KEY POINTS
• Approximately 15% of all strokes affect young patients, in the most productive period of life. Recent evidence suggests an increasing incidence with variable mortality.
• Stroke is an important health problem in Latin America, but there are only few well designed population based stroke incidence studies.
• The 3 most frequent vascular risk factors in our popula tion were hypertension, sedentarism, and smoking. At least 1 of them was present in nearly 75% of patients with ischemic disease.
• The recognition of etiologies and vascular risk factors among young patients with stroke is of outmost impor tance to develop public health policies with the aim of primary and secondary prevention
Approximately 15% of all strokes affect young patients, with substantial consequences because affects people in their most productive period of their lives. Currently we lack specific management guidelines for this group of patients. Moreover the list of stroke etiologies is more diverse compared to older people; and almost one-third remains cryptogenic despite extensive work up1-3.
Recent evidence suggests an increasing incidence4 with variable mortality (between 2-6%). There is worldwide variation attributed to better control of vascular risk factors (VRF) in richer countries2,3,5.
Stroke is an important health problem in Latin America (LA)6, but there are only few well designed population based stroke incidence studies in LA7-13. The reported prevalence of stroke in an Argentine town with 75 000 inhabitants, was of 868 per 100 000 population, and less than 25% were patients under 60 years of age14. Two national hospital registries identified hypertension as the most prevalent VRF in our country12,13.
We organized the AISYF study (Argentinean Initiative to study Stroke in the Young and Fabry disease) first national, multicenter, and prospective study of stroke in young patients in Argentina, in order to identify their etiologies, clinical characteristics and VRF.
Materials and methods
We performed a prospective, multicenter, nationwide study of ischemic and hemorrhagic stroke in young adults in Argentina. The study was conducted at 22 public and private centers across the country, coordinated by the Department of Neuro-logy at Hospital Británico de Buenos Aires between January 2011 and December 2015. The protocol was approved by our institutional review board. All patients gave their informed con sent. Results regarding the identification of patients with Fabry disease in young patients with stroke were already published15.
Inclusion criteria: we enrolled patients between the ages of 18 and 55 years with either a TIA (defined as a transient episode of neurological dysfunction caused by focal brain, spinal cord, or retinal ischemia, without acute infarction), an brain infarct (defined as a focal neurological deficit due to in farction of central nervous system tissue)16 or an intracerebral hemorrhage (defined as a focal neurological deficit associated with focal collection of blood within the brain parenchyma)17 within 180 days of their stroke.
Exclusion criteria: patients with ischemic stroke following subarachnoid hemorrhage, cancer or trauma and patients with a hemorrhagic stroke due to a vascular malformation (e.g., aneurysm, arteriovenous malformation or cavernous hemangioma) or suspected to be related to cancer, trauma or anticoagulation. We also excluded those with either an epidural or subdural hemorrhage.
Stroke Subtype Classification and Etiological Workup: demographic data, cardiovascular risk factors, clinical and neuroimaging data were collected. VRF computed tomogra phy and/or magnetic resonance images were performed in all patients. They all underwent comprehensive etiological investigations, including brain and vascular imaging, electrocardiography, echocardiography, extensive laboratory testing. Information on comorbidities and VRFs were collected using a standardized case report form. All variables analyzed were checked for completeness, range, and outliers. TOAST criteria were used to define the clinical subtypes of ischemic stroke18. Intracerebral hemorrhage was categorized by location17,19.
Analysis of the demographic characteristics: the follow ing VRFs were evaluated: -Hypertension (systolic blood pressure > 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg in 2 recordings outside the acute event or treat ment with antihypertensive medication at least during the previous 2 weeks).
-Diabetes (randomized blood glucose > 200 mg/dl or fas-ting blood glucose ≥ 126 mg/dl in 2 determinations). - Smoking (one or more cigarettes per day in the last 2 months). - Alcohol abuse (intake > 100 g/day = one liter of wine or two liters of beer per day in the last 2 months, or acute intoxication during the 24 hours prior to stroke). - Overweight (body mass index between 25 and 29.9 kg/m2) and obesity (body mass index> 30 kg/m2). -Sedentarism (walking <1 km/day). -Hyperlipidemia (total fasting cholesterol ≥ 200 mg/dl and/or fasting LDL cholesterol ≥130 mg/dl and/or fasting triglycerides ≥ 200 mg / dl).
In addition, personal and family history of previous cere brovascular events, history of migraine, use of contraceptives, sympathicomimetic vasoconstrictors, anti-migraine drugs and use of illicit drugs were recorded.
All data was presented as percentage for categorical vari ables or as mean and standard error for numerical variables. The t Test was used for numerical variables, and for categorical variables the x2 test was performed. A two-tailed value of p < 0.05 was considered statistically significant.
Results
A total of 311 patients were enrolled, we excluded 18 for incomplete information (Fig. 1). The remaining 293 had a mean age of 43.3 years (range 20-55 years); 158 were men (53.9%), 242 suffered cerebral infarctions (82.6%), 24 (8.2%) cerebral hemorrhages and 27 (9.2%) presented TIA. The frequency of stroke increased with age in both genders, with a higher prevalence in females among younger patients, and males among older patients (Fig. 2). VRFs were: hypertension 120 (41%), smoking 92 (31.4%), dyslipidemia 81 (27.6%), overweight/obesity: 74 (25.3%), diabetes 36 (12.3%), previous stroke/TIA 37 (12.6%), alcohol abuse 18 (6.1%), previous TIA 17 (5.8%) and previous hemorrhagic stroke 6 (2%).
In patients with ischemic stroke: 69 (25.7%) had no VRF, 71 (26.3%) had 1 VRF, and 129 (48%) had 2 or more VRFs. The most frequent were sedentarism, hypertension and smoking. Men had significantly higher frequency of sedentarism, arterial hypertension, obesity, alcohol con sumption and diabetes compared with females (Table 1). Etiologies according to the TOAST classification were: large vessel atheromatosis: n = 33 (12.3%), cardioembo lism: n = 20 (7.5%), small artery disease: n = 31 (11.5%), other causes: non atherosclerotic vasculopathy n = 73 (27.1%) and undetermined etiology: n = 112 (41.6%). Undetermined etiologies included 16 patients (5.9%) who presented 2 or more causes, 55 (20.4%) with a negative evaluation (cryptogenic), and 41 (15.2%) with an incom plete evaluation. Older patients presented a lower preva lence of undetermined causes and a higher frequency of large and small vessel disease (Fig. 3). Among the less common causes of stroke (including non-atherosclerotic vasculopathies and patients with more than one probable cause) we found: 31 with arterial dissections, 12 with cerebral venous thrombosis, 3 had migraine infarcts, 24 had thrombophilia including 2 with simultaneous altera tion of 2 coagulation factors (antiphospholipid syndrome in 11 patients, factor V Leiden in 8, prothrombin 20210 in 4, protein C deficiency in 1 and protein S deficiency in 2), 5 patients with vasculitis, 12 with strokes associated with vasoconstrictors or contraceptives and 1 patient each with CADASIL, extrinsic carotid artery compression, Klippel- Trenaunay-Weber syndrome and Fabry disease15.
We divided our patients into 2 groups according to their ages (< 45 years and ≥ 45 years). Those older than 45 years had significantly higher frequency of hypertension, diabetes, sedentarism, overweight/obesity, hypercholes terolemia, atherosclerotic disease and small vessel disease. Oral contraceptive use, presence of patent foramen ovale and non-atherosclerotic etiologies were significantly higher in younger patients (Table 2).
Clinical characteristics of patients with cerebral hemor rhage are summarized in Table 3. Compared with ischemic stroke patients the prevalence of arterial hypertension (p = 0.009) and alcohol consumption (p = 0.002) were significantly higher in those with brain hemorrhages.
Discussion
The recognition of etiologies and VRF among young patients with stroke is of outmost importance to develop public health policies with the aim of primary and sec ondary prevention. The 3 most frequent VRFs in our population including ischemic and hemorrhagic stroke were: hypertension, sedentarism, and smoking. One of the most relevant findings was the presence of at least 1 VRF in nearly 75% of patients with ischemic disease, and the frequency increased with age20-22. Other factors previ ously reported in association with stroke, such as diabetes and alcoholism23 were rare in our cohort, except in those with hemorrhagic strokes in whom alcohol consumption was significantly higher than in ischemic stroke24. Recent population studies have identified smoking as one of the most prevalent VRFs26. In our population smoking affected almost one-third, and it was the most frequent VRF among those under 45 years25,26. A similar finding was recently reported by European studies of stroke in young patients20,22.
The slight predominance of women among our younger patients; was probably related to specific risk factors at this age, such as the use of oral contraceptives, preg nancy, puerperium and migraine9,20,21,27. In a study of young Mexican women, migraine, smoking and use of oral contraceptives were the main risk factors for brain infarcts9. Moreover, the risk of stroke in youngers than 45 years suffering from migraine with aura is twice than that of those without aura28,29 and smoking associated with contraceptive use increases the risk of stroke20,30.
There is limited information of stroke in young patients in LA9-11. In a report of 106 Brazilian patients, (15 to 40 years old), 86% had ischemic strokes of which 16% were undetermined11. In a large Mexican cohort with infarcts, 300 were younger than 40 years and 32% were crypto genic10.
In our study 25% of patients with cerebral ischemia, and especially the younger ones, presented non-athero sclerotic vascular etiologies, including cervicocephalic arterial dissections, cerebral venous thrombosis and thrombophilia1,26. These pathologies require a high index of suspicion because they are frequently mis-diagnosed31.
The majority of stroke studies in young people used the etiologic classification of TOAST21,32-35. Our series, as in other similar studies, has a high proportion of patients included within indeterminate causes, probably due to 2 fundamental reasons: the operational classification criteria used in each study and the depth of the complementary examinations performed26. In some studies migraine in farcts, or those occurring during pregnancy or the puer perium are included as other identifiable causes35, while in other publications these etiologies are considered as indeterminate. Likewise, there is a greater frequency of indeterminate causes in the youngers due not only to the existence of less recognized etiologies in this age group but also to the greater presence of VRFs in the elderly which may explain different pathophysiological mecha nisms of the cerebrovascular attack.
The strength of our study is that this is, in our know-ledge, the first large, multicentric, national and prospec tive analysis of stroke in young patients in Argentina. Limitations include that not every center had access to all diagnostic tests and this may explain the high frequency of patients in the undetermined category. In addition, the majority of our patients were evaluated in urban medical centers and therefore it is possible that those in rural areas may present different clinical char acteristics. These limitations are frequent problems in developing countries7,36.
In conclusion, our results have multiple implications for the daily clinical practice and the development of public health policies. It demonstrated that VRFs are very fre quent even in young adults and therefore it is mandatory to develop strategies for stroke prevention in the younger population.