Three hundred and twenty-four subjects who were aged between 16 and 44 years and who were affected with ischemic stroke between January 2000 and December 2005 were included in the study. The mean age at onset was 36.7 years, with 162 women and 162 men. This sample represents 2.7% of the total number of ischemic strokes in patients aged 16 to 80 years who were registered in the study period in the Lombardy region (324/12108) [14
In total, 303 patients (93.5%) were admitted to the neurological ward, and the remaining patients were admitted to other wards (cardiology, internal medicine, neurosurgery, or intensive care units).
No significant differences were observed in the demographic data between the centers with respect to gender or the age of onset distributions.
The baseline characteristics and the risk factor frequencies, stratified by the age groups, are reported in . Among the risk factors with frequencies that were greater than 5% (), one or more were present in 71.5% of the patients (33.3% of the patients had one risk factor, 22.8% of the patients had two risk factors, and 15.4% of the patients had three or more risk factors). No risk factor was present in 28.5% of the patients.
Baseline characteristics and risk factors stratified by age classes.
Multivariate analyses, which were adjusted for hospital and gender, indicated that hypercholesterolemia is more frequent in older patients (Group 1 versus Group 3: OR 0.97, 95%
CI 0.53–1.80, P
= 0.04), obesity is more frequent in older patients (Group 2 versus Group 3: OR 0.22, 95%
CI 0.07–0.75, P
= 0.01), and migraine is more frequent in younger patients (Group 2 versus Group 3: OR 2.30, 95%
CI 1.13–4.70, P
Hypertension was present in 50% of the patients with LVD, in 56.9% of those with SVD, in 38.7% of those with CE, in 36.2% of those with OTH, and in 32.2% of those with UND. Smoking was present in 33.3% of the patients with LVD, in 52.9% of those with SVD, in 8.1% of those with CE, in 18.1% of those with OTH, and in 10.3% of those with UND. Hypercholesterolemia was present in 16.7% of the patients with LVD, in 31.4% of those with SVD, in 21% of those with CE, in 14.9% of those with OTH, and in 19.5% of those with UND. The frequency of hypertension was higher in the LVD than in CE (P = 0.02), OTH (P < 0.001) and UND (P = 0.003) categories. Hypertension was also more prevalent in patients with SVD than in those with CE (P < 0.001) or OTH (P < 0.001). The prevalence of smoking was higher in patients with LVD or SVD; however, the difference was significant only for SVD when compared to OTH (P = 0.02) and UND (P = 0.004). Hypercholesterolemia was present in all of the stroke subtypes, without any significant differences between them.
The time delay from stroke onset to hospital arrival was <3 hours for 107 patients (33.0%), between 3 and 6 hours for 67 patients (20.7%), between 6 and 24 hours for 68 patients (21.0%), and greater than 24 hours for 63 patients (19.4%). The time delay was unknown for 19 patients (5.9%). No significant changes of time delay over 5-year period of study were observed.
For the diagnostic workup, EKGs and brain CT scans were acquired for all of the patients. Brain MRIs were acquired in 237 patients (73.1% of the total). All of the patients underwent a minimum of one extracranial circulation evaluation: an extracranial duplex ultrasonography was performed in 238 patients (73.4%), an extracranial MR angiography was performed in 67 patients (20.7%), and an extracranial CT angiography was performed in 27 patients (8.3%). Intracranial MR angiography was performed in 118 patients (36.4%), intracranial CT angiography was performed in 19 patients (5.9%), and digital subtraction angiography was performed in 102 patients (31.5%). Cardiologic assessments included transthoracic echocardiography (192 patients, 59.3%), transesophageal echocardiography (146 patients, 45.1%), transthoracic echocardiography with contrast injection (76 patients, 23.5%), and transcranial Doppler ultrasonography (32 patients, 9.9%).
The type of diagnostic workup that was performed changed over the study period. When we performed the analysis over the 5-year period, the proportion of the patients who underwent a brain MRI progressively increased from 60% (2000) to 80.6% (2005), P = 0.1. The frequency of the patients who underwent a minimum of one noninvasive angiographic study of the cerebral circulation (extracranial and intracranial MR angiography, extracranial and intracranial CT angiography) increased: 26.7% (2000), 37.2% (2001), 50% (2002), 39.2% (2003), 50.7% (2004), 69.4% (2005), P < 0.001. A reduction in the use of digital angiography was observed: 53.3% (2000), 27.9% (2001), 42% (2002), 21.6% (2003), 26.7% (2004), 30.6% (2005), P = 0.03, . The proportion of the patients who underwent a minimum of one study to detect a PFO increased from 56.7% (2000) to 66.7% (2005), P = 0.08.
Changes over time (2000–2005) of the frequency of angiographic studies.
Coagulation testing (i.e., testing for the presence of antiphospholipid antibodies, protein C, protein S, antithrombin III, factor V Leiden, and prothrombin gene analyses and homocysteine plasma levels) was performed in 271 patients (83.6% of the patients). The findings of these tests were abnormal in at least one test in 84 patients (30.9%). In 19/324 patients (5.9%), thrombophilia was recognized as the underlying cause of stroke: antiphospholipid syndrome was present in ten patients, mutations in the factor V Leiden or prothrombin genes were detected in four patients, a deficiency of protein S was observed in one patient, and hyperhomocysteinemia (persistently elevated plasma levels >20μ
mol/L measured at admission and after three months) was observed in four patients.
With respect to the stroke type classification, LACI was diagnosed in 44 patients (13.6%), TACI was diagnosed in 35 patients (10.8%), PACI was diagnosed in 130 patients (40.1%), and POCI was diagnosed in 115 (35.5%). No significant changes over the study period were observed in the relative percentages of these diagnoses (P = 0.08).
Multivariate analyses, which were adjusted for age and center, indicated that the patients who were aged between 16 and 29 years (Group 1) exhibited a lower frequency of LACI classification than did older patients (40–44 years of age: OR = 0.42 (95%
CI = 0.19–0.91), P
The stroke etiology was classified according to the TOAST criteria: LVD in 30 patients (9.3%), SVD in 51 patients (15.7%), CE in 62 patients (19.1%), UND in 87 (26.9%), and OTH in the remaining 94 patients (29%).
In the cardioembolic group, high-risk factors were identified in 17 patients (valvular heart disease in 5 patients, dilated cardiomyopathy in 5 patients, atrial fibrillation in 4 patients, myocardial infarction in 2 patients, and congenital cardiac malformation in 1 patient). Low- or uncertain-risk sources were identified in 45 patients (PFO in 32 patients and PFO with atrial septal aneurysm in 13 patients).
The subtypes of uncertain diagnoses included an incomplete evaluation in 23 patients, multiple possible etiologies in 13 patients, and a negative evaluation in 51 patients.
The category of ischemic stroke as a result of other causes included the following items: arterial dissection (47), thrombophilia (19), systemic lupus erythematosus (7), inflammatory bowel disease (1), drug abuse (5), migrainous infarct (5), vascular malformations (5), hematological disease (2), Sneddon's syndrome (2), and CADASIL (1).
The stroke etiology was stratified by age, as illustrated in . According to the multivariate model, the risk of exhibiting a LVD or SVD etiology versus all of the other causes (i.e., CE, UND and OTH) was significantly lower when the patient was under 30 years of age at onset (P
= 0.02; OR = 0.21 (95%
CI = 0.07–0.56)) when compared to older patients.
TOAST criteria stratified by age classes.
Only UND and CE showed changes over time. The proportion of undetermined cases decreased as follows: 60% (2000), 30.2% (2001), 22% (2002), 31.4% (2003), 18.7% (2004), 20.8% (2005), P < 0.001. Cardioembolic strokes increased as follows: 3.3% (2000), 18.6% (2001), 12% (2002), 25.5% (2003), 21.3% (2004), 23.6% (2005), P = 0.11, .
: TOAST classification: changes over time (2000–2005) of undetermined and cardioembolic strokes.
The medical treatments and surgical procedures that were used are listed in . Percutaneous PFO closure was reserved for cases of PFO that were associated with an atrial septal aneurysm, severe right-left shunting, and/or deep vein thrombosis. The exclusion of other possible causes of ischemia was also required for this procedure to be performed. No significant changes over the study period were observed.
Therapeutic approach in young patients with ischemic stroke.
The mortality rate during hospitalization was 2/324 (0.6%). The intrahospital complications included the following: epileptic seizures (6), urinary tract infections (18), respiratory infections (6), fever (12), and depressive symptoms (19). Among the surviving patients, 76.1% returned home, and 23.9% were transferred to a neurorehabilitation department. Multivariate analyses, which were adjusted for age and center, indicated that the probability of returning home was significantly higher in LACI (95.3%) than in non-LACI (73.1%) patients (P
= 0.02; OR = 11.03 (95%
CI = 2.48–49.08)). A higher probability of returning home was also observed in patients who were aged between 16 and 29 years (80.4%) and for those who were between 30 and 40 years of age (78.4%) when compared to the older group (71.9%). These differences, however, did not reach statistical significance (P
= 0.13 and P
= 0.08, resp.).