Risk Factors for Stroke

The incidence of stroke increases dramatically with advancing age, and increasing age is the most powerful risk factor for stroke. The incidence of stroke doubles each decade past 55 years of age. Half of all strokes occur in people older than 70 to 75 years. Overall, stroke incidence rate is 1.25 times greater in men than women. Men develop ischemic strokes at higher rates than women up to the age of 75 years. With an estimated 20% of the population being older than 65, greater than 10 million octogenarians, and an increasing life expectancy in the United States, it is predicted that in the near future, the incidence of stroke will reach 1 million per year. Compared to whites, African Americans have approximately a twofold increased risk of first-ever stroke. The rate of cerebral infarction is higher in African Americans and Hispanic Americans than in whites; this could be partially explained by the higher prevalence of DM and arterial hypertension experienced by these ethnic minorities. African Americans had been thought to have higher rates of intracranial atherosclerotic occlusive disease compared with whites, but this may actually reflect ascertainment bias (Sacco et al., 1995; Wityk et al., 1996). Furthermore, the stroke incidence and case fatality rates are also markedly different among the major ethnic groups in Auckland, New Zealand. Maori and Pacific Islands people have a higher rate of mortality within 28 days of stroke when compared with Europeans, especially men (Bonita et al., 1997). Chinese, Koreans, and Japanese also may have increased rates of intracranial hemorrhage and intracranial atherosclerotic cerebrovascular disease compared to whites. In comparison with the United States and Western Europe, where hemorrhagic stroke represents 20% or less of all stroke subtypes, 40% of the stroke subtypes are hemorrhagic (intracerebral or subarachnoid hemorrhage) in Japan. Conversely, in India, where ischemic stroke accounts for 80% of all strokes, 10% to 15% of strokes occur in people younger than 40 years and are mostly related to intracranial atherosclerosis.

Heredity and Risk of Stroke

Heredity seems to play a role in the pathogenesis of cerebral infarction. An increased risk is seen with a family history of stroke among first-degree relatives. Genetic factors have been linked with the pathogenesis of ischemic stroke, but specific genetic variants remain largely unknown, and some purported genetic associations have not been replicated (Chinnery et al., 2010). There are a number of genetic causes of stroke. Some inherited diseases, such as the hereditary dyslipoproteinemias, predispose to accelerated atherosclerosis. A number of inherited diseases are associated with nonatherosclerotic vasculopathies, including Ehlers-Danlos (especially type IV) syndrome, Marfan syndrome, Rendu-Osler-Weber disease, and Sturge-Weber syndrome. Familial atrial myxomas, hereditary cardiomyopathies, and hereditary cardiac conduction disorders are examples of inherited cardiac disorders that predispose to stroke. Deficiencies of protein C and S or antithrombin (AT) are examples of inherited hematological abnormalities that can cause stroke. Finally, rare inherited metabolic disorders that can cause stroke include mitochondrial encephalopathy, lactic acidosis, and stroke-like episodes (MELAS), Fabry disease, and homocystinuria. The presence of the apolipoprotein epsilon-2 allele in elderly individuals and deletion of the gene for the angiotensin-converting enzyme may increase the risk for stroke, but the association with stroke subtype is unclear (Agerholm-Larsen et al., 1997; Slooter et al., 1997; Szolnoki et al., 2001). The most significant findings of the ongoing Siblings With Ischemic Stroke Study (SWISS) to date have been the relationship between age and stroke in probands and sibs, and the lack of a tight association among ischemic stroke subtypes (Adams Jr. et al., 1993) within families (Meschia et al., 2005; Meschia et al., 2006; Wiklund et al., 2007).

Common Modifiable Risk Factors

At least 25% of the adult population has arterial hypertension, defined as systolic blood pressure (SBP) greater than 140 mm Hg or diastolic blood pressure (DBP) greater than 90 mm Hg. Prehypertension is defined as SBP between 120 and 139 mm Hg or DBP between 80 and 90 mm Hg. Optimal blood pressure is defined as SBP less than 120 mm Hg and DBP less than 80 mm Hg (Chobanian et al., 2003). Similar guidelines have been developed by the European Society of Hypertension–European Society of Cardiology Guidelines Committee (2003). The JNC7 Report (Table 51A.2) emphasizes that patients at risk, including those with DM or a history of stroke, should be treated with medications. Unfortunately, arterial hypertension remains poorly treated worldwide, and in the United States many patients are either undertreated or untreated (Hajjar and Kotchen, 2003). Arterial hypertension predisposes to ischemic stroke by aggravating atherosclerosis and accelerating heart disease, increasing the relative risk for stroke an estimated three- to fourfold. The risk is greater for patients with isolated systolic hypertension and elevated pulse pressure. Arterial hypertension is also the most important modifiable risk factor for stroke and the most powerful risk factor for all forms of vascular dementia (vascular cognitive impairment). Lowering blood pressure in stroke survivors helps prevent recurrent stroke and is more important than the specific hypotensive agent used, although there is some suggestion that beta-blockers are less preferred than other agents based on recent meta-analyses (Lindholm et al., 2005). Blood pressure treatment that results in a modest reduction in SBP of 10 to 12 mm Hg and 5 to 6 mm Hg diastolic is associated with a 38% reduction in stroke incidence (MacMahon and Rodgers, 1996). Treatment of isolated systolic hypertension in the elderly is also effective for reducing stroke risk. The Systolic Hypertension in the Elderly Program showed a 36% reduction in nonfatal plus fatal stroke over 5 years in the age 60-and-older group when isolated systolic hypertension was treated. Treating systolic hypertension also slows the progression of carotid artery stenosis. The PROGRESS trial evaluated the effects of perindopril and indapamide on the risk for stroke in patients with histories of stroke or transient ischemic attack (TIA). Regardless of blood pressure at entry, patients clearly benefited from treatment (PROGRESS Collaborative Group, 2001).

About 171 million people worldwide have type 2 DM, including 18 million Americans. It is estimated that these numbers will grow to 366 million people worldwide and 30 million Americans by 2030 (Fonseca et al., 2002). Diabetes mellitus increases the risk of ischemic cerebrovascular disease an estimated two- to fourfold as compared with the risk in people without diabetes. In addition, DM increases morbidity and mortality after stroke. Macrovascular disease is the leading cause of death among patients with DM. The mechanisms of stroke secondary to diabetes may be caused by cerebrovascular atherosclerosis, cardiac embolism, or rheological abnormalities. The excess stroke risk is independent of age or blood pressure status. Diabetes associated with arterial hypertension adds significantly to stroke risk. There is a fourfold increase in the relative risk of cardiovascular events among patients with diabetes and hypertension compared to those without the two conditions. Diabetic persons with retinopathy and autonomic neuropathy appear to be a group at particularly high risk for ischemic stroke. High insulin levels increase the risk for atherosclerosis and may represent a pathogenetic factor in cerebral small-vessel disease. Presently, no evidence exists that tighter diabetic control or normal HbA_1c levels over time decrease the risk for stroke or stroke recurrence. Moreover, optimal target blood glucose levels in stroke patients remain largely unknown (Gray et al., 2004; Van den Berghe et al., 2006). The UK Glucose Insulin in Stroke Trial (GIST-UK) failed to demonstrate any clinical benefit of insulin-induced euglycemia (target glucose 72-126 mg/dL). Furthermore, mortality appeared to be higher among patients with the greatest glucose reduction (Gray et al., 2009). Moreover, in the Spectroscopic Evaluation of Lesion Evolution in Stroke: Trial of Insulin for Acute Lactic Acidosis (SELESTIAL), the infusion of glucose-potassium-insulin did not have a favorable impact on cerebral infarct growth (McCormick et al., 2010).

High total cholesterol and high low-density lipoprotein (LDL) concentration are correlated with atherosclerosis. Dyslipidemia is a recognized risk factor for ischemic stroke. Meta-analyses have suggested that ischemic stroke risk increases with increasing serum cholesterol, and the reduction in stroke risk associated with 3-hydroxy 3-methylglutaryl coenzyme A reductase inhibitor (statin) therapies is related to reduction in LDL cholesterol (Amarenco et al., 2004; Tirschwell et al., 2004). Lipid-modifying therapy with statins has definitively established that reduction of LDL cholesterol reduces cardiovascular risk. Statins benefit stroke survivors as well. Lipid-lowering agents may slow progression of atherosclerotic plaque growth and may possibly cause a regression in plaque formation.

The Scandinavian Simvastatin Survival Study (4S) (Scandinavian Simvastatin Survival Study Group, 1999) investigated cholesterol lowering in persons with coronary heart disease and hypercholesterolemia and reported a highly significant relative reduction in the total mortality rate, major coronary events, and number of cardiac revascularization procedures. Post hoc analysis also showed a 28% reduction in fatal or nonfatal stroke and TIAs (Pedersen et al., 1998). The Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) study investigated cholesterol lowering with pravastatin in patients with a previous myocardial infarction (MI) or unstable angina who had cholesterol levels between 155 and 271 mg/dL and reported a remarkable reduction in MI, cardiac revascularizations, and cardiovascular deaths, as well as a 20% reduction in the risk for stroke (Long-Term Intervention with Pravastatin in Ischaemic Disease [LIPID] Study Group, 1998). Similar findings were associated with atorvastatin in the Myocardial Ischemia Reduction with Aggressive Cholesterol Lowering (MIRACL) trial and the Anglo-Scandinavian Cardiac Outcomes Trial-Lipid Lowering Arm (ASCOT-LLA) studies. MIRACL showed a 50% relative risk reduction (RRR) (P = .045) in stroke among high-risk coronary disease patients (Schwartz et al., 2001). The ASCOT-LLA study demonstrated a favorable trend for fatal and nonfatal stroke, with a 27% RRR in patients at low risk for coronary events (hazard ratio, 0.73; 95% confidence interval [CI], 0.56-0.96; P = .0236), though not to the prespecified endpoint of P = .01 (Sever et al., 2003).

Although the Heart Protection Study (HPS) of patients at high risk with DM, coronary artery disease, or other atherosclerotic vascular disease did show an overall reduction in stroke risk, a subgroup analysis of the Heart Protection Study (HPS) did not show a reduction in the risk for stroke among patients with prior cerebrovascular disease (Heart Protection Study Collaborative Group, 2004). This subgroup analysis was limited in that the mean time from event to enrollment was 4.3 years, and the LDL reduction was 38 mg/dL. Subsequently, the Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) study was published (SPARCL Investigators, 2006). This was a study of 4731 patients who had suffered either a stroke or TIA, had no known coronary heart disease, had LDL cholesterol levels of 100 to 190 mg/dL, and who were randomized to placebo or 80 mg of atorvastatin. A 56-mg/dL reduction in LDL treatment was noted in patients on atorvastatin. During a median follow-up of 4.9 years, 11.2% of atorvastatin-treated patients and 13.1% of placebo patients had a fatal or nonfatal stroke for an adjusted hazard ratio of 0.84 (95% CI, 0.71-0.99). A small increase in hemorrhagic stroke was reported in the atorvastatin group. Moreover, recent studies evaluating withdrawal of statins in acute ischemic stroke showed a higher incidence of death or dependency at 90 days (Blanco et al., 2007). Current guidelines of the American Heart Association and proposed modifications of the NCEP-III guidelines would therefore suggest that all patients at risk for stroke or who have had a cerebral infarction should be treated to a goal LDL level of below 70 mg/dL (Grundy et al., 2004; Sacco et al., 2006).

Atrial fibrillation, the most common sustained cardiac arrhythmia in the general population, affects about 1% of adults, is the most common cause for cardioembolic stroke, and is a risk factor for future cardiovascular disease. An estimated 1 to 2 million Americans have chronic nonvalvular atrial fibrillation (NVAF), a condition that is associated with an overall risk for stroke of approximately five- to sixfold, and a mortality rate approximately twice that of age- and sex-matched individuals without atrial fibrillation. The prevalence of atrial fibrillation increases with advancing age and is 0.5% for patients aged 50 to 59 years and 8.8% for those aged 80 to 89 years. Approximately 70% of individuals with atrial fibrillation are between 65 and 85 years of age. NVAF is associated with a substantial risk for stroke. Heart failure, arterial hypertension, DM, prior stroke or TIA, and age older than 75 years increase the risk for embolism in patients with NVAF. High-risk patients have a 5% to 7% yearly risk for thromboembolism. The CHADS2 Score represents a validated quantification of risk, with congestive heart failure, hypertension, age older than 75 years, and a history of DM being assigned 1 point; stroke or TIA are assigned 2 points (Gage et al., 2001). Ischemic stroke rates increase from 1.9 to 18.2 events per 100 patient-years with CHADS 2 scores of 0 and 6, respectively. Warfarin therapy, with the International Normalized Ratio (INR) value adjusted to between 2.0 and 3.0, decreases the relative risk for stroke in patients with NVAF by approximately two-thirds. High-risk patients, regardless of age, sustain particular benefit from warfarin anticoagulation. Left atrial enlargement also increases the risk for stroke in men. Likewise, left ventricular hypertrophy as demonstrated by electrocardiography (ECG) in men with preexisting ischemic heart disease is a major risk factor for stroke.

Cigarette smoking is the leading cause of preventable death in the United States. Cigarette smoking is a major risk factor for coronary artery disease, stroke, and peripheral arterial disease, an independent risk factor for ischemic stroke in men and women of all ages, and a leading risk factor of carotid atherosclerosis in men. The risk for stroke in smokers is two to three times greater than in nonsmokers. The mechanisms of enhanced atherogenesis promoted by cigarette smoking are incompletely understood but include reduced capacity of the blood to deliver oxygen, cardiac arrhythmias, increased blood coagulability, and triggering of arterial thrombosis and arterial spasm. Tobacco also increases carotid artery plaque thickness. More than 5 years may be required before a reduction in stroke risk is observed after cessation of smoking. Switching to pipe or cigar smoking is of no benefit. Counseling, nicotine replacement products, varenicline (a nicotinic receptor partial agonist),and bupropion are efficacious smoking cessation treatments. Selective blockers of the cannabinoid receptor type 1, such as rimonabant, have been proposed for treatment of multiple cardiometabolic risk factors including smoking and abdominal obesity (Gelfand and Cannon, 2006).

There is a J-shaped association between alcohol consumption and ischemic stroke; light to moderate use (up to two drinks a day) evenly distributed throughout the week offers a reduced risk, whereas heavy alcohol consumption is associated with an increased risk for total stroke. Heavy drinking may precipitate cardiogenic brain embolism. Alcohol consumption increases the risk for hemorrhagic stroke; alcohol-induced hypertension predisposes to spontaneous intracranial hemorrhage. Furthermore, active drinkers have a higher frequency of obstructive apneas and more severe hypoxemia. Conversely, moderate alcohol consumption may reduce the risk for ischemic stroke and may elevate HDL concentration. Elimination or reduction of alcohol consumption is recommended for heavy drinkers. Restriction of alcohol consumption to fewer than two drinks daily is recommended for light to moderate drinkers and nonpregnant women.

The prevalence of obesity (body mass index of 30 or higher) has increased nationwide. More than 61% of adult Americans are overweight, and 27% are obese. Obesity, particularly abdominal or truncal, is an important risk factor for cardiovascular disease in men and women of all ages. There is some evidence that physical activity can reduce the risk for stroke. Regular exercise lowers arterial blood pressure, decreases insulin resistance, increases HDL cholesterol, and is associated with lower cardiovascular morbidity and mortality.

Numerous studies have established an association between obstructive sleep apnea (OSA) and stroke; moreover the severity of OSA is much higher in stroke patients than in controls (Dyken, 2000). Habitual snoring increases the risk for stroke and adversely affects the outcome of patients admitted to the hospital with stroke. Mounting evidence also suggests that inflammation, lipoprotein(a) concentration, impaired fibrinolysis, and increased thrombotic potential are important nontraditional cardiovascular risk factors.

Atherosclerotic lesions of the carotid artery bifurcation are a common cause of stroke. Asymptomatic carotid artery disease carries a greater risk for vascular death from coronary artery disease than from stroke. Persons with an asymptomatic carotid bruit have an estimated annual risk for stroke of 1.5% at 1 year and 7.5% at 5 years. Asymptomatic carotid artery stenosis of less than 75% carries a stroke risk of 1.3% annually; with stenosis of greater than 75%, the combined TIA and stroke rate is 10.5% per year, with most events occurring ipsilateral to the stenosed carotid artery. Plaque composition may be an important factor in the pathophysiology of carotid artery disease. Plaque structure rather than degree of carotid artery stenosis may be a critical factor in determining stroke risk. Ultrasonographic carotid artery plaque morphology may identify a subgroup of patients at high risk for stroke. Ulcerated, echolucent, and heterogeneous plaques with a soft core represent unstable plaques at high risk for producing arterioarterial embolism.

Patients who suffer TIAs are at greater risk than normal controls for stroke or death from vascular causes. The risk for stroke is approximately three times higher. Symptomatic carotid artery stenosis of greater than 70% carries an annual risk for stroke of approximately 15%. Approximately 10% to 15% of those experiencing a stroke have TIAs before their stroke. Patients with hemispheric TIAs are at greater risk for ipsilateral stroke than patients with retinal TIAs. Patients with a first stroke are at greater risk for recurrent stroke, especially (but not exclusively) early after the first stroke. Those who suffer a recurrent stroke have a higher mortality than patients with first stroke. If the recurrence is contralateral to the first stroke, prognosis for functional recovery is poor. The risk for stroke recurrence is also increased by the presence of underlying dementia.

The aorta is the most frequent site of atherosclerosis. Protruding atheroma may be the cause of otherwise unexplained TIAs or strokes. Aortic-arch atheromatosis detected by transesophageal echocardiography is an independent risk factor for cerebral ischemia; the association is particularly strong with mobile and thick atherosclerotic plaques more than 4 mm in thickness (French Study of Aortic Plaques in Stroke Group, 1996). The prevalence of ulcerated plaques was 16.9% among patients with cerebrovascular diseases, compared to 5.1% among patients with other neurological diseases. Remarkably, ulcerated plaques were found in 61% of cryptogenic cerebral infarcts, compared to 22% of cerebral infarcts with a known cause.

Other Risk Factors for Stroke

Hemostatic factors may be important in assessing the risk for cerebrovascular disease. Elevated hematocrit, hemoglobin concentration, and increased blood viscosity may be indicators of risk for ischemic stroke. Elevation of plasma fibrinogen is an independent risk factor for the development of cerebral infarction. Epidemiological studies have shown a correlation between elevated plasma fibrinogen levels and both ischemic stroke incidents and mortality. An elevated plasma fibrinogen level may reflect progression of atherogenesis. Plasminogen activator inhibitor-1 excess and factor VII are independent risk factors for coronary heart disease. Compared with white Americans, black Americans have higher mean levels of fibrinogen, factor VIII, von Willebrand factor, and AT, and lower mean levels of protein C. Fibrinogen levels are closely correlated with other stroke risk factors such as cigarette smoking, arterial hypertension, diabetes, obesity, hematocrit levels, and spontaneous echocardiographic contrast. Antiphospholipid (aPL) antibodies are a marker for an increased risk for thrombosis, including TIAs and stroke, particularly in those younger than 50 years. In older patients, the presence of aPL (either lupus anticoagulants [LAs] or anticardiolipin [aCL]) among patients with ischemic stroke does not predict either increased risk for subsequent vascular occlusive events over 2 years or a differential response to aspirin or warfarin therapy. As such, routine screening for aPL in older patients with ischemic stroke does not appear warranted (Levine et al., 2004) The factor V Leiden mutation is associated with deep venous thrombosis in otherwise healthy individuals with additional prothrombotic risk factors. An overall association of the factor V Leiden mutation and arterial thrombosis has not been found. As opposed to homozygous mutations, heterozygous factor V Leiden and prothrombin gene mutations have no clear association with increased stroke risk (Fields and Levine, 2005). Elevated von Willebrand factor is a risk factor for MI and ischemic stroke. Elevated levels of fasting total homocysteine (normal 5-15 mM), a sulfhydryl-containing amino acid, have been associated with an increased risk for stroke and thrombotic events in case-controlled studies. Metabolism of homocysteine requires vitamin B₆ (pyridoxine), vitamin B₁₂ (cyanocobalamin), folate, and betaine. Plasma homocysteine concentrations may be reduced by the administration of folic acid alone or in combination with vitamins B₆ and B₁₂. Conversely, serum folate concentrations of 9.2 nM or less have been associated with elevated plasma levels of homocysteine, and a decreased folate concentration alone may be a risk factor for ischemic stroke, particularly among blacks (Giles et al., 1995). However, folic acid supplementation does not have a major impact on stroke reduction (Lee et al., 2010).

Stroke is uncommon among women of childbearing age, estimated at 4.4/100,000 (Petitti et al., l997). The relative risk for ischemic stroke is increased among users of high-dose estrogen oral contraceptives, particularly with coexistent arterial hypertension, cigarette smoking, and increasing age. Based on an odds ratio of 1.93, the risk is increased to 8.5/100,000, which translates into a number needed to harm of 24,000 women to cause one ischemic stroke (Gillum et al., 2000). Thus, for a healthy young woman without any other stroke risk factors, the risk of stroke associated with oral contraceptives is small and probably outweighed by their benefits. New agents containing lower doses of estrogen and progestin have reduced the frequency of oral contraceptive–related cerebral infarction. Two postmenopausal hormone replacement studies with equine estrogen (Premarin) showed no benefit in reducing the incidence of stroke in a cohort of women with coronary heart disease (Hulley et al., 1998). The Women’s Estrogen for Stroke Treatment study of estradiol replacement in postmenopausal women status post stroke also failed to show a reduction in recurrent stroke (Viscoli et al., 2001). In addition, the Women’s Health Initiative, a prospective randomized trial of estrogen therapy in healthy postmenopausal women, was halted prematurely because the risks outweighed the benefits. Absolute excess risks per 10,000 person-years attributable to estrogen plus progestin were sevenfold for coronary heart disease events, eightfold for strokes, eightfold for pulmonary thromboembolism, and eightfold for invasive cancers, whereas absolute risk reductions per 10,000 person-years were sixfold for colorectal cancers and fivefold for hip fractures (Rossouw et al., 2002). The risk for thrombosis associated with pregnancy is high in the postpartum period. The risk for cerebral infarction is increased in the 6 weeks after delivery but not during pregnancy.

A diurnal and seasonal variation of ischemic events occurs. Circadian changes in physical activity, catecholamine levels, blood pressure, blood viscosity, platelet aggregability, blood coagulability, and fibrinolytic activity may explain the circadian variations in onset of myocardial and cerebral infarction. Although an early-morning peak occurs for all subtypes of stroke, most clinical trials on the use of platelet antiaggregants or other antithrombotic agents do not take these circadian variations into account. Rhythmometric analyses support the notion that stroke is a chrono-risk disease, in which cold temperatures also represent a risk factor. A history of recent infection, particularly of bacterial origin and within 1 week of the event, is also a risk factor for ischemic stroke in patients of all ages. A number of recent reports suggest that Chlamydia pneumoniae, a causative organism of respiratory infections, may have a role in carotid and coronary atherosclerosis. Some studies have also identified an association with chronic infections with Helicobacter pylori and cytomegalovirus (Bittner, 1998; Ross, 1999). These findings have not been confirmed, however (Lindsberg and Grau, 2003).

Transient Ischemic Attacks

An estimated 400,000 individuals experience a TIA each year. A TIA is a prognostic indicator of stroke, with one-third of untreated TIA patients having a stroke within 5 years. About 1 in 10 patients with TIA experience a stroke in the next 3 months. The interval from the last TIA is an important predictor of stroke risk; of all patients who subsequently experience stroke, 21% do so within 1 month and 51% do so within 1 year of the last TIA. In one series, patients with TIA had a 3-month stroke risk of 10.5%, equal to the recurrence rate following a stroke. Furthermore, 50% of those strokes following a TIA occurred within 48 hours of TIA onset (Johnston et al., 2000). Cardiac events are the principal cause of death in patients who have had a TIA. The 5% to 6% annual mortality rate after TIA is mainly caused by MI, similar to the 4% annual cardiac mortality rate in patients with stable angina pectoris.

A TIA is a temporary and “non-marching” neurological deficit of sudden onset; attributed to focal ischemia of the brain, retina, or cochlea; and lasting less than 24 hours. Yet most TIAs last only a few minutes. Episodes that last longer than 1 hour are usually due to small infarctions. With the advent of diffusion-weighted magnetic resonance imaging (DW-MRI) sequences, the time-based definition of TIA is inadequate because infarctions are sometimes evident on DW-MRI in patients whose clinical manifestations resolved completely within a few hours. Thus, a “tissue-based” modification of the TIA definition has been proposed, suggesting that any transient episode, regardless of duration, associated with a clinically appropriate lesion by MRI be defined as a stroke, and that otherwise prolonged events (>1-6 hours in duration) be defined as stroke rather than TIA when otherwise clinically appropriate (Albers et al., 2002). Furthermore, DW-MRI is useful in predicting the risk for early stroke; patients with TIA and DW-MRI lesions are at greater risk for experiencing a subsequent stroke than patients without a lesion (Coutts et al., 2005). Johnston et al. also introduced a new unified score for risk stratification in patients with TIAs, known as the ABCD2 score. The ABCD2 score is based on age, blood pressure (≥140/90 mm Hg), clinical features, TIA duration, and diabetes (Table 51A.3). ABCD2 scores of 4 or greater indicate a moderate to high stroke risk and justify prompt hospital admission (Johnston et al., 2007).

Table 51A.3 ABCD2 Score

The onset of TIA symptoms is sudden, reaching maximum intensity almost immediately. To qualify as a TIA, therefore, an episode should also be followed by complete clinical recovery. TIAs involving the carotid circulation should be distinguished from those involving the vertebrobasilar circulation. Headaches often occur in patients with TIAs. As such, migraine with aura may sometimes be indistinguishable from TIA.

Agreement between physicians to define the likelihood of a TIA, even among fellowship-trained neurologists, remains poor (Castle et al., 2010). The following symptoms are considered typical of TIAs in the carotid circulation: ipsilateral amaurosis fugax, contralateral sensory or motor dysfunction limited to one side of the body, aphasia, contralateral homonymous hemianopia, or any combination thereof. The following symptoms represent typical TIAs in the vertebrobasilar system: bilateral or shifting motor or sensory dysfunction, complete or partial loss of vision in the homonymous fields of both eyes, or any combination of these symptoms. Perioral numbness also occurs. Isolated diplopia, vertigo, dysarthria, and dysphagia should not be considered as being caused by a TIA unless they occur in combination with one another or with any of the other symptoms just listed (Box 51A.1). Older patients with isolated vertebrobasilar symptoms and a significant history of cardiovascular risk factors should, however, be evaluated for possible TIA or stroke, because they are at substantially higher risk for cerebrovascular events (Norrving et al., 1995).

Occlusive disease in the subclavian arteries or the innominate artery can give rise to extracranial steal syndromes. The most well-defined syndrome is subclavian steal syndrome (SSS). In SSS, reversal of flow in the vertebral artery is caused by a high-grade subclavian artery stenosis or occlusion proximal to the origin of the vertebral artery from the aortic arch or innominate artery, with resultant symptoms of brainstem ischemia, usually precipitated by actively exercising the ipsilateral arm. The left side is involved most frequently. With innominate artery occlusion, the origin of the right carotid is also subject to the consequences of reduced pressure. SSS can be suspected by the presence of a reduced or delayed radial pulse and diminished blood pressure in the affected arm relative to the contralateral arm. A subclavian steal may be symptomatic or asymptomatic. Many patients have angiographic evidence of reversed vertebral blood flow without ischemic symptoms. Transcranial Doppler ultrasonography may detect transient retrograde basilar blood flow. Retrograde vertebral artery flow is a benign entity. Brainstem infarction is an uncommon complication of the subclavian steal syndrome.

Transient global amnesia (TGA) is characterized by a reversible antegrade and retrograde memory loss, except for a total amnesia of events that occur during the attacks and inability to learn newly acquired information. During the attacks, patients remain alert without motor or sensory impairments and often ask the same questions repeatedly. Patients are able to retain personal identity and carry on complex activities. TGA most commonly affects patients in their 50s and older. Men are affected more commonly than women. The attacks begin abruptly and without warning. A typical attack lasts several hours (mean, 3-6 hours) but seldom longer than 12 hours. Onset of TGA may follow physical exertion, sudden exposure to cold or heat, or sexual intercourse. Although a large number of conditions have been associated with transient episodes of amnesia, in most instances, TGA is of primary or unknown cause. TGA has been documented in association with epilepsy, migraine, intracranial tumors, overdose of diazepam, cardiac arrhythmias secondary to digitalis intoxication, and as a complication of cerebral and coronary angiography. Many reports have suggested a vascular causal factor for this heterogeneous syndrome. Bilateral hippocampal and parahippocampal complex ischemia, possibly of migrainous origin, in the distribution of the posterior cerebral arteries is a potential mechanism. Acute confusional migraines in children and TGA have a number of similar features. Others have suggested an epileptic causal factor for a minority of patients. Venous hypertension with transient hypoxemia in the context of incompetent internal jugular vein valves has also been suggested as a possible mechanism for TGA (Nedelmann et al., 2005