] Stroke CerebrooascDis 1994;4:188-193 © 1994 National Stroke Association
Ischemic Stroke in Young Adults: Results from the University of Wisconsin Stroke Registry Mark T. Szmanda, D.O., Douglas A. Dulli, M.D., Ross L. Levine, M.D., and Nancy Bee, B.S.
We prospectively evaluated 128 consecutive young adults aged 18-50 years who suffered from at least one ischemic stroke. Men (92 of 128, 72%) predominated and had a mean age of 41 ± 8 years. Women (36 of 128, 28%) had a mean age of 40 ± 8 years (ns). Risk factors that separated male and female groups included previous stroke, which as seen overall in 34% (43 of 128; M/F = 38/5, P = 0.002), and stroke in the family, which was seen overall in 21% (27 of128; M/F = 22/5, P = 0.005). Thirty-day mortality was seen in 3% (4 of 128), all of whom were men. Stroke causes included atherosclerotic in 22% (28 of 128; M/F = 19/9, ns), cardioembolic in 17% (22 of 128, M/F = 17/5, ns), arteriopathic in 17% (22 of 128, M/F = 11/11, P < O.OO2forfemale preponderance), and coagulopathic in 15% (19 of 128, M/F = 18/1, P = 0.002 for male preponderance). Stroke causes remained "undetermined," including small deep stroke and mixed causes, in 16%( 21 of 128; M/F = 17/4, ns), "uncertain," including migraine-related and mitral valve prolapse, in 9% (11 of 128; M/F = 3/8, P = 0.002 for female preponderance), and "unknown"in 4% (5 of128; M/F = 5/0, ns). These data, as part ofthe University of Wisconsin Stroke Registry, compare favorably to similar, previously published series from otherinstitutions. Composite data from several ofthese series are also included. Key Words: Epidemiology-Ischemic stroke, adult (young) - Thromboembolic stroke.
Although cerebral ischemia certainly occurs in young adults, it appears to be characterized by different etiologic and prognostic features than in older persons. In addition, the cause of ischemic stroke in young adults often remains either undetermined, uncertain, or unknown in up to one-third of the cases (1-7).
Young stroke patients are thought to have both the greatest recovery potential and the greatest likell-
From the Department of Neurology, University of Wisconsin Medical School, and the Neurology Service, Middleton VAH, Madison, WI, U.S.A. Address correspondence to Dr. R L Levine at 2500 Overlook Terrace, Madison, WI 53705, U.S.A. 188
J STROKE CEREBROVASCDIS, VOL. 4, NO.3, 1994
hood of finding a remediable lesion, which often justifies extensive investigation (8). However, our knowledge of ischemic stroke in young adults is based on literature data, which have drawbacks including relatively few patients in each study, differing age limits, as well as a consecutive versus nonconsecutive data collection (7). In many previous studies, evaluations did not specifically study male versus female differences, if any. We prospectively studied demographics, vascular risk factors, causes, and male versus female differences in consecutive ischemic stroke patients aged 18-50 years. Etiologic diagnoses were established using prospective criteria, and data were compared to similar, previously published series from other institutions.
YOUNG ADULT STROKE
Studies and Methods The study population comprised inpatients and outpatients in the Department of Neurology of the University of Wisconsin Hospitals between 1988 and 1993. A total of 128 young adults between 18 and 50 years of age who had at least one ischemic stroke was prospectively evaluated. Patients were seen in consultation by the University of Wisconsin Stroke Service, and data were entered into the University of Wisconsin Stroke Registry. Stroke was defined as a focal neurological deficit of relatively abrupt onset that persisted for greater than 24 h and was associated with the appropriate computer tomographic or magnetic resonance imaging findings. Patients with intracerebral hemorrhage or transient ischemic attacks (TIAs) only were excluded from this study. Stroke localization was not used as an identifying factor in this study except to note when small deep stroke had occurred. Of interest, however, is that none of our patients had ocular stroke. In addition, TIAs as premonitory symptoms were extremely scarce in our patients. All patients underwent careful evaluation for coagulopathies, careful evaluation of cardiac function often including transesophageal echocardiography, and careful evaluation of cerebrovascular status with Doppler/ultrasound and, especially when no other cause was evident, cerebral angiography. Thirty-day mortality was also documented. Stroke risk factors such as smoking, hypertension, previous ischemic stroke, ischemic stroke in the family, diabetes, coronary arterial disease, serum cholesterol greater than 200 mg/L, illicit drug use, history of migraine, and oral contraceptive use were documented. All patient records and radiographs were independently reviewed by two of us, and stroke causes were assigned after this review. Cases were deemed "atherosclerotic" based on ' the results of cerebral angiography and extracranial vascular Doppler/ultrasound in those patients without an evident cardiac source of emboli. Cases were deemed "cardioembolic" based on the results of transthoracic or transesophageal echocardiography plus normal-appearing cerebral vasculature ipsilateral to the clinical stroke. Cases were deemed "arteriopathic" based on the results of angiography, which showed arteritis, dysplasia, dissection, or vasculopathy. These cases often had associated medical or autoimmune disease, trauma, or illicit drug use. Cases were deemed "coagulopathic" based on the results of appropriate laboratory blood tests combined with angiographic and echo cardiographic data that otherwise excluded categorization elsewhere. Cases were deemed "undetermined" based on test results that revealed either
"mixed" etiologic presentation or at least one isolated small deep stroke without any clear etiology. Cases were deemed "uncertain" based on their relationship, either alone or in combination, with migraine, oral contraceptive agents, or mitral valve prolapse. Finally, cases were deemed "unknown" based on the absence of any clearer relationships. Data from this current series were added, in a composite fashion, to selected series from the world literature in order to gain further insight into the pathogenesis of ischemic stroke in young adults. Series were chosen (1,2,4-7) based on their similarity of study population mean age, risk factor definition, dependence on neurodiagnostic studies for clinical diagnoses, and pathogenic categorization. Unfortunately, male versus female data were not readily available from these literature series. Statistical evaluations were performed on male versus female data in this current series. Student's t test, Chi-square analysis, and Fisher's exact test were employed where appropriate.
Results UWStroke Registry There were 92 men and 36 women (male:female ratio, 2.6) in our series of 128 young adults (mean age, 41 years; age range, 18-50) who presented with at least one completed ischemic stroke. Male and female ages were not significantly different (Table 1). All patients were white. Risk factors were as reported in Table 1. Clinical history of previous ischemic stroke (p = 0.002) and history of stroke in the family (p = 0.05) were significantly associated with male sex. Areas of "silent" infarction on neurodiagnostic images were not considered as evidence of past stroke for this current study. Arterial hypertension (p = 0.10) and evidence of coronary artery disease (p = 0.08) approached significance in their association with male sex. In addition, the four patients who died within 30 days of their strokes were all men, and all four died from brain herniation. Only 9% of patients had premonitory TIAs. Causes of ischemic stroke were as reported in Table 2. Strokes were felt to be "atherosclerotic" in 28 of 128 patients (22%),25 of whom had confirmatory anglographic evidence of atherothrombotic occlusive disease. Strokes were felt to be "cardioembolic" in 22 of 128 patients (17%), all of whom had nonocclusive or normal cerebral vasculature ipsila teral to their stroke. Conventional cardiac sources of emboli included sep-
J STROKE CEREBROVASC DIS, VOL. 4, NO.3,
1994
189
M. T. SZMANDA ET AL.
Table 1.
Demographic and risk factor characteristics in 128 youngadults with ischemic stroke No. (all)
Demographics Age (years; mean ± SD) Number in all Risk factors Smoking Hypertension Previous stroke Stroke in family Diabetes Coronary disease Cholesterol >200 mglL Premonitory TIAs 30-day mortality
40.5 128
± 8.0
59 (46%) 45 (35%) 43 (34%) 27 (21%) 22 (17%) 21 (16%) 17 (13%) 11 (9%) 4 (3%)
Male
Female
p'
40.7 ± 7.9 92
40.4 ± 8.4 36
ns
46 37 38 22 16 18 12 6 4
13 8 5 5 6 3 5 5 0
ns 0.10 0.002 0.05 ns
0.08 ns ns ns
'Student's t test, Chi-square, or Fisher's exact test used for male versus female data.
tal defects in 7, acute myocardial infarction in 6, prosthetic valvular disease in 3, and lone atrial fibrillation in 2. Strokes were felt to be "arteriopathic" in 22 of 128 patients (17%), and these included arterial dissection in 10, intracranial arteritis in 5, and fibromuscular dysplasia in 2. Females has significantly more arteriopathy (11 of 36 versus 11 of 92, P < 0.025) than males. This subgroup also included two females with both cocaine abuse and an arteriopathy. Strokes were felt to be "coagulopathic" in 19 of 128 patients (15%), and these included anticardiolipin antibody/lupus anticoagulant in 9, platelet disorders in 5, and polycythemia vera in 2. Males had significantly more coagulopathy (18 of 92 versus 1 of 36, P < 0.002) than females. History of previous ischemic stroke and his-
tory of stroke in the family were evenly distributed across these pathogenic subgroups. Strokes were felt to be of "undetermined" cause in 21 of 128 patients (16%). This included 15 patients with small deep stroke who were not otherwise classifiable as to stroke cause. Males had more small deep strokes (13 of 92 versus 3 of 36, p :::= 0.2) than females. This subgroup also included five patients with a "mixed" stroke cause, three of whom had both an evident cardiac source of emboli and mild-to-moderate extracranial vascular occlusive disease ipsilateral to their stroke. Strokes were felt to be of "uncertain" cause in 11 of 128 patients (9%), including probable migraine-related stroke in 7 (4 women), oral contraceptive agents in 2, and mitral valve prolapse in 2
Table 2. Causes of ischemic stroke in 128 young adults
Atherosclerotic Angiography Doppler only Two or more risk factors Cardioembolic Ateriopathic Coagulopathic Undetermined" Uncertain' Unknown
No. (all)
Male
Female
28 (22%) 25 3 17 22 (17%) 22 (17%) 19 (15%) 21 (16%) 11 (9%) 5 (4%)
19 16 3 11 17 11 18 17 3 5
9 9 6 5 i1 1 4 8 0
ns
ns
Ischemic Stroke in Young Adults: Results from the University of Wisconsin Stroke Registry Mark T. Szmanda, D.O., Douglas A. Dulli, M.D., Ross L. Levine, M.D., and Nancy Bee, B.S.
We prospectively evaluated 128 consecutive young adults aged 18-50 years who suffered from at least one ischemic stroke. Men (92 of 128, 72%) predominated and had a mean age of 41 ± 8 years. Women (36 of 128, 28%) had a mean age of 40 ± 8 years (ns). Risk factors that separated male and female groups included previous stroke, which as seen overall in 34% (43 of 128; M/F = 38/5, P = 0.002), and stroke in the family, which was seen overall in 21% (27 of128; M/F = 22/5, P = 0.005). Thirty-day mortality was seen in 3% (4 of 128), all of whom were men. Stroke causes included atherosclerotic in 22% (28 of 128; M/F = 19/9, ns), cardioembolic in 17% (22 of 128, M/F = 17/5, ns), arteriopathic in 17% (22 of 128, M/F = 11/11, P < O.OO2forfemale preponderance), and coagulopathic in 15% (19 of 128, M/F = 18/1, P = 0.002 for male preponderance). Stroke causes remained "undetermined," including small deep stroke and mixed causes, in 16%( 21 of 128; M/F = 17/4, ns), "uncertain," including migraine-related and mitral valve prolapse, in 9% (11 of 128; M/F = 3/8, P = 0.002 for female preponderance), and "unknown"in 4% (5 of128; M/F = 5/0, ns). These data, as part ofthe University of Wisconsin Stroke Registry, compare favorably to similar, previously published series from otherinstitutions. Composite data from several ofthese series are also included. Key Words: Epidemiology-Ischemic stroke, adult (young) - Thromboembolic stroke.
Although cerebral ischemia certainly occurs in young adults, it appears to be characterized by different etiologic and prognostic features than in older persons. In addition, the cause of ischemic stroke in young adults often remains either undetermined, uncertain, or unknown in up to one-third of the cases (1-7).
Young stroke patients are thought to have both the greatest recovery potential and the greatest likell-
From the Department of Neurology, University of Wisconsin Medical School, and the Neurology Service, Middleton VAH, Madison, WI, U.S.A. Address correspondence to Dr. R L Levine at 2500 Overlook Terrace, Madison, WI 53705, U.S.A. 188
J STROKE CEREBROVASCDIS, VOL. 4, NO.3, 1994
hood of finding a remediable lesion, which often justifies extensive investigation (8). However, our knowledge of ischemic stroke in young adults is based on literature data, which have drawbacks including relatively few patients in each study, differing age limits, as well as a consecutive versus nonconsecutive data collection (7). In many previous studies, evaluations did not specifically study male versus female differences, if any. We prospectively studied demographics, vascular risk factors, causes, and male versus female differences in consecutive ischemic stroke patients aged 18-50 years. Etiologic diagnoses were established using prospective criteria, and data were compared to similar, previously published series from other institutions.
YOUNG ADULT STROKE
Studies and Methods The study population comprised inpatients and outpatients in the Department of Neurology of the University of Wisconsin Hospitals between 1988 and 1993. A total of 128 young adults between 18 and 50 years of age who had at least one ischemic stroke was prospectively evaluated. Patients were seen in consultation by the University of Wisconsin Stroke Service, and data were entered into the University of Wisconsin Stroke Registry. Stroke was defined as a focal neurological deficit of relatively abrupt onset that persisted for greater than 24 h and was associated with the appropriate computer tomographic or magnetic resonance imaging findings. Patients with intracerebral hemorrhage or transient ischemic attacks (TIAs) only were excluded from this study. Stroke localization was not used as an identifying factor in this study except to note when small deep stroke had occurred. Of interest, however, is that none of our patients had ocular stroke. In addition, TIAs as premonitory symptoms were extremely scarce in our patients. All patients underwent careful evaluation for coagulopathies, careful evaluation of cardiac function often including transesophageal echocardiography, and careful evaluation of cerebrovascular status with Doppler/ultrasound and, especially when no other cause was evident, cerebral angiography. Thirty-day mortality was also documented. Stroke risk factors such as smoking, hypertension, previous ischemic stroke, ischemic stroke in the family, diabetes, coronary arterial disease, serum cholesterol greater than 200 mg/L, illicit drug use, history of migraine, and oral contraceptive use were documented. All patient records and radiographs were independently reviewed by two of us, and stroke causes were assigned after this review. Cases were deemed "atherosclerotic" based on ' the results of cerebral angiography and extracranial vascular Doppler/ultrasound in those patients without an evident cardiac source of emboli. Cases were deemed "cardioembolic" based on the results of transthoracic or transesophageal echocardiography plus normal-appearing cerebral vasculature ipsilateral to the clinical stroke. Cases were deemed "arteriopathic" based on the results of angiography, which showed arteritis, dysplasia, dissection, or vasculopathy. These cases often had associated medical or autoimmune disease, trauma, or illicit drug use. Cases were deemed "coagulopathic" based on the results of appropriate laboratory blood tests combined with angiographic and echo cardiographic data that otherwise excluded categorization elsewhere. Cases were deemed "undetermined" based on test results that revealed either
"mixed" etiologic presentation or at least one isolated small deep stroke without any clear etiology. Cases were deemed "uncertain" based on their relationship, either alone or in combination, with migraine, oral contraceptive agents, or mitral valve prolapse. Finally, cases were deemed "unknown" based on the absence of any clearer relationships. Data from this current series were added, in a composite fashion, to selected series from the world literature in order to gain further insight into the pathogenesis of ischemic stroke in young adults. Series were chosen (1,2,4-7) based on their similarity of study population mean age, risk factor definition, dependence on neurodiagnostic studies for clinical diagnoses, and pathogenic categorization. Unfortunately, male versus female data were not readily available from these literature series. Statistical evaluations were performed on male versus female data in this current series. Student's t test, Chi-square analysis, and Fisher's exact test were employed where appropriate.
Results UWStroke Registry There were 92 men and 36 women (male:female ratio, 2.6) in our series of 128 young adults (mean age, 41 years; age range, 18-50) who presented with at least one completed ischemic stroke. Male and female ages were not significantly different (Table 1). All patients were white. Risk factors were as reported in Table 1. Clinical history of previous ischemic stroke (p = 0.002) and history of stroke in the family (p = 0.05) were significantly associated with male sex. Areas of "silent" infarction on neurodiagnostic images were not considered as evidence of past stroke for this current study. Arterial hypertension (p = 0.10) and evidence of coronary artery disease (p = 0.08) approached significance in their association with male sex. In addition, the four patients who died within 30 days of their strokes were all men, and all four died from brain herniation. Only 9% of patients had premonitory TIAs. Causes of ischemic stroke were as reported in Table 2. Strokes were felt to be "atherosclerotic" in 28 of 128 patients (22%),25 of whom had confirmatory anglographic evidence of atherothrombotic occlusive disease. Strokes were felt to be "cardioembolic" in 22 of 128 patients (17%), all of whom had nonocclusive or normal cerebral vasculature ipsila teral to their stroke. Conventional cardiac sources of emboli included sep-
J STROKE CEREBROVASC DIS, VOL. 4, NO.3,
1994
189
M. T. SZMANDA ET AL.
Table 1.
Demographic and risk factor characteristics in 128 youngadults with ischemic stroke No. (all)
Demographics Age (years; mean ± SD) Number in all Risk factors Smoking Hypertension Previous stroke Stroke in family Diabetes Coronary disease Cholesterol >200 mglL Premonitory TIAs 30-day mortality
40.5 128
± 8.0
59 (46%) 45 (35%) 43 (34%) 27 (21%) 22 (17%) 21 (16%) 17 (13%) 11 (9%) 4 (3%)
Male
Female
p'
40.7 ± 7.9 92
40.4 ± 8.4 36
ns
46 37 38 22 16 18 12 6 4
13 8 5 5 6 3 5 5 0
ns 0.10 0.002 0.05 ns
0.08 ns ns ns
'Student's t test, Chi-square, or Fisher's exact test used for male versus female data.
tal defects in 7, acute myocardial infarction in 6, prosthetic valvular disease in 3, and lone atrial fibrillation in 2. Strokes were felt to be "arteriopathic" in 22 of 128 patients (17%), and these included arterial dissection in 10, intracranial arteritis in 5, and fibromuscular dysplasia in 2. Females has significantly more arteriopathy (11 of 36 versus 11 of 92, P < 0.025) than males. This subgroup also included two females with both cocaine abuse and an arteriopathy. Strokes were felt to be "coagulopathic" in 19 of 128 patients (15%), and these included anticardiolipin antibody/lupus anticoagulant in 9, platelet disorders in 5, and polycythemia vera in 2. Males had significantly more coagulopathy (18 of 92 versus 1 of 36, P < 0.002) than females. History of previous ischemic stroke and his-
tory of stroke in the family were evenly distributed across these pathogenic subgroups. Strokes were felt to be of "undetermined" cause in 21 of 128 patients (16%). This included 15 patients with small deep stroke who were not otherwise classifiable as to stroke cause. Males had more small deep strokes (13 of 92 versus 3 of 36, p :::= 0.2) than females. This subgroup also included five patients with a "mixed" stroke cause, three of whom had both an evident cardiac source of emboli and mild-to-moderate extracranial vascular occlusive disease ipsilateral to their stroke. Strokes were felt to be of "uncertain" cause in 11 of 128 patients (9%), including probable migraine-related stroke in 7 (4 women), oral contraceptive agents in 2, and mitral valve prolapse in 2
Table 2. Causes of ischemic stroke in 128 young adults
Atherosclerotic Angiography Doppler only Two or more risk factors Cardioembolic Ateriopathic Coagulopathic Undetermined" Uncertain' Unknown
No. (all)
Male
Female
28 (22%) 25 3 17 22 (17%) 22 (17%) 19 (15%) 21 (16%) 11 (9%) 5 (4%)
19 16 3 11 17 11 18 17 3 5
9 9 6 5 i1 1 4 8 0
ns
ns
