prevalence of cardiovascular disease and associated risk factors among adult Population in the Gulf region: a systematic review. Adv Public Health. 2015;27:e235101. [Google Scholar]
- Al-Shammri S, Shahid Z, Ghali A, et al. Risk factors, subtypes and outcome of ischaemic stroke in Kuwait-a hospital-based study. Med Princ Pract Int J Kuwait Univ Health Sci Cent. 2003;12(4):218–223. doi: 10.1159/000072287. [DOI] [PubMed] [Google Scholar]
- Amarenco P, Bogousslavsky J, Caplan LR, et al. Classification of stroke subtypes. Cerebrovasc Dis. 2009;27(5):493–501. doi: 10.1159/000210432. [DOI] [PubMed] [Google Scholar]
- American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care. 2014;37(Suppl 1):S81–90. doi: 10.2337/dc14-S081. [DOI] [PubMed] [Google Scholar]
- Andersen ZJ, Olsen TS, Andersen KK, et al. Association between short-term exposure to ultrafine particles and hospital admissions for stroke in Copenhagen, Denmark. Eur Heart J. 2010;31(16):2034–2040. doi: 10.1093/eurheartj/ehq188. [DOI] [PubMed] [Google Scholar]
- Ashkanani A, Hassan KA, Lamdhade S. Risk factors of stroke patients admitted to a general hospital in Kuwait. Int J Neurosci. 2013;123(2):89–92. doi: 10.3109/00207454.2012.732633. [DOI] [PubMed] [Google Scholar]
- Bahou Y, Hamid H, Hadidi A. Ischaemic stroke in Jordan: a 2-year hospital-based study of subtypes and risk factors. East Mediterr Health J Rev Santé Méditerranée Orient Al-Majallah Al-Ṣiḥḥīyah Li-Sharq Al-Mutawassiṭ. 2004;10(1–2):138–146. [PubMed] [Google Scholar]
- Bejot Y, Caillier M, Ben Salem D, et al. Ischaemic stroke subtypes and associated risk factors: a French population based study. J Neurol Neurosurg Psychiatry. 2008;79(12):1344–1348. doi: 10.1136/jnnp.2008.150318. [DOI] [PubMed] [Google Scholar]
- Borgie M, Garat A, Cazier F, et al. Traffic-related air pollution. A pilot exposure assessment in Beirut, Lebanon. Chemosphere. 2014;96:122–128. doi: 10.1016/j.chemosphere.2013.09.034. [DOI] [PubMed] [Google Scholar]
- Brook RD. Cardiovascular effects of air pollution. Clin Sci (Lond) 2008;115(6):175–187. doi: 10.1042/CS20070444. [DOI] [PubMed] [Google Scholar]
- Chacar HR, Salameh P. Public schools adolescents’ obesity and growth curves in Lebanon. Leban Med J. 2011;59(2):81. [PubMed] [Google Scholar]
- Chen X, Wen W, Anstey KJ, et al. Prevalence, incidence, and risk factors of lacunar infarcts in a community sample. Neurology. 2009;73(4):266–272. doi: 10.1212/WNL.0b013e3181aa52ea. [DOI] [PubMed] [Google Scholar]
- Corea F, Silvestrelli G, Baccarelli A, et al. Airborne pollutants and lacunar stroke: a case crossover analysis on stroke unit admissions. Neurol Int. 2012;4(2):e11. doi: 10.4081/ni.2012.e11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Deleu D, Inshasi J, Akhtar N, et al. Risk factors, management and outcome of subtypes of ischemic stroke: a stroke registry from the Arabian Gulf. J Neurol Sci. 2011;300(1–2):142–147. doi: 10.1016/j.jns.2010.08.023. [DOI] [PubMed] [Google Scholar]
- El-Roueiheb Z, Tamim H, Kanj M, et al. Cigarette and waterpipe smoking among Lebanese adolescents, a cross-sectional study, 2003–2004. Nicotine Tob Res Off J Soc Res Nicotine Tob. 2008;10(2):309–314. doi: 10.1080/14622200701825775. [DOI] [PubMed] [Google Scholar]
- El Sayed MM, Adeuja AO, El-Nahrawy E, et al. Characteristics of stroke in Hofuf, Saudi Arabia. Ann Saudi Med. 1999;19(1):27–31. doi: 10.5144/0256-4947.1999.27. [DOI] [PubMed] [Google Scholar]
- Gebeily S, Fares Y, Kordahi M, et al. Cerebral white matter hyperintensities (WMH): an analysis of cerebrovascular risk factors in Lebanon. Int J Neurosci. 2014;124(11):799–805. doi: 10.3109/00207454.2014.884087. [DOI] [PubMed] [Google Scholar]
- Giroud M, Lemesle M, Quantin C, et al. A hospital-based and a population-based stroke registry yield different results: the experience in Dijon, France. Neuroepidemiology. 1997;16(1):15–21. doi: 10.1159/000109666. [DOI] [PubMed] [Google Scholar]
- Goldstein LB, Jones MR, Matchar DB, et al. Improving the reliability of stroke subgroup classification using the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria. Stroke J Cereb Circ. 2001;32(5):1091–1098. doi: 10.1161/01.str.32.5.1091. [DOI] [PubMed] [Google Scholar]
- Hajat C, Dundas R, Stewart JA, et al. Cerebrovascular risk factors and stroke subtypes: differences between ethnic groups. Stroke. 2001;32(1):37–42. doi: 10.1161/01.str.32.1.37. [DOI] [PubMed] [Google Scholar]
- Hamad A, Hamad A, Sokrab TE, et al. Stroke in Qatar: a one-year, hospital-based study. J Stroke Cerebrovasc Dis Off J Natl Stroke Assoc. 2001;10(5):236–41. doi: 10.1053/jscd.2001.30382. [DOI] [PubMed] [Google Scholar]
- Ihle-Hansen H, Thommessen B, Wyller TB, et al. Risk factors for and incidence of subtypes of ischemic stroke. Funct Neurol. 2012;27(1):35–40. [PMC free article] [PubMed] [Google Scholar]
- Jackson C, Sudlow C. Are lacunar strokes really different? Stroke J Cereb Circ. 2005;36(4):891–901. doi: 10.1161/01.STR.0000157949.34986.30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- James PA, Oparil S, Carter BL, et al. evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8) JAMA. 2014;311(5):507–520. doi: 10.1001/jama.2013.284427. [DOI] [PubMed] [Google Scholar]
- Khan U, Porteous L, Hassan A, et al. Risk factor profile of cerebral small vessel disease and its subtypes. J Neurol Neurosurg Psychiatry. 2007;78(7):702–706. doi: 10.1136/jnnp.2006.103549. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Khan FY, Yasin M, Abu-Khattab M, et al. Stroke in Qatar: a first prospective hospital-based study of acute stroke. J Stroke Cerebrovasc Dis Off J Natl Stroke Assoc. 2008;17(2):69–78. doi: 10.1016/j.jstrokecerebrovasdis.2007.11.004. [DOI] [PubMed] [Google Scholar]
- Kim BJ, Kim JS. Ischemic stroke subtype classification: an Asian viewpoint. J Stroke. 2014;16(1):8. doi: 10.5853/jos.2014.16.1.8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kolominsky-Rabas PL, Wiedmann S, Weingartner M, et al. Time trends in incidence of pathological and etiological stroke subtypes during 16 years: the Erlangen Stroke Project. Neuroepidemiology. 2015;31:24–29. doi: 10.1159/000371353. [DOI] [PubMed] [Google Scholar]
- Nasreddine L, Naja F, Chamieh MC, et al. Trends in overweight and obesity in Lebanon: evidence from two national cross-sectional surveys (1997 and 2009) BMC Public Health. 2012;12:798. doi: 10.1186/1471-2458-12-798. [DOI] [PMC free article] [PubMed] [Google Scholar]
- O’Donnell MJ, Fang J, Mittleman MA, et al. Fine particulate air pollution (PM2.5) and the risk of acute ischemic stroke. Epidemiol Camb Mass. 2011;22(3):422–431. doi: 10.1097/EDE.0b013e3182126580. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pendlebury ST, Giles MT, Rothwell PM. Transient Ischemic Attack and Stroke: Diagnosis, Investigation, and Management. 1st ed. New York: Cambridge University Press; 2009. Epidemiology, risk factors, pathophysiology and causes of transient ischemic attack and stroke [Internet] [cited 2015 May 12]. Available from: http://www.langtoninfo.co.uk/web_content/9780521735124_excerpt.pdf. [Google Scholar]
- Rahim HFA, Sibai A, Khader Y, et al. Non-communicable diseases in the Arab world. Lancet. 2014 Jan 25;383(9914):356–367. doi: 10.1016/S0140-6736(13)62383-1. [DOI] [PubMed] [Google Scholar]
- Sallam A-R, Al-Aghbari K, Awn H. The Clinical Profile of Stroke: A Yemeni Experience. J Med J. 2009;43(2):115–121. [Google Scholar]
- Schulz UGR, Rothwell PM. Differences in vascular risk factors between etiological subtypes of ischemic Stroke: importance of population-based studies. Stroke. 2003;34(8):2050–2059. doi: 10.1161/01.STR.0000079818.08343.8C. [DOI] [PubMed] [Google Scholar]
- Sibai A, Hwalla N. WHO STEPS-Chronic disease risk factor surveillance: data book for Lebanon [Internet] [cited 2015 May 22]. Available from: http://www.who.int/chp/steps/2008_STEPS_Lebanon.pdf.
- Stone NJ, Robinson JG, Lichtenstein AH, et al. 2013 ACC/AHA Guideline on the Treatment of Blood Cholesterol to Reduce Atherosclerotic Cardiovascular Risk in Adults A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. 2014;129(25 suppl 2):S1–45. doi: 10.1161/01.cir.0000437738.63853.7a. [DOI] [PubMed] [Google Scholar]
- Turin TC, Kita Y, Rumana N, et al. Ischemic stroke subtypes in a Japanese population: Takashima Stroke Registry, 1988–2004. Stroke J Cereb Circ. 2010;41(9):1871–1876. doi: 10.1161/STROKEAHA.110.581033. [DOI] [PubMed] [Google Scholar]
- Ueshima H, Sekikawa A, Miura K, et al. Cardiovascular disease and risk factors in Asia: a selected review. Circulation. 2008;118(25):2702–2709. doi: 10.1161/CIRCULATIONAHA.108.790048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- World Health Statistics 2015. World Health Organization; 2015. [Google Scholar]
- WHO | Lebanon [Internet] WHO; [cited 2015 May 26]. Available from: http://www.who.int/countries/lbn/en/ [Google Scholar]