Historical note

In the 18th and 19th centuries, an indulgent lifestyle was thought to be a major cause of apoplexy, but probably only in combination with other factors (166). Some physicians noted an association between apoplexy and tobacco in people of certain constitutions, but this was neither a widespread nor a popular observation (181). It was commonly asserted that if apoplexy was to be avoided, it was important to lead a balanced and moderate lifestyle, with constant attention to “habitual discharges” (214). Muscular exertion of any kind, but especially “straining at stool,” was a common medical explanation for apoplexy. In 1788, Rowley suggested that violent passions of the mind, cold weather, tight clothing around the neck, constipation, and everything in the least bit flatulent should be avoided (189). Clarke’s advice in 1824 was to have regular bowel movements, a calm life, a moderate diet, and avoid smoking and drinking to excess (60).

As the 18th century progressed, the theory of blood pressure was gradually accepted. By the beginning of the 19th century, bloodletting appeared to change almost imperceptibly from being a means of restoring the balance of the humors to a means of reducing blood pressure (181). In fact, Cheyne conveyed the idea that in “plethoric” or “sanguineous” apoplexies, blood has a pressure of its own: “we find that every vessel within the heads has been in a state of excitement… the brain is torn up by the blood which they (arteries) had driven out of the course of the circulation” (53). Schiller notes that in the 1840s, cerebral hemorrhage was related to an increased “impulse” and as the century progressed, blood pressure began to be measured more frequently and was found to be high in nearly all cases of intracerebral hemorrhage (198). By 1892, Osler was advising that measures should be taken to reduce arterial pressure after apoplexy, which was most rapidly and satisfactorily achieved by venesection (165). Beginning in 1967, a landmark series of randomized trials compared antihypertensive medication with placebo in patients with severe diastolic hypertension (16). Subsequent trials comparing treatment versus placebo (or routine care) in patients with moderate, then mild, diastolic and systolic hypertension demonstrated incontrovertibly that a reduction in elevated blood pressure reduced stroke incidence in primary prevention (212; 62; 19).

Stroke is the second leading cause of death and the third leading cause of long-term disability worldwide, whereas it is the fifth-ranked cause of death in the United States. It has been estimated that in 2017 there were 11.9 million first-stroke patients, 104.2 million prevalent and 6.2 million fatal stroke cases, and 132.1 million stroke-related disability-adjusted life years lost globally (122). Low- and middle-income countries bear a particularly high stroke burden, with about 80% of all incident strokes, 77% of all stroke survivors, 87% of all deaths from stroke, and over 85% of all stroke-related disability-adjusted life years (122). The largest INTERSTROKE case-control study, with 26,919 participants in 32 countries, showed that in over 90% of cases, the risk of first stroke was significantly and independently related to 10 potentially modifiable risk factors: arterial hypertension, diabetes mellitus, abdominal obesity, lipids, cardiac causes, alcohol, tobacco, dietary patterns, insufficient physical activity, and psychosocial factors. In addition, the population-attributable risk of stroke for these 10 factors was similar across different regions of the world, ethnic groups, men and women, and younger and older age groups, although differences in the population-attributable risk may exist among geographical regions and specific populations (160).

Stroke has been identified as a priority area for World Health Organization actions to reduce the burden of noncommunicable diseases. WHO’s noncommunicable diseases campaign targets include a 25% relative risk reduction in the overall mortality from cardiovascular diseases by 2025. Primary prevention strategies are, therefore, essential to reducing the burden of stroke worldwide (44). One stroke prevention strategy is the population approach, usually implemented by governments or health organizations to reduce the prevalence of risk factors in the general population. Another strategy, the high-risk approach, is commonly implemented by the health workforce to reduce the prevalence and level of risk factors in high-risk individuals. Optimal primary stroke prevention requires an integrated approach with widely implemented, simple, and effective steps (44).

Primary prevention addresses all measures for avoiding a stroke or transient ischemic attack. Prevention of stroke and transient ischemic attack includes both conventional approaches to vascular risk factor management (blood pressure lowering, cholesterol reduction with statins, smoking cessation, and antiplatelet therapy) and more specific interventions, such as carotid revascularization or anticoagulation for atrial fibrillation. Modifiable risk factors are also highly prevalent in young patients (18 to 55 years), emphasizing the need for vigorous primary prevention measures in the younger population (230).

Hypertension and stroke

Despite control efforts to respond to the emerging high blood pressure burden, the prevalence of hypertension, and thus, the adverse impact on cardiovascular morbidity and mortality, is increasing globally. Of the estimated 1.39 billion of adults worldwide who have hypertension, two thirds are living in low- and middle-income countries (151). By 2025, one in three adults (1.56 billion) will be affected by hypertension globally.

Arterial hypertension is the most powerful stroke risk factor after age. Individuals with hypertension have up to a 5-fold higher risk of stroke. Phase 2 of the INTERSTROKE study, a large case-control set of 26,919 individuals from 32 countries worldwide, showed that globally 47.9% of stroke is attributable to hypertension (BP ≥ 140/90 mm), with higher figures for intracerebral hemorrhage (56.4%) and lower figures for ischemic stroke (45.7%) (160).

In a pooled analysis of 61 prospective studies including about 1 million individuals, the risk of stroke increased progressively, with blood pressure values as low as 115/75 mm/Hg (132). Each increment of 20/10 mm/Hg doubles the risk of cardiovascular disease. Such a correlation is consistent at all ages.

Over half of hypertensive individuals have additional risk factors, the most common being metabolic syndrome, obesity, lipid disorders, and diabetes mellitus (223). Such factors have an additive effect on the risk of cardiovascular disease, including stroke.

The International Society of Hypertension and the European Society of Cardiology continue to define arterial hypertension as systolic blood pressure of 140 mmHg or higher and/or diastolic blood pressure of 90 mmHg or higher on repeated measurements over an interval between 1 and 4 weeks (239; 223). The American College of Cardiology and the American Heart Association proposed new high blood pressure clinical practice guidelines that delineate four categories for blood pressure levels (236). Normal blood pressure is set at lower than 120/80 mm Hg, whereas elevated blood pressure is defined as systolic blood pressure between 120 to 129 mm Hg and diastolic blood pressure less than 80 mm Hg. Accordingly, arterial hypertension is defined as having a reading of over 130/80 mm Hg, based on data from observational studies of blood pressure, cardiovascular risk, and the benefit of blood pressure control in clinical trials. Stage 1 hypertension is 130 to 139 mm Hg/80 to 89 mm Hg and stage 2 is more than 140/90 mm Hg. In the new categories, a third stage of hypertension is no longer posited (236). This redefinition may affect statistics for arterial hypertension prevalence among adults in the United States by as much as 14% (156). Moreover, the recommended threshold for pharmacological treatment of hypertension is a blood pressure reading of over 130/80 mm Hg coupled with an estimated 10-year atherosclerotic cardiovascular disease risk of at least 10%. The lower blood pressure goal for treated hypertensive patients is defined as less than 130/80 mm Hg, regardless of atherosclerotic cardiovascular disease risk (236).

Therefore, the cornerstone of primary and secondary stroke prevention is control of elevated blood pressure. To date, the efficacy of blood pressure-lowering in reducing incident and recurrent strokes has been amply demonstrated.

Many trials have shown the benefit of blood pressure control in all age groups and all degrees of hypertension. One study showed a 35% reduction in total strokes and a 44% reduction in fatal strokes over 5 years with a standardized blood pressure lowering regimen, compared with routine care (17). Treatment of isolated systolic hypertension to less than 140 mmHg in elderly patients is also beneficial for reducing risk of stroke (65; 20). A meta-analysis of multiple treatment trials showed that a mean reduction in diastolic blood pressure of 5 to 6 mm Hg correlates with a 35% to 40% reduction in incidence of stroke. This treatment effect appears to be valid in different races and ages. Furthermore, a meta-analysis of nine prospective studies that included 420,000 people followed for 10 years found that the risk continued to decrease with lower blood pressure, without a lower limit in blood pressure (228). This suggests that antihypertensive therapy targeted toward the lowest tolerated blood pressure may be beneficial in patients at risk of stroke. However, intensive lowering of blood pressure versus standard treatment does not ameliorate cardiovascular disease, including stroke in individuals with diabetes mellitus and hypertension (33). Lowering blood pressure provides similar relative protection at all levels of baseline cardiovascular risk, but progressively greater absolute risk reductions as baseline risk increases (36).

In a large-scale meta-analysis of 37 randomized clinical trials of pharmacological blood pressure treatment involving 186,988 individuals (divided into seven groups based on systolic blood pressure at baseline) without a previous history of cardiovascular disease at entry, the mean age of participants was 65.3 years; 51% were men; the mean blood pressure at baseline was 157/89 mm Hg, 27% were on angiotensin-converting enzyme inhibitors; 27% received a calcium-channel blocker; 21% were on diuretics; 20% were under antiplatelet treatment; and 19.5% received lipid-lowering therapy (38). After a mean follow-up of 4.2 years, these individuals showed an average reduction of 5 mm Hg in systolic blood pressure, which lowered their rate of cardiovascular events by about 10% and their rate of stroke by 13%. The relative risk reductions were proportional to the magnitude of blood pressure lowering.

A relevant question here is whether specific classes of antihypertensive drugs, beyond their blood pressure lowering effect, might have beneficial effects on specific hypertensive patient subgroups. There is some evidence that specific classes of antihypertensive drugs have different effects or their pharmacological actions differ in patient subgroups. Currently, there are five classes of first-line blood pressure lowering drugs: diuretics, beta blockers, calcium channel blockers, angiotensin-converting inhibitors, and angiotensin II receptor blockers.

Clinical trials in hypertensive cardiovascular high-risk patients. Older published reports of controlled trials of antihypertensive treatment in the primary prevention of cardiovascular diseases included only small numbers of subjects (16; 123; 206). After these results, further clinical trials with adequate numbers of patients were designed.

The Medical Research Council (MRC) trial of treatment of mild hypertension. The MRC trial was a single-blind study that recruited 17,354 patients with mild hypertension, aged 35 to 64 years, randomly allocated at entry to take bendrofluazide or propranolol or placebo tablets (18). The stroke rate was reduced on active treatment: 60 strokes occurred in the treated group and 109 in the placebo group, giving rates of 1.4 and 2.6 per 1000 patient-years of observation, respectively (p < 0.01). The incidence of all cardiovascular events was reduced on active treatment: 286 events occurred in the treated group and 352 in the placebo group, giving rates of 6.7 and 8.2 per 1000 patient-years, respectively (p < 0.05). For mortality from all causes, treatment made no difference to the rates. Comparison of the two active drugs showed that the reduction in stroke rate on bendrofluazide was greater than that on propranolol: 18 events occurred in the bendrofluazide group and 42 in the propranolol group, giving rates of 0.8 and 1.9 per 1000 patient-years, respectively (p = 0.002).

The Medical Research Council (MRC) trial of treatment of hypertension in older adults. This MRC trial was a randomized, placebo-controlled, single-blind trial that randomized 4396 patients, aged 65 to 74, to receive diuretic (hydrochlorothiazide 25 mg or 50 mg plus amiloride 2.5 mg or 5 mg daily), beta blocker (atenolol 50 mg daily), or placebo (20). Patients had mean systolic pressures of 160 to 209 mm/Hg and mean diastolic pressures less than 115 mm/Hg during an 8-week run in and were not taking antihypertensive treatment. Both treatments reduced blood pressure below the level in the placebo group. Compared with the placebo group, actively treated subjects (diuretic and beta-blocker groups combined) had a 25% (95% CI 3% to 42%) reduction in stroke (p = 0.04), 19% (95% CI -2% to 36%) reduction in coronary events (p = 0.08), and 17% (95% CI 2% to 29%) reduction in all cardiovascular events (p = 0.03). After adjusting for baseline characteristics, the diuretic group had significantly reduced risks of stroke (31%, 95% CI 3% to 51%, p = 0.04), coronary events (44%, 95% CI 21% to 60%, p = 0.0009), and all cardiovascular events (35%, 95% CI 17% to 49%, p = 0.0005) compared with the placebo group. The beta-blocker group showed no significant reductions in these endpoints.

Intervention as a Goal in Hypertension Treatment (INSIGHT) study. INSIGHT was a prospective, randomized, double-blind trial of 6321 patients, aged 55 to 80 years, with hypertension (blood pressure ≥ 150/90 mm/Hg, or ≥ 160 mm/Hg systolic). All patients had at least one additional cardiovascular risk factor. Patients were randomly assigned to nifedipine 30 mg in a long-acting gastrointestinal-transport-system formulation (n = 3157), or to hydrochlorothiazide 25 mg plus amiloride 2.5 mg (n = 3164). Dose titration was done using dose doubling, with an addition of atenolol 25 to 50 mg or enalapril 5 to 10 mg. Primary outcome was cardiovascular death, myocardial infarction, heart failure, or stroke. Primary outcome occurred in 200 (6.3%) patients in the nifedipine group and in 182 (5.8%) patients in the hydrochlorothiazide plus amiloride group (RR 1.10, 95% CI 0.91 to 1.34). The authors concluded that nifedipine once daily or hydrochlorothiazide plus amiloride were found to be equally effective in preventing overall cardiovascular or cerebrovascular complications (43).

UK Prospective Diabetes Study (UKPDS 39). UKPDS was a randomized controlled trial comparing an angiotensin-converting enzyme inhibitor (captopril) with a beta blocker (atenolol) in patients with type 2 diabetes, with the aim of a blood pressure of <150/<85 mm/Hg; 1148 hypertensive patients with type 2 diabetes were also included (24). Of the 758 patients allocated to a tight control of blood pressure, 400 were allocated to captopril and 358 to atenolol; 390 patients were allocated to less tight control of blood pressure. Captopril and atenolol were reported to be equally effective in reducing blood pressure to a mean of 144/83 mm/Hg and 143/81 mm/Hg, respectively. Captopril and atenolol were found to be equally effective in reducing the risks of fatal and nonfatal stroke.

Losartan Intervention for Endpoint Reduction in Hypertension Study (LIFE Trial). The LIFE trial was an investigator-initiated, prospective, community-based, double-blind, randomized, active-controlled, parallel-group study. Eligible patients were men or women aged 55 to 80 years, with previously treated or untreated hypertension and electro-cardiographically documented left ventricular hypertrophy. After a 2-week placebo run-in period, patients entered a minimum 4-year, double-blind, active-treatment period. Active treatment continued for 4 years after the last patient was enrolled and until at least 1040 patients experienced a primary cardiovascular event. Antihypertensive therapy was titrated to achieve a goal blood pressure of less than 140/90 mm/Hg. Patients initially received losartan 50 mg or atenolol 50 mg once a day. After 2 months, hydrochlorothiazide 12.5 mg was added when blood pressure did not reach the goal, whereas after 4 months, the dose of losartan or atenolol was doubled to 100 mg plus hydrochlorothiazide 12.5 mg when blood pressure remained inadequately controlled. At 6 months, additional open-label antihypertensive medication (excluding beta-blockers, angiotensin-converting inhibitors, and angiotensin II receptor blockers), including upward titration of hydrochlorothiazide, could be added to reach goal blood pressure. If blood pressure was less than or equal to 160/95 mm/Hg, upward titration with additional open-label therapy was mandatory. The primary endpoint of LIFE was the composite of cardiovascular morbidity and mortality (defined as stroke, myocardial infarction, and cardiovascular death). By adjusting intention-to-treat analysis, losartan was seen to be superior to atenolol in reducing the risk of combined cardiovascular morbidity and mortality, that is, a significant 13% relative risk reduction versus atenolol (p = 0.021). Losartan reduced the risk of fatal and nonfatal stroke, a major cause of death and disability, by 25% compared with atenolol (64). The superior risk reduction in stroke from losartan use versus atenolol in the LIFE study suggests an important role for losartan in patients with hypertension beyond the effects of blood pressure reduction alone (118). Both losartan and atenolol substantially reduced systolic blood pressure and diastolic blood pressure (-30.2/16.6 mm/Hg with losartan, -29.1/16.8 with atenolol).

Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). The ALLHAT trial, a randomized, double-blind, active-controlled clinical trial, was conducted to determine whether treatment with a calcium channel blocker or an angiotensin-converting enzyme inhibitor could lower the incidence of coronary heart disease or other cardiovascular disease events versus treatment with a diuretic (08). A total of 33,357 subjects aged 55 years or older, with hypertension and with at least one other risk factor, were enrolled. Participants were randomly assigned to receive chlorthalidone 12.5 to 25 mg/d (n = 15,255), amlodipine 2.5 to 10 mg/d (n = 9048), or lisinopril 10 to 40 mg/d (n = 9054) with a planned follow-up of 4 to 8 years. The mean follow-up was 4.9 years. The primary outcome was combined fatal coronary heart disease or nonfatal myocardial infarction, analyzed by an intention to treat, which evidenced no differences among treatments. Likewise, all-cause mortality did not differ among groups. Five-year systolic blood pressures were significantly higher in the amlodipine (0.8 mm/Hg, p = 0.03) and lisinopril (2 mm/Hg, p < 0.001) groups compared with chlorthalidone, and 5-year diastolic blood pressure was significantly lower with amlodipine (0.8 mm/Hg, p < 0.001). Secondary outcomes were all-cause mortality, stroke, and combined coronary artery diseases (primary outcome, coronary revascularization, or angina with hospitalization, heart failure, and peripheral arterial disease). Regarding amlodipine versus chlorthalidone, secondary outcomes were similar except for a higher 6-year rate of heart failure with amlodipine. As for lisinopril versus chlorthalidone, lisinopril had reported higher 6-year rates of combined cardiovascular diseases, stroke (6.3% vs. 5.6%; RR 1.15, 95% CI 1.02 to 1.30), and heart failure. In the ALLHAT study, the stroke rate results for lisinopril versus chlorthalidone were analyzed for these subgroups: age younger than 65 years and 65 years or older, men and women, blacks and non-blacks, diabetics and non-diabetics. A significant differential effect was seen according to race (p = 0.01 and p = 0.04 for interaction, respectively). The reported relative risks were 1.4% (95% CI 1.17 to 1.68) and 1.00 (95% CI 0.85 to 1.17) for stroke in black versus non-black patients in the lisinopril group.

The International Verapamil-Trandolapril Study (INVEST). INVEST was a randomized, open-label, blinded endpoint study of 22,576 hypertensive coronary artery disease patients aged 50 years or older. Patients were randomly assigned to either calcium antagonist strategy (verapamil sustained release) or atenolol. Trandolapril and/or hydrochlorothiazide were administered to achieve blood pressure goals of less than 140 mm/Hg (systolic) and less than 90 mm/Hg (diastolic). When diabetes or renal impairment was present, blood pressure goals were set to be less than 130 mm/Hg (systolic) and less than 85 mm/Hg (diastolic). Primary endpoints were first occurrence of death (all causes), nonfatal myocardial infarction, or nonfatal stroke. After a follow-up of 61,835 patient-years (mean 2.7 years per patient), 2269 patients had a primary outcome event, with no statistically significant difference between treatment strategies (9.93% in verapamil group and 10.17% in atenolol group, RR 0.98, 95%CI 0.90 to 1.06). Two-year blood pressure control was similar among groups. No difference in the incidence of stroke was evidenced among the groups (174).

Controlled Onset Verapamil Investigation of Cardiovascular End Points (CONVINCE) trial. CONVINCE was a double-blind, randomized clinical trial in which a total of 16,602 participants diagnosed as having hypertension with one or more additional risk factors for cardiovascular disease were enrolled (34). After a mean follow-up of 3 years, the study was closed before unblinding the results. Initially, 8241 participants received 180 mg of controlled-onset, extended-release verapamil, and 8361 participants received either 50 mg of atenolol or 12.5 mg of hydrochlorothiazide. Primary endpoints were first occurrence of stroke, myocardial infarction, or cardiovascular disease-related death. Systolic and diastolic blood pressure were reduced by 13.6 and 7.8 mm/Hg for participants assigned to the verapamil group and by 13.5 and 7.1 mm/Hg for participants assigned to the atenolol or hydrochlorothiazide groups. There were 364 primary cardiovascular disease-related events that occurred in the verapamil group versus 365 in the atenolol or hydrochlorothiazide groups (HR 1.02, 95% CI 0.88 to 1.18). For fatal or nonfatal strokes, the hazard ratio was reported to be 1.15 (95% CI 0.90 to 1.48); for fatal or nonfatal myocardial infarctions, 0.82 (95% CI 0.65 to 1.03); and for cardiovascular disease-related deaths, 1.09 (95% CI 0.87 to 1.37).

The Valsartan Antihypertensive Long-term Use Evaluation (VALUE) trial. The VALUE trial was designed to verify the hypothesis that for the same blood pressure control, valsartan was capable of reducing cardiac morbidity and mortality better than amlodipine in hypertensive patients at high cardiovascular risk (114). In total, 15,245 patients, aged 50 years or older, with treated or untreated hypertension with a high risk of cardiac events participated in this randomized, double-blind, parallel-group trial comparing valsartan and amlodipine. Duration of treatment was event-driven, and the trial lasted until at least 1450 patients had reached a primary endpoint, defined as a composite of cardiac mortality and morbidity. Patients were followed for a mean of 4.2 years. Blood pressure was reduced by both treatments, but the effects of the amlodipine-based regimen were more pronounced, especially in the early period (blood pressure 4.0/2.1 mm/Hg was lower in the amlodipine group than in the valsartan group after 1 month, 1.5/1.3 mm/Hg after 1 year, p < 0.001). The primary composite endpoint was reported to occur in 810 patients in the valsartan group (10.6%, 25.5 per 1000 patient-years) and in 789 patients in the amlodipine group (10.4%, 24.7 per 1000 patient-years, HR 1.04, 95% CI 0.94 to 1.15, p = 0.49). Stroke incidence (a secondary endpoint) was seen to be lower in the amlodipine group than in the valsartan group.

The Anglo-Scandinavian Cardiac Outcomes Trial-Blood Pressure Lowering Arm (ASCOT-BPLA). ASCOT-BPLA was a multicenter, prospective, randomized controlled trial including 19,257 patients with hypertension, aged 40 to 79 years, and having at least three other cardiovascular risk factors (66). Patients were assigned to either amlodipine 5 to 10 mg plus perindopril 4 to 8 mg as required (amlodipine-based regimen n = 9639) or atenolol 50 to 100 mg plus bendroflumethiazide 1.25 to 2.50 mg with potassium as required (atenolol-based regimen n = 9618). The primary endpoint was nonfatal myocardial infarction (including silent myocardial infarction) and fatal coronary artery disease. Analysis was by intention to treat. The study was stopped prematurely after 5.5 years’ median follow-up and accumulated over a total of 106,153 patient-years of observation. Though not significant, compared with the atenolol-based regimen, fewer individuals on the amlodipine-based regimen had a primary endpoint (429 vs. 474, unadjusted HR 0.90, 95% CI 0.79 to 1.02), fatal and nonfatal stroke (327 vs. 422, HR 0.77, 95% CI 0.66 to 0.89), total cardiovascular events and procedures (1362 vs. 1602, HR 0.84, 95% CI 0.78 to 0.90), and all-cause mortality (738 vs. 820, HR 0.89, 95% CI 0.81 to 0.99). The incidence of developing diabetes was reported to be lower with the amlodipine-based regimen (567 vs. 799, HR 0.70, 95% CI 0.63 to 0.78).

The Systolic Blood Pressure Intervention Trial (SPRINT). SPRINT was a multicenter, randomized, controlled, open-label trial comprising 9361 patients aged over 50 years with an increased cardiovascular risk and systolic blood pressure ranging from 130 to 180 mm Hg (207). Patients with diabetes or previous stroke were excluded, and participants were randomized to either one of two intervention groups: the standard treatment group with a systolic blood pressure target of less than 140 mm Hg or the intensive treatment group with a target of less than 120 mm Hg. All major blood pressure–lowering agents were included, though those with the strongest evidence for reduction in cardiovascular outcomes were encouraged, including thiazide-type diuretics (mainly chlorthalidone) as the first-line agent, loop diuretics (for participants with advanced chronic kidney disease), and beta-adrenergic blockers (for those with coronary artery disease). At 1 year, the mean systolic blood pressure was 121.4 mm Hg in the intensive treatment group and 136.2 mm Hg in the standard treatment group, whereas the mean diastolic blood pressure was 68.7 mm Hg in the intensive treatment group and 76.3 mm Hg in the standard treatment group. Thus, the average difference in systolic blood pressure between both intervention groups was 14.8 mm Hg, whereas the average difference in diastolic blood pressure was 7.6 mm Hg. With a mean follow-up of 3.26 years, the mean systolic blood pressure was 121.5 mm Hg in the intensive treatment group and 134.6 mm Hg in the standard treatment group. Participants in the intensive treatment group required more low blood pressure medications than those in the standard treatment group (2.8 vs. 1.8, respectively).

The primary composite outcome was the first occurrence of either myocardial infarction, stroke, acute decompensated heart failure, acute coronary syndrome not resulting in myocardial infarction, or death from cardiovascular causes. Patients assigned to a target of 120 mm Hg or less had fewer primary outcome events than those assigned to a blood pressure of 140 mm Hg (hazard ratio of 0.75).

A nonsignificant reduction in stroke (secondary outcome) was observed in both intervention groups [hazard ratio of 0.89 (CI 0.63-1.25)]. The highest reduction in secondary outcome occurred in death from cardiovascular causes (hazard ratio of 0.57), whereas there was no reduction in acute coronary syndrome in either group. Serious adverse effects (hypotension, syncope, bradycardia, electrolyte imbalance, and acute renal failure) occurred more frequently in patients assigned to a target of less than 120 mm Hg.

In the past few months, the individual patient data were pooled from two randomized controlled blood pressure trials, the Systolic Blood Pressure Intervention Trial (SPRINT) and the Action to Control Cardiovascular Risk in Diabetes Blood Pressure (ACCORD-BP), which assessed the benefit of blood pressure treatment and cardiovascular outcomes (42). Additionally, the primary composite endpoint and the length of follow-up were harmonized between both studies. Both trials assessed the benefit of intensive blood pressure control (systolic blood pressure goal less than 120 mm Hg) versus standard blood pressure control (less than 140 mm Hg) in patients with elevated cardiovascular risk and no diabetes mellitus (SPRINT trial) and patients with type 2 diabetes mellitus (ACCORD-BP trial). The pooled cohort of the studies comprised 14,094 patients (mean age of 66 years; 39% females; 33% with type 2 diabetes mellitus), with a mean systolic blood pressure at baseline of 139 mm Hg (42).

In this pooled analysis, intensive blood pressure lowering resulted in a significant 18% reduction in the primary composite endpoint (unstable angina, myocardial infarction, acute cardiac decompensation, stroke, and cardiovascular death). The stroke rate was reduced, though not significantly, by 15% in patients with intensive blood pressure treatment; the hazard ratio for stroke was not significantly lower in patients with diabetes mellitus (HR 0.79) compared to those with no diabetes mellitus (HR 0.88) in the intensive treatment arm (42).

Clinical trials in hypertensive and normotensive high cardiovascular risk patients.

Heart Outcomes Prevention Evaluation (HOPE) trial. The HOPE study enrolled 9297 patients 55 years old or older with a history of cardiovascular disease or diabetes and at least one other cardiovascular risk factor (hypertension, elevated total cholesterol levels, low high-density lipoprotein cholesterol levels, cigarette smoking, or documented microalbuminuria) (249). HOPE was a 2-by-2 factorial design study, with patients randomized to either ramipril (n = 4645) at a dose of 2.5 mg once-a-day for a week, 5 mg for the next 3 weeks, and then 10 mg, or matching placebo (n = 4652). The mean blood pressure at entry was 139/79 mm/Hg for both the ramipril and placebo groups. In the ramipril group, there was a reported 22% reduction in the combined primary endpoint, a 26% reduction in cardiovascular deaths, a 20% reduction in myocardial infarction, a 32% reduction in stroke, and a 16% reduction in total mortality. Ramipril was seen to lower blood pressure only 3.3 mm/Hg, identical to the placebo group. Therefore, the observed benefit must have included some special properties not related to the effects of anti-hypertension drugs.

European Trial On Reduction of Cardiac Events Among Patients with Stable Coronary Artery Disease (the EUROPA study). The EUROPA study included patients with previous myocardial infarction, angiographic evidence of coronary artery disease, coronary revascularization, or a positive stress test only (81). After a run-in period of 4 weeks, in which all patients received perindopril, 12,218 patients were randomly assigned perindopril 8 mg once daily (n = 6110), or matching placebo (n = 6108). Mean follow-up was 4.2 years, and the primary endpoint was cardiovascular death, myocardial infarction, or cardiac arrest; 603 (10%) placebo and 488 (8%) perindopril patients experienced the primary endpoint, which yields a 20% relative risk reduction (95% CI 9 to 29, p = 0.0003) with perindopril. Stroke was infrequent and the incidence was not different in the two groups (1.6% with perindopril and 1.7% with placebo).

Systolic Hypertension in the Elderly Program (SHEP). The SHEP, a randomized trial that included a total of 4736 patients aged 60 years or older with isolated systolic hypertension who were randomly assigned to 12.5 mg/day of chlorthalidone (with the possibility of adding 25 mg/day of atenolol or 0.05 mg/day of reserpine) or placebo, showed that antihypertensive drug treatment reduced the incidence of both hemorrhagic and ischemic strokes (175). The analysis of the specific ischemic stroke subtypes demonstrated a significant benefit in reducing lacunar strokes (adjusted RR 0.53, 95% CI 0.32 to 0.88), but not significant effect on atherothrombotic and cardioembolic strokes.

A Coronary disease Trial Investigating Outcome with Nifedipine gastrointestinal therapeutic system (ACTION trial). In the ACTION trial, 3825 patients with treated stable symptomatic coronary disease were randomly assigned to double-blind addition of nifedipine (gastrointestinal therapeutic system) 60 mg once daily and 3840 patients, to placebo (180). The primary endpoint was the combination of death, acute myocardial infarction, refractory angina, new overt heart failure, debilitating stroke, and peripheral revascularization. Mean follow-up was 4.9 years. Analysis was by intention to treat; 310 patients allocated nifedipine died (1.64 per 100 patient-years) compared with 291 people allocated placebo (1.53 per 100 patient-years, HR 1.07, 95% CI 0.91 to 1.25). Primary endpoint rates were 4.60 per 100 patient-years for nifedipine and 4.75 per 100 patient-years for placebo (HR 0.97, 95% CI 0.88 to 1.07). With nifedipine, the rate of death and any cardiovascular event or procedures was 9.32 per 100 patient-years versus 10.50 per 100 patient-years for placebo (HR 0.89, 95% CI 0.83 to 0.95). The difference was mainly attributable to a reduction in the need for coronary angiography and interventions in patients assigned nifedipine, despite an increase in peripheral revascularization. The incidence of debilitating stroke was reduced by 22% in the nifedipine group versus placebo.

Summary of the benefit of different antihypertensive drug classes in primary stroke prevention in high-risk patients for vascular disease. A meta-analysis of outcome data showed that beta-blocker-based therapy is associated with a higher risk of stroke versus other antihypertensive treatments (137). In 12 trials, the total number of events was 1650 in 51,963 patients taking a beta blocker as opposed to 1594 events in 53,882 patients taking other antihypertensive drugs (RR 1.16, 95% CI 1.04 to 1.30). These data suggest that beta-blocker-based therapy is much less efficient at reducing the risk of stroke when compared with other treatment options for hypertension (238).

Several meta-analyses have been performed that confirm the differential effects of antihypertensive drug classes. Consecutive overviews of actively controlled trials compared the incidence of stroke between patients randomly assigned to initial treatment with old drugs (diuretics and/or beta-blockers) and those started on newer agents (calcium channel blockers, angiotensin-converting enzyme inhibitors, beta blockers, or angiotensin II type-1 receptor blockers) (209; 210; 208; 231; 227). Compared with older drugs, calcium channel blockers provided significantly better protection against stroke (pooled OR 0.86, 95% CI 0.81 to 0.92), whereas the opposite was the case for angiotensin-converting enzyme inhibitors (pooled OR 1.10, 95% CI 1.01 to 1.20) that reduced the risk of cardiovascular events, but not of stroke (227). Compared with controls, treatment initiated with angiotensin II type-1 (AT1) receptor blockers lessened the incidence of stroke by 26% (95% CI 14% to 36%). Overall, newer compared with older agents provided 7% (95% CI 2% to 11%) better protection against stroke (251). A meta-analysis of 13 major hypertension trials including more than 103,000 subjects and 4040 incident cases of stroke suggests that calcium channel blockers reduce the risk of stroke more effectively than other classes of antihypertensive drugs (15). Another meta-analysis showed that the mean relative risk of an event with an angiotensin II type-1 receptor blocker compared with a comparator was 0.79 (95% CI 0.69 to 0.9) for stroke, 0.96 (95% CI 0.85 to 1.09) for coronary heart disease, 0.83 (95% CI 0.72 to 0.97) for heart failure, and 0.9 (95% CI 0.83 to 0.96) for a major cardiovascular event (37).

An important point is whether the very elderly benefit from antihypertensive treatment. The benefit of lowering blood pressure in the very elderly is supported by the results of the Study on Cognition and Prognosis in the Elderly (SCOPE), which randomly assigned 4964 patients aged 70 to 89 years with mild to moderate hypertension to double-blind treatment with candesartan 8 to 16 mg daily or placebo (170; 218). Open-label antihypertensive therapy (mostly thiazide diuretics) was added as needed to control the blood pressure. Assignment to candesartan was associated with a 27.8% (95% CI 1.3% to 47.2%) relative risk reduction of nonfatal strokes and a nonsignificant reduction of all strokes (RR reduction 23.6%, 95% CI -0.7% to 42.1%) as compared with placebo. This was achieved with a lower mean blood pressure (mean difference 3.2/1.6 mm/Hg) in the active treatment group.

Taken together, these results suggest that in primary prevention there may be marginal cerebrovascular benefits for regimens that include an angiotensin-converting inhibitor, an angiotensin II type-1 receptor blocker, and calcium channel blocker, but the available data are not definitive. Hypertension should be treated in the very elderly to reduce their risk of nonfatal stroke (91; 29; 83).

In a meta-analysis of 123 large-scale randomized controlled blood-pressure-lowering trials between January 1966 and July 2015 comprising more than 600,000 participants, every reduction of 10 mm Hg in systolic blood pressure led to a significant reduction in major cardiovascular events and all-cause mortality. In the 54 studies comparing stroke risk (136,682 participants in the intervention groups and 128,641 participants in control groups), every reduction of 10 mm Hg in systolic blood pressure reduced stroke risk by 27%. The reduction in stroke risk was higher in patients without a history of cardiovascular disease as compared with patients with baseline cardiovascular disease. When the five classes of blood-pressure-lowering drugs were compared, the highest benefit in stroke risk reduction was observed with calcium channel blockers, whereas beta-blockers exhibited the least effectiveness in stroke prevention (77).

Clinical trials in persons at intermediate-cardiovascular risk. The Heart Outcomes Prevention Evaluation (HOPE)-3 Trial is an international multicenter double-blind, randomized, placebo-controlled trial using a 2-by-2 factorial design in 12,705 patients considered to be at intermediate risk of cardiovascular disease (a rough estimate for 10-year adverse cardiovascular event risk is 10%) (143). A combination of blood-pressure-lowering therapy (candesartan at a dose of 16 mg per day plus hydrochlorothiazide at a dose of 12.5 mg per day) and cholesterol-lowering therapy (rosuvastatin at a dose of 10 mg per day) was given to 3180 of these patients; 3181 patients received rosuvastatin therapy plus placebo; 3176 patients were assigned to blood-pressure-lowering therapy (candesartan–hydrochlorothiazide) plus placebo; and 3168 patients were assigned to a dual-placebo group.

The mean age of participants was 65.7 years, 54% were men, and 38% had a history of arterial hypertension. Nearly one-third of the participants were Chinese, 27% were Hispanic, and 20% were Caucasian. At baseline, the mean blood pressure was 138/82 mm Hg and the mean LDL cholesterol level was 128 mg per deciliter (3.3 mmol per liter). The median follow-up was 5.6 years.

A nonsignificant 20% relative risk reduction of fatal and nonfatal stroke was observed in the group receiving blood pressure therapy (candesartan plus hydrochlorothiazide) as compared to placebo (143). The greatest reduction in fatal and nonfatal stroke events occurred in the patient group receiving combined blood-pressure-lowering and cholesterol-lowering therapies as compared to the dual-placebo group; the relative difference in this case was 44%, and the absolute difference was 0.8% (248).

Recommendations of the American Heart Association / American Stroke Association (148).

	• Regular blood pressure screening and appropriate treatment, including both lifestyle modification and pharmacological therapy, are recommended (Class I, Level of Evidence A).
	• Annual screening for high blood pressure and health-promoting lifestyle modification are recommended for patients with prehypertension (systolic blood pressure of 120 to 139 mm Hg or diastolic blood pressure of 80 to 89 mm Hg) (Class I; Level of Evidence A).
	• Patients who have hypertension should be treated with antihypertensive drugs to a target blood pressure of less than 140/90 mm Hg (Class I; Level of Evidence A).
	• Successful reduction of blood pressure is more important in reducing stroke risk than the choice of a specific agent, and treatment should be individualized on the basis of other patient characteristics and medication tolerance (Class I; Level of Evidence A).
	• Self-measured blood pressure monitoring is recommended to improve blood pressure control (Class I; Level of Evidence A).

The latest recommendation of the American College of Cardiology and American Heart Association (ACC/AHA) Guideline on the Primary Prevention of Cardiovascular Disease for adults with confirmed arterial hypertension and a 10-year atherosclerosis cardiovascular disease risk of at least 10% is a blood pressure target lower than 130/80 mm Hg (28). Pharmacological treatment should be coupled with nonpharmacological interventions, including weight loss, a heart-healthy dietary pattern, dietary sodium reduction and potassium supplementation, limited alcohol consumption, and increased physical activity with a structured exercise program (28).

Lipid abnormalities and stroke

Abnormalities of serum lipids [triglyceride, cholesterol, low-density lipoprotein (LDL), and high-density lipoprotein (HDL)] are known risk factors for vascular disease. Cholesterol and LDL have a direct relationship with the incidence of coronary heart disease, whereas HDL has an inverse relationship. Whether increased serum cholesterol levels are a risk factor for stroke remains controversial. A meta-analysis of prospective studies of hypercholesterolemia found a relative stroke risk of 1.2 among those with cholesterol levels higher than 220 mg/dL. Low HDL and elevated triglycerides have been found to increase stroke risk (191). In the Copenhagen City Heart Study, total cholesterol was positively associated with risk of non-hemorrhagic stroke, but only for levels greater than 8 mmol/L (> 309 mg/dL), corresponding to levels in the upper 5% of the study cohort (136). An analysis of cohort studies by stroke subtypes has shown that the risk of non-hemorrhagic stroke increases as LDL concentration increases (a 15% increase for each 1 mml/L increase in LDL concentration). The risk of hemorrhagic stroke increases as LDL concentration falls (a 19% increase for each 1 mmol/L in LDL concentration decrease) (130).

There are several possible explanations for these different results and for the apparent lack of association between cholesterol levels and ischemic stroke. Most studies have selected cohorts of patients to study coronary heart disease, and these patients have been young subjects at high risk for myocardial infarction. Ischemic stroke is a disease that occurs typically in older men and women in comparison to patients with coronary heart disease, and in these studies the patients have a greater risk of dying of myocardial infarction rather than of stroke (240). Moreover, cerebrovascular events were not analyzed according to stroke subtypes (ischemic or hemorrhagic) in most studies. The Multiple Risk Factor Intervention Trial (MRFIT) showed that the risk of death due to ischemic stroke increased with increasing serum cholesterol levels. Conversely, in the same study, there was a negative association between hemorrhagic stroke and lower serum cholesterol levels, suggesting a possible U-shaped relationship between cholesterol and stroke (125). Because of the U-shaped relationship, counting hemorrhagic strokes together with ischemic strokes could have masked a small, but significant relationship between cholesterol levels and ischemic stroke (11). Finally, regarding the etiology of ischemic stroke, cholesterol is an important risk factor for atherosclerosis, but the cause of stroke could be cardioembolic or due to other causes unrelated to cholesterol levels. Unfortunately, however, there are very limited data on the correlation between cholesterol and ischemic stroke subtypes.

In the randomized trials of cholesterol-lowering drugs or diets that were conducted before the introduction of statins, coronary events were reduced by only about 10% to 15% (130). The stroke risk did not appear to be altered in a meta-analysis of these trials (RR 1.0), but the 95% CI (0.8 to 1.6) includes the possibility of a 10% to 15% reduction in stroke (104). Numerous statin trials in patients with known coronary heart disease (21; 195; 25; 237; 08; 103) and primary prevention trials including high-risk populations have demonstrated a decrease in stroke incidence in patients treated with statins (195; 25). It is not known whether these findings might be due to the cholesterol reduction effect of statins or to pleiotropic effects of statins, such as improved endothelial function, decreased platelet aggregability, and reduced vascular inflammation (12; 153; 161). Evidence has shown that statins can stabilize vulnerable carotid plaque or reduce plaque progression (41; 120; 155). Proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors, a new class of lipid-lowering drugs, have shown significant benefit in reducing cardiovascular events, although safety concerns have emerged due to increased adverse neurocognitive events associated with the use of PCSK9 inhibitors compared with placebo. PCSK9 inhibitors were found to stabilize carotid plaque on MR plaque imaging in three asymptomatic patients with mild to severe internal carotid stenosis who showed atherosclerotic plaque progression while on atorvastatin therapy alone (162).

A meta-analysis of 49,008 patients included in four randomized controlled trials of statins as lipid-lowering therapy found that they significantly reduced cases of incident ischemic strokes by 30% in patients without cardiovascular disease. Patients at high cardiovascular risk showed greater risk reduction with statin therapy (150).

Clinical trials with statins and stroke prevention in patients with coronary heart disease and in high-risk patients for vascular disease, mainly diabetics and hypertensives, without coronary heart disease.

The Scandinavian Simvastatin Survival Study (4S). The 4S trial showed that simvastatin 10 to 40 mg/day, given for 6 months after a myocardial infarction or unstable angina in men with serum total cholesterol above 270 mg/dL, lowered mortality by 30% and major coronary events by 34% after 5 years of follow-up (21). Post-hoc analysis showed that the incidence of strokes and transient ischemic attacks was lowered by 30%, but this was mainly caused by the reduction in the number of transient ischemic attacks. When transient ischemic attack was excluded from the analysis, the difference was not significant.

West of Scotland Coronary Prevention (WOSCOP) Study. The WOSCOP double-blinded trial randomized 6595 men (45 to 64 years of age) with hypercholesterolemia and no history of myocardial infarction, and after 4.9 years of follow-up, showed that pravastatin (40 mg per day) reduced the incidence of fatal and nonfatal coronary events by 31%, all cardiovascular deaths by 32%, and death from any cause by 22% in hypercholesterolemic men (272 ± 23 mg/dL on average) (202). There was no significant reduction in the incidence of stroke.

The Cholesterol and Recurrent Event (CARE) study. CARE included patients with myocardial infarction 3 to 20 months before randomization, and it studied pravastatin 40 mg/day versus placebo (195). Patients had cholesterol concentrations within the normal range or slightly high (total cholesterol less than 240 mg/dL and LDL cholesterol between 115 and 174 mg/dL). The study found a 24% relative risk reduction (RRR) of a fatal or nonfatal coronary event in patients treated with pravastatin after 5 years of treatment. A stroke occurred in 3.7% of the patients in the placebo group compared to 2.5% in the pravastatin group (RRR = 31%). A subanalysis found a 27% reduction in stroke or transient ischemic attack, and also that all categories of stroke were reduced (179).

The Long-Term Intervention with Pravastatin in Ischemic Disease (LIPID) Study. The LIPID trial randomized patients with a history of myocardial infarction or hospitalized for unstable angina and initial plasma total cholesterol levels of 155 to 271 mg/dL on treatment with pravastatin 40 mg daily versus placebo (25). After 6.1 years of follow-up, death from coronary heart disease (primary endpoint) occurred in 8.3% of the patients in the placebo group and 6.4% of those in the pravastatin group, a RRR of 24%. The incidences of all pre-specified cardiovascular outcomes (myocardial infarction, death from coronary heart disease or nonfatal myocardial infarction, stroke, and coronary revascularization) were consistently lower among patients receiving pravastatin. For stroke, the RRR was 19% (237).

The Kyushu Lipid Interventional Study (KLIS). The KLIS aimed to investigate the effects of pravastatin in the primary prevention of coronary events and cerebral infarction in Japanese men aged 45 to 74 years with serum total cholesterol 220 mg/dL or greater (26). The coronary events included myocardial infarction, coronary artery surgery and angioplasty, cardiac death, and sudden death. A total of 5640 patients were recruited: 3061 were allocated to the pravastatin (10 to 20 mg per day) treatment and 2579 to conventional treatment. Patients were followed for 5 years on average. The primary endpoint occurred in 2.9% of patients. There was a nonsignificant reduction of 14% in the pravastatin group (RR 0.86, 95% CI 0.61 to 1.20) and a nonsignificant 22% stroke risk reduction (RR 0.78, 95% CI 0.54 to 1.13). Unfortunately, due to unsuccessful randomization, a protocol-based analysis was carried out with adjustment for coronary risk factors at baseline using the Cox proportional hazards model.

Myocardial Ischemia Reduction with Aggressive Cholesterol Lowering (MIRACL). The MIRACL study enrolled 3086 patients hospitalized for acute coronary syndromes (199). Subjects were randomized immediately after admission to 80 mg atorvastatin daily or to placebo for 16 weeks. There was a significant reduction in the risk of the primary outcome: combined death, nonfatal myocardial infarction, resuscitated cardiac arrest, or recurrent symptomatic myocardial ischemia, with new objective evidence requiring emergency rehospitalization (RR 0.84, 95% CI 0.70 to 1.00, p = 0.48). Also, there was a significant risk reduction in stroke (235). Of 31 nonfatal strokes, 22 occurred in the placebo group and nine in the atorvastatin group (OR 0.40, 95% CI 0.19 to 0.88, p = 0.02). Of the 36 fatal and nonfatal strokes, 24 occurred in the patients treated with placebo and 12 in the group treated with atorvastatin (OR 0.49, 95% CI 0.2 to 0.98, p = 0.04) (235).

Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto del Miocardio (GISSI) study. The aim of this study was to test the efficacy of a low-dose pravastatin regimen (20 mg per day) in patients with myocardial infarction (27). A total of 4271 acute myocardial infarction patients (6 months or less) with total blood cholesterol 200 mg/dL or greater were randomized to low-dose pravastatin or no treatment. The publication of results from other trials prompted the Data Safety and Monitoring Board and the Steering Committee to change the protocol and, later, to stop randomization with a decreased statistical power of the study. Mainly because of the on-course modification of the study protocol, 402 of 2133 (18.8%) patients in the control group were given a cholesterol-lowering treatment during follow-up. Conversely, 296 of 2138 (13.8%) patients permanently stopped taking their tablets. The analysis carried out excluding patients randomized to pravastatin treatment, and currently not assuming the drug indicated that the low-dose pravastatin had efficacy (total cholesterol -12.5%, LDL -18.8%, triglycerides -7.9%, HDL +3.4%). During the study, 256 (6.0%) patients either died or had a nonfatal stroke or myocardial infarction: 136 (6.4%) in the control group and 120 (5.6%) in the pravastatin group (RR 0.90, 95% CI 0.71 to 1.15, p = 0.41). One hundred and sixty patients died: 88 (4.1%) in the control group and 72 (3.4%) in the pravastatin group (RR 0.84, 95% CI 0.61 to 1.14, p = 0.26). The reduction of cardiovascular events was more evident in the by-treatment analysis, with coronary heart disease deaths being significantly decreased (RR 0.60, 95% CI 0.38 to 0.96, p = 0.04).

The A to Z Trial. The A to Z trial was a study designed to investigate two issues concerning contemporary care of patients with acute coronary syndromes (35; 69). The first issue was in regards to antithrombotic treatment. The second issue was to compare early initiation of an intensive statin regimen with delayed initiation of a less intensive regimen in patients with acute coronary syndrome. The study was a randomized, double-blind trial that included patients with acute coronary syndromes receiving 40 mg per day of simvastatin for 1 month, followed by 80 mg per day thereafter (n = 2265) compared with acute coronary syndrome patients receiving placebo for 4 months followed by 20 mg per day of simvastatin (n = 2232) (69). Follow-up was at least 6 months and up to 24 months. A total of 343 patients (16.7%) in the placebo plus simvastatin group experienced the primary endpoint (the composite of cardiovascular death, nonfatal myocardial infarction, readmission for acute coronary syndromes, and stroke), compared with 309 (14.4%) in the simvastatin-only group (40/80 mg) (OR 0.89, 95% CI 0.76 to 1.04, p = 0.14). Cardiovascular death occurred in 109 (5.4%) and 83 (4.1%) patients in the two groups (OR 0.75, 95% CI 0.57 to 1.00, p = 0.05), but no differences were observed in other individual components of the primary endpoint, including stroke. No difference was evident during the first 4 months between the groups for the primary endpoint (OR 1.01, 95% CI 0.83 to 1.25, p = 0.89), but from 4 months to the end of the study, the primary endpoint was significantly reduced in the simvastatin-only group (OR 0.75, 95% CI 0.60 to 0.95, p = 0.02).

The Greek Atorvastatin and Coronary Heart Disease Evaluation (GREACE) trial. In this trial, patients with established coronary heart disease and LDL greater than 100 mg/dL were randomized to atorvastatin (10 to 80 mg/day) or usual care (31). After 3 years of follow-up, the composite endpoint (death, nonfatal myocardial infarction, unstable angina, coronary heart failure, revascularization, or stroke) was lowered by 51%, and all components of the primary endpoint were significantly reduced. The incidence of stroke was lowered by 47%. By the end of the follow-up, 26% of patients in the usual care group were receiving some form of lipid-lowering therapy.

Heart Protection Study (HPS). The HPS investigated 20,536 adults aged 40 to 80 years with coronary disease, other occlusive arterial disease, or diabetes. This trial included 3280 patients with stroke before randomization, including 1822 strokes without established coronary heart disease. Subjects were randomly assigned to 40 mg simvastatin daily or placebo for 5 years. All-cause mortality was significantly reduced, mainly because of a highly significant reduction in the coronary death rate (103). Stroke rate was also significantly reduced by 25% [444 strokes in the simvastatin group (4.3%) versus 585 strokes in the placebo group (5.5%), OR 0.75, 95% CI 0.66 to 0.79, p < 0.0001]. The reduction in stroke was mainly the result of a reduction in ischemic stroke, and similar numbers in each treatment group had a hemorrhagic stroke. In addition, there was a significant reduction in the number of patients who, although not having a stroke during follow-up, had at least one episode of transient ischemic attack [204 (2.0%) versus 250 (2.4%), p = 0.02].

Pravastatin in Elderly Individuals at Risk of Vascular Disease (PROSPER). PROSPER was a randomized controlled trial in which 5804 men (n = 2804) and women (n = 3000) aged 70 to 82 years with a history of either risk factors (62% had hypertension, 11% diabetes, and 28% smoked) or vascular disease (44% had cardiovascular disease and 11% had stroke before randomization) were assigned to pravastatin (40 mg per day, n = 2891) or placebo (n = 2931) (201). Follow-up was 3.2 years on average, and the primary endpoint was a composite of coronary death, nonfatal myocardial infarction, and fatal or nonfatal stroke. Pravastatin lowered LDL cholesterol concentrations by 34% and reduced the incidence of the primary endpoint to 408 events, compared with 473 on placebo (OR 0.85, 95% CI 0.74 to 0.97, p = 0.014). Coronary heart disease death and nonfatal myocardial infarction risk was also reduced (OR 0.81, 95%CI 0.69 to 0.94, p = 0.006). Stroke risk was unaffected (OR 1.03, 95% CI 0.81 to 1.31), but the hazard ratio for transient ischemic attack was 0.75 (95% CI 0.55 to 1.00). PROSPER confirmed that statins could be used in elderly patients as well in younger patients to prevent any cardiovascular events but did not confirm a positive effect on the incidence of stroke in this population. Caution should be taken in interpreting these results for several reasons: (1) a lack of effect on stroke might be due to a lack of statistical power, (2) a too-short trial duration, and (3) the dose of pravastatin used was likely low. A higher dose of 40 mg per day might have produced more benefit (215; 13).

Antihypertensive and Lipid-Lowering treatment to prevent Heart Attack Trial (ALLHAT-LLT). ALLHAT-LLT enrolled 10,355 patients (52) and was part of the large-scale ALLHAT study of high-risk hypertensive patients in the primary care setting. Eligibility for ALLHAT-LLT was based on LDL cholesterol levels in patients aged over 55 years already enrolled in the main ALLHAT study. Patients with either an LDL cholesterol concentration of 120 to 189 mg/dL or 100 to 129 mg/dL, with coronary heart disease and a triglyceride concentration of less than 350 mg/dL were randomly assigned to receive either pravastatin 40 mg per day or usual care consisting of measures, at the discretion of the primary care physician, to reduce LDL cholesterol. These measures could include statins or other lipid-lowering drugs. LDL cholesterol levels in the pravastatin group were reduced by 28% from baseline, whereas the reduction in the usual care group was 11%. Accordingly, differences in cardiovascular endpoints were small and failed to reach statistical significance. After 4.8 years of follow-up, the relative risk of all-cause mortality was 0.99 (95% CI 0.89 to 1.11), and the relative risk of coronary heart disease death and nonfatal myocardial infarction was 0.91 (95% CI 0.79 to 1.04). There was no significant difference even in stroke incidence (4.07% in the pravastatin group and 4.5% in the usual care group, RR 0.91, 95% CI 0.75 to 1.09, p = 0.31).

The Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT). In the ASCOT, 19,342 hypertensive patients (systolic blood pressure > 160 or diastolic blood pressure > 100 mm/Hg) who also had at least three risk factors among ventricular hypertrophy, ECG abnormalities, non-insulin-dependent diabetes mellitus, peripheral vascular disease, transient ischemic attacks, male gender and age over 55 years, microalbuminuria, smoking, total cholesterol/HDL cholesterol ratio greater than 6, and early coronary heart disease were randomized to beta blockers with or without diuretics or amlodipine with or without angiotensin-converting enzyme inhibitors (197). The patients who had a total cholesterol concentration of less than 6.5 mmol/dL were offered random assignment in a factorial design to either atorvastatin 10 mg per day or placebo. A total of 10,297 patients were assigned to the lipid group, and the follow-up was scheduled to be 5 years. The independent Data and Safety Monitoring Board committee of the study recommended stopping the lipid group early because of the highly evident efficacy of the atorvastatin group on the primary endpoint. Stroke reduction was 27%, and it was obtained in addition to the 40% stroke reduction found in a population of patients who were well controlled for their hypertension after the blood pressure goal (< 140/90 mm/Hg) was achieved (197; 12).

Collaborative Atorvastatin Diabetes Study (CARDS). In the CARDS, 2838 patients aged 40 to 75 years were randomized to placebo (n = 1410) or atorvastatin 10 mg daily (n = 1428) (61). Study entrants had no documented previous history of cardiovascular disease, an LDL cholesterol concentration of 4.14 mmol/L or lower, a fasting triglyceride amount of 6.78 mmol/L or less, and at least one of the following: retinopathy, albuminuria, current smoking, or hypertension. The primary endpoint was time to first occurrence of the following: acute coronary heart disease events, coronary revascularization, or stroke. The trial was terminated 2 years earlier than expected because the pre-specified early stopping rule for efficacy had been met. Median duration of follow-up was 3.9 years, and 127 patients allocated placebo (2.6 per 100 person-years at risk) and 83 allocated atorvastatin (1.54 per 100 person-years at risk) had at least one major cardiovascular event (RR 37%, 95% CI -52 to -17, p = 0.001). Treatment would be expected to prevent at least 37 major vascular events per 1000 such people treated for 4 years. Assessed separately, acute coronary heart disease events were reduced by 36% (-55 to -9), coronary revascularization by 31% (-59 to -16), and rate of stroke by 48% (-69 to -11). Atorvastatin reduced the death rate by 27% (-48 to 1, p = 0.059). No excess of adverse events was noted in the atorvastatin group.

Aggressive Lipid-Lowering Initiation Abates New Cardiac Events (ALLIANCE) study. The 4-year ALLIANCE trial was a population-based, randomized, open-label trial involving 2443 patients aged 18 to 75 years (121). The study assessed the effects of aggressive lipid lowering in the prevention of ischemic events in patients with hypercholesterolemia (LDL cholesterol level 130 to 250 mg/dL if not on lipid-lowering therapy or 110 to 200 mg/dL if on therapy) and established coronary artery disease (prior myocardial infarction, coronary artery bypass grafting or other coronary intervention, or unstable angina). Patients were randomized to usual care or atorvastatin 10 to 80 mg per day to achieve LDL levels of 80 mg/dL or less. The primary outcome measure was occurrence of an ischemic event, defined as cardiovascular death, nonfatal myocardial infarction, resuscitated cardiac arrest, unstable angina, or coronary revascularization, including coronary artery bypass grafting. A total of 289 (23.7%) patients in the atorvastatin group compared with 333 (27.7%) patients in the usual care group experienced a primary outcome (OR 0.83, 95% CI 0.71 to 0.97, p = 0.02). This reduction in morbidity was largely due to fewer nonfatal myocardial infarctions (4.3% vs. 7.7%, p = 0.0002). Levels of LDL were more reduced (34.3% vs. 23.3%, p < 0.0001) in patients receiving atorvastatin compared with those receiving usual care. Stroke was a secondary endpoint, and it was experienced by 2.9% of the patients treated with atorvastatin compared to 3.2% of the patients in the usual care group (OR 0.87, 95% CI 0.55 to 1.38, p = 0.552).

Pravastatin or Atorvastatin Evaluation and Infection Therapy – Thrombolysis in Myocardial Infarction (TIMI) 22 study. The TIMI 22 study enrolled 4162 patients who had been hospitalized for acute coronary syndrome within the preceding 10 days (45). The study compared 40 mg of pravastatin daily with 80 mg of atorvastatin daily. The primary endpoint was a composite of death from any cause, myocardial infarction, documented unstable angina requiring rehospitalization, revascularization (performed at least 30 days after randomization), or stroke. After a mean follow-up of 24 months, the rates of the primary endpoint were 26.3% in the pravastatin group and 22.4% in the atorvastatin group, reflecting a 16% reduction in the hazard ratio in favor of atorvastatin (95% CI 5% to 26%, p = 0.005). Stroke was infrequent, but the rates did not differ significantly between the two groups.

Treating to New Targets (TNT) study. In the TNT study, a total of 10,001 patients with clinically evident coronary heart disease and LDL cholesterol levels of less than 130 mg per deciliter were randomly assigned to a double-blinded therapy and received either 10 mg or 80 mg of atorvastatin per day (126). Patients were followed for a median of 4.9 years. The primary endpoint was the occurrence of a first major cardiovascular event, defined as death from coronary heart disease, nonfatal non-procedure-related myocardial infarction, resuscitation after cardiac arrest, or fatal or nonfatal stroke. The mean LDL cholesterol levels were 77 mg per deciliter during treatment with 80 mg of atorvastatin and 101 mg per deciliter during treatment with 10 mg of atorvastatin. A primary event occurred in 434 patients (8.7%) receiving 80 mg of atorvastatin, compared with 548 patients (10.9%) receiving 10 mg of atorvastatin, representing an absolute reduction in the rate of major cardiovascular events of 2.2% and a 22% relative reduction in risk (OR 0.78, 95% CI 0.69 to 0.89, p < 0.001). As compared to patients given 10 mg of atorvastatin, patients given 80 mg of atorvastatin also had significant reductions in the risk of fatal or nonfatal stroke (OR 0.75, 95% CI 0.59 to 0.96, p = 0.02). There was no difference between the two treatment groups in overall mortality.

Incremental Decrease in End Point through Aggressive Lipid Lowering (IDEAL) study. The IDEAL study was a prospective, randomized, open-label, blinded endpoint evaluation trial with a median follow-up of 4.8 years, which enrolled 8888 patients aged 80 years or younger with a history of acute myocardial infarction (173). Patients were randomly assigned to receive a high dose of atorvastatin (80 mg per day, n = 4439), or a usual dose of simvastatin (20 mg per day, n = 4449). The primary endpoint was the occurrence of a major coronary event, defined as coronary death, confirmed nonfatal acute myocardial infarction, or cardiac arrest with resuscitation. A major coronary event occurred in 463 simvastatin patients (10.4%) and in 411 atorvastatin patients (9.3%) (OR 0.89, 95% CI 0.78 to 1.01, p = 0.07). Nonfatal acute myocardial infarction occurred in 321 (7.2%) and 267 (6.0%) patients in the two groups (OR 0.83, 95% CI 0.71 to 0.98, p = 0.02), but no differences were seen in the other two components of the primary endpoint. Death from any cause occurred in 374 patients (8.4%) in the simvastatin group and in 366 (8.2%) patients in the atorvastatin group (OR 0.98, 95% CI 0.85 to 1.13, p = 0.81). Fatal or nonfatal stroke occurred in 174 patients (3.9%) treated with simvastatin and in 151 patients (3.4%) treated with atorvastatin (OR 0.87, 95% CI 0.70 to 1.08, p = 0.20).

MEGA study. In this prospective, randomized, open-labeled blinded study, patients with hypercholesterolemia (> 156 mg/dL) and no history of coronary heart disease or stroke were randomly assigned to a diet alone or diet plus 10 to 20 mg pravastatin daily group; 3966 patients were randomly assigned to the diet alone group and 3866 to the diet plus pravastatin group (157). Mean follow-up was 5.3 years. Coronary heart disease was significantly lower in the diet plus pravastatin group than in the diet alone group (66 vs. 101 events, HR 0.67, 95% CI 0.49 to 0.91, p = 0.01). Treatment with pravastatin was associated with a lower incidence of stroke than diet alone, although this difference was not significant (HR 0.83, 95% CI 0.57 to 1.21, p = 0.33).

Atorvastatin Study for Prevention of Coronary Heart Disease Endpoints in Non-Insulin-Dependent Diabetes Mellitus (ASPEN) study. In this double-blind, parallel group study, a total of 2410 subjects with type 2 diabetes was randomly assigned to receive 10 mg of atorvastatin or placebo in a 4-year period (119). The composite primary endpoint comprised cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, recanalization, coronary artery bypass surgery, resuscitated cardiac arrest, and worsening or unstable angina requiring hospitalization. When atorvastatin was compared to placebo, composite primary endpoint rates were 13.7% and 15.0%, respectively (HR 0.90, 95% CI 0.73 to 1.12, p = ns). There was no significant reduction in the incidence of stroke.

Study of the Effectiveness of Additional Reductions in Cholesterol and Homocysteine (SEARCH) study. In this study, 12,064 myocardial infarction survivors were randomized to more versus less intensive cholesterol-lowering treatment using simvastatin 80 mg versus 20 mg daily (200). After 6 years of median follow-up, the annual overall rate of major vascular events was approximately 3%. There was no significant reduction in the incidence of stroke.

Justification for the Use of Statins in Prevention: an Interventional Trial Evaluating Rosuvastatin (Jupiter) study. In this study, 17,802 apparently healthy men and women with LDL-cholesterol levels of less than 130 mg/dL and high-sensitivity C-reactive protein levels of 2.0 mg/L or higher were randomly assigned to rosuvastatin 20 mg daily or placebo (185). The composite primary endpoint comprised myocardial infarction, stroke, arterial revascularization, hospitalization for unstable angina, or death from cardiovascular causes. The trial was stopped after a median follow-up of 1.9 years. The rates of the primary endpoint were 0.77 and 1.36 per 100 person-years of follow-up in the rosuvastatin and placebo groups, respectively (HR 0.56, 95% CI 0.46 to 0.69, p = 0.002). The rates of stroke were 0.18 and 0.34 per 100 person-years of follow-up in the rosuvastatin and placebo groups, respectively (HR 0.52, 95% CI 0.34 to 0.79, p = 0.002).

Summary of the benefit of statins in stroke prevention in patients with coronary heart disease and in high-risk patients for vascular disease. The Cholesterol Treatment Trialist’s (CTT) collaborators reported on the results of a prospective meta-analysis of data from 90,056 individuals in 14 randomized trials of statins: 42,131 (47%) had preexisting coronary heart disease, 21,575 (24%) were women, 18,686 (21%) had a history of diabetes, and 49,689 (55%) had a history of hypertension (32). During a mean of 5 years, there were 8186 deaths, 14,348 individuals had major vascular events, and 5103 individuals developed cancer. There was a 12% proportional reduction in all-cause mortality per mmol/L reduction in LDL cholesterol (RR 0.88, 95% CI 0.84 to 0.91, p < 0.0001). This reflected a 19% reduction in coronary mortality (RR 0.81, 95% CI 0.76 to 0.85, p < 0.0001) and nonsignificant reductions in both noncoronary vascular mortality (RR 0.93, 95% CI 0.83 to 1.03, p = 0.2) and nonvascular mortality (RR 0.95, 95% CI 0.90 to 1.01, p = 0.1). There were corresponding reductions in myocardial infarction or coronary death (RR 0.77, 95% CI 0.74 to 0.80, p < 0.0001), in the need for coronary revascularization (RR 0.76, 95% CI 0.73 to 0.80, p < 0.0001), in fatal or nonfatal stroke (RR 0.83, 95% CI 0.78 to 0.88, p < 0.0001), and combining these, of 21% in any such major vascular event (RR 0.79, 95%CI 0.77 to 0.81, p < 0.0001) (32). Regarding cerebrovascular events, data were available on a total of 2957 first strokes after randomization. There were 2282 strokes among 65,138 patients in nine trials that sought information on stroke type, of which 204 (9%) were attributed definitely to hemorrhage, 1565 (69%) were confirmed to be ischemic, and 513 (22%) were of unknown type. Overall, there was a significant 17% proportional reduction in the incidence of first stroke of any type [1340 (3.0%) statin versus 1617 (3.7%) control: RR 0.83, 95% CI 0.78 to 0.88, p < 0.0001] per mmol/L LDL cholesterol. This overall reduction in stroke reflected a highly significant 19% proportional reduction (RR 0.81, 99% CI 0.74 to 0.89, p < 0.0001) in stroke not attributed to hemorrhage (ie, presumed ischemic) per mmol/L LDL cholesterol reduction, and no apparent difference in hemorrhagic stroke. The overall reduction in presumed ischemic stroke reflected a highly significant 22% proportional reduction in confirmed ischemic stroke (RR 0.78, 99% CI 0.70 to 0.87, p < 0.0001) per mmol/L LDL cholesterol reduction and a 12% proportional reduction in stroke of unknown type (RR 0.88, 99% CI 0.7 to 1.02, p = 0.03). There was no significant reduction in stroke during the first year after randomization (RR 0.96, 99% CI 0.79 to 1.17, p = 0.6), but there were significant reductions of about 20% to 25% during each of the subsequent 3 years and favorable trends thereafter. During an average of 5 years of treatment, the reduction in the overall incidence of stroke was about one fifth per mmol/L LDL cholesterol decrease, meaning that eight fewer participants have stroke per 1000 among those with preexisting coronary artery disease at baseline, compared with five fewer per 1000 among participants with no such history (32).

Another meta-analysis of randomized trials of statins in combination with other preventive strategies, including 165,792 individuals, shows that each 1 mmol/L (39 mg/dL) decrease in LDL cholesterol equates to a 21.1% reduction in relative risk for stroke (95% CI 6.3 to 33.5, p = 0.009). Incidence of all strokes was reduced by 18% (95% CI 13 to 23, p < 0.0001). Incidence of fatal stroke was reduced by 13% (95% CI -3 to 27, p = 0.10), although this did not reach statistical significance without evidence of heterogeneity across the trials (10).

In the Three-City Study, a noninstitutionalized cohort of 7484 people in France (mean age of 74 years, 63% women) with no known history of vascular disease was followed for an average of 9.1 years (09). At baseline, 27.3% of participants reported using lipid-lowering drugs (13.5% statins and 13.8% fibrates); the most common statins used were simvastatin and pravastatin, and the most common fibrate was fenofibrate. During the follow-up, 292 stroke events occurred, confirmed either by CT scan or brain MRI (227 infarcts, 57 hemorrhages, and the remainder indeterminate). Increased risk of stroke was associated with older age, male sex, high blood pressure, diabetes, and high body mass index, whereas no association with baseline blood lipid concentrations was found. A 34% decrease in stroke risk (hazard ratio of 0.66, 0.49 to 0.90) was observed in lipid-lowering users compared to nonusers. There were no differences in stroke risk reduction between statin users (hazard ratio of 0.68, 0.45 to 1.01) and fibrate users (hazard ratio of 0.66, 0.44 to 0.98). On the other hand, a higher risk reduction occurred in lipid-lowering users with ischemic stroke (hazard ratio of 0.63, 0.45 to 0.84) in comparison to hemorrhagic stroke (hazard ratio of 0.72, 0.37 to 1.42).

The beneficial effect of statins on prevention of major cardiovascular events at different ages has been confirmed in another meta-analysis (57). This meta-analysis comprised data from 28 large statin trials among 186,854 participants. Patients were divided into six age groups (55 years or younger, 56-60 years, 61-65 years, 66-70 years, 71-75 years, and older than 75 years). During 4.9 years of follow-up, significant cardiovascular risk reduction by statin therapy occurred in all age groups, including individuals older than 75 years. Overall, a significant 24% relative risk reduction of major coronary events and a 16% relative risk reduction in any stroke was observed with statin therapy. Nevertheless, in the primary prevention setting the evidence was less extensive in people over 75 years (57).

Recommendations of the American Heart Association / American Stroke Association (148).

	• In addition to therapeutic lifestyle changes, treatment with an HMG coenzyme-A reductase inhibitor (statin) medication is recommended for the primary prevention of ischemic stroke in patients estimated to have a high 10-year risk for cardiovascular events (Class I, Level of Evidence A).
	• Niacin may be considered for patients with low HDL cholesterol or elevated Lp(a), but its efficacy in preventing ischemic stroke in patients with these conditions is not established. Caution should be used with niacin because it increases the risk of myopathy (Class IIb; Level of Evidence B).
	• Fibric acid derivatives may be considered for patients with hypertriglyceridemia, but their efficacy in preventing ischemic stroke is not established (Class IIb; Level of Evidence C).
	• Treatment with nonstatin lipid-lowering therapies such as fibric acid derivatives, bile acid sequestrants, niacin, and ezetimibe may be considered in patients who cannot tolerate statins, but their efficacy in preventing stroke is not established (Class IIb; Level of Evidence C). Treatment of adults with diabetes with a statin, especially those with additional risk factors, is recommended to lower risk of a first stroke (Class I, Level of Evidence A).

The 2019 American College of Cardiology and American Heart Association (ACC/AHA) Guideline on the Primary Prevention of Cardiovascular Disease in patients with high blood cholesterol levels also recommends moderate-intensity statin therapy for patients with a 10-year risk of atherosclerosis cardiovascular disease ranging from 7.5% to 20% (intermediate-risk patients) and for diabetic patients aged 40 to 75 years regardless of estimated 10-year atherosclerosis cardiovascular disease risk (class of recommendation: I; level of evidence: A) (28). The 2019 ACC/AHA guideline recommends a low-density lipoprotein cholesterol reduction of at least 30% to 50% or more in intermediate- and high-risk patients (higher than 20% 10-year atherosclerosis cardiovascular disease risk) for optimal prevention (28).

According to the latest ACC and AHA lipid guidelines, high statin therapy is the mainstay lipid-lowering therapy, particularly for adult patients with LDL-cholesterol levels of at least 190 mg/dL. In such cases, the target reduction is at least 50% from the LDL-cholesterol baseline level. Ezetimibe may be added to the maximum tolerated statin dose when LDL-cholesterol remains over 100 mg/dL. Moreover, proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors may be considered for patients with high cardiovascular risk and LDL-cholesterol levels that remain higher than 100 mg/dL despite combined statin and ezetimibe therapies (184).

Other risk factors, clinical trials, evidence-based recommendations

Diabetes mellitus and stroke. The American Diabetes Association recommends rigorous comprehensive control of blood sugar levels for adherent patients with type 1 or type 2 diabetes to prevent microvascular complications (75). In addition, strict blood pressure control in patients with type 2 diabetes significantly lowers stroke risk (148). In a large observational nationwide population-based cohort study involving 406,721 patients with diabetes mellitus and 2,086,440 control subjects followed over a mean of 7.3 years, patients with diabetes mellitus were found to have a 54% higher risk of first-ever stroke (ischemic or hemorrhagic) compared to control subjects (250). In this study, nonoptimal glycemic control was directly associated with higher frequency of comorbid stroke risk factors (arterial hypertension, reduced kidney function, and higher blood cholesterol levels) and with a 2-fold higher risk of stroke in patients with nonoptimal glycemic levels compared to control subjects (250).

An HbA1c level lower than 7% is a desirable goal in most adult patients with type 2 diabetes when achieved safely by proper pharmacological treatment along with lifestyle optimization (75; 147). However, this target should be individualized, taking into account age, life expectancy, disease duration, comorbidities, and hypoglycemia risk and its consequences. Metformin, a biguanide class drug that reduces peripheral insulin resistance and concurrently decreases intestinal glucose absorption, has also shown pleiotropic effects and is currently recommended as the first glucose-lowering agent of choice to prevent cardiovascular disease, including stroke, in noncontraindicated adult patients with type 2 diabetes mellitus (28). When optimal glycemic control is not achieved, a glucagon-like peptide-1 receptor agonist may be added to metformin in patients older than 55 years with high cardiovascular disease risk (carotid, coronary, or lower-limb artery stenosis greater than 50%, left ventricular hypertrophy, or nephropathy).

Li and colleagues conducted a metaanalysis comparing the effect on stroke risk of three new antihyperglycemic therapies (134). This metaanalysis included 19 randomized controlled trials: eight from glucagon-like peptide-4 agonists, five from sodium-glucose cotranporter-2 inhibitors, and six from dipeptidyl peptidase-4 inhibitors.

The survey involved 155,027 patients with type 2 diabetes mellitus with a median of diabetes duration ranging from 9.2 to 15.8 years, a median of HbA1c ranging from 7.3% to 8.91%, and a median follow-up ranging from 1.3 to 10.5 years. Most patients had no history of previous stroke. Glucagon-like peptide-1 (GLP-1) agonists significantly reduced the risk of nonfatal stroke by 15% and total stroke by 16% compared to placebo. Neither sodium-glucose cotransporters-2 inhibitors nor glucagon-like peptidase-4 inhibitors reduced stroke occurrence (134). GLP-1 agonists may have neuroprotective and atherosclerotic properties by decreasing the production of proinflammatory factors, promoting antiapoptotic mechanisms, reducing oxidative stress, and limiting the formation and accumulation of advanced glycation end-products (90).

Addition of lipid-lowering therapy is advisable as primary prevention of cardiovascular disease in adults with type 2 diabetes mellitus. A moderate-intensity statin therapy is suggested in individuals between 40 and 75 years of age with type 2 diabetes and no atherosclerotic cardiovascular disease whereas a high-intensity statin therapy should be prescribed in individuals with type 2 diabetes and high atherosclerotic cardiovascular disease risk or those aged 50 to 70 years. Adding ezetimibe to the maximum tolerated dose of statin therapy is advisable in adult patients with a higher 10-year atherosclerosis cardiovascular disease risk (over 20%), the goal being to lower low-density lipoprotein cholesterol by at least 50% (76).

A study assessed the effect of low doses of aspirin in primary cardiovascular disease prevention in Caucasian patients over 40 years of age with any type of diabetes (30). In this trial, 7740 patients were randomly assigned to 100 mg of aspirin per day and compared with 7740 patients assigned to placebo. The mean age in both groups was 63 years, 62% were male, nearly 45% of patients had a body mass index greater than 30, 55% of patients had had diabetes for more than 9 years, and 60% of patients had a moderate-to-high vascular risk score. After a mean follow-up of 7.4 years, the aspirin group showed a nonsignificant 12% reduction in nonfatal ischemic stroke and a nonsignificant 15% reduction in transient ischemic attack occurrence. However, patients assigned to the aspirin group exhibited a significantly higher risk of any major bleeding, mostly from the gastrointestinal tract. The rate of intracranial hemorrhages was no different in either group.

Current recommendations concerning acetylsalicylic acid use (75 to 162 mg per day) as primary prevention of cardiovascular events, including stroke, in patients with type 2 diabetes mellitus and a high cardiovascular risk require that these patients be made aware of the benefits versus the bleeding risks of this strategy (76).

The latest U.S. Preventive Services Task Force (USPSTF) statement reports that aspirin use provides small net benefits in the primary prevention of both nonfatal myocardial infarction and ischemic stroke in individuals aged 40 to 59 years, with a 10-year risk of 10% or more (225). This information is based on an analysis of 13 randomized clinical trials involving 161,680 participants. Therefore, aspirin use (75-100 mg/daily) in individuals aged between 40 and 59 years should be individualized. Evidence indicates that aspirin use in those older than 60 years is associated with a high risk of bleeding complications.

Cigarette smoking and stroke. In 2015, it was estimated that 25% of men and about 5% of women worldwide were daily smokers, despite the well-known harm of tobacco use to cardiovascular health (87). To date, a body of convincing evidence confirms the epidemiologic association between cigarette smoking and all types of stroke (56; 229). Compared to never-smokers and those who have quit in the more distant past, the risk of stroke is two to four times higher in current and recent smokers (160; 168; 229). The risk of stroke increases at a rate of 12% for each increment of five cigarettes smoked per day (168). Moreover, smoking was found to be responsible for 10.9% of PAR, making it the leading preventable stroke risk factor (160).

It is well established that quitting smoking has profound benefits for current and long-term health at any age, even among heavy and lifelong smokers. In fact, the Nurses’ Health Study and the Framingham Study both demonstrated that the risk of ischemic stroke is reduced to that of non-smokers after 2 and 5 years, respectively (241; 116).

Recommendations of the American Heart Association / American Stroke Association (148).

	• Counseling, in combination with drug therapy using nicotine replacement, bupropion, or varenicline, is recommended for active smokers to assist in quitting smoking (Class I; Level of Evidence A).
	• On the basis of epidemiological studies, which show a consistent and overwhelming relationship between smoking and both ischemic stroke and subarachnoid hemorrhage, abstention from cigarette smoking is recommended for patients who have never smoked (Class I; Level of Evidence B).

Alcohol and stroke. In 2016, one third of the world’s population were current drinkers (intake of one or more drinks in the past 12 months) and the mean number of standard alcoholic drinks consumed daily was 1.7 for men and 0.73 for women (88). According to WHO’s “Global status report on alcohol and health 2018,” the most commonly recorded form of alcohol consumption worldwide is spirits, followed by beer and wine. In the next few years, the total per capita alcohol consumption in persons aged 15 years and older is projected to increase mainly in the WHO regions of the Americas, South-East Asia, and the Western Pacific (243). Evidence of the benefits and consequences of different alcohol consumption patterns on cardiovascular health comes mainly from observational studies. The available evidence suggests a dose-dependent relationship (J-shaped curve) between alcohol intake and stroke risk (58). Thus, compared with abstention, low or moderate habitual alcohol consumption (one or fewer drinks/day for women and one to two drinks/day for men) may be associated with a lower risk of ischemic stroke (128; 58). Conversely, heavy drinking (more than four drinks/day) is associated with an increased risk of all types of stroke, especially subarachnoid and intracerebral hemorrhages (128). The relationship between alcohol intake and stroke is complex, however, and some uncertainties still persist (113). Although the beneficial effects of low and moderate alcohol consumption on ischemic stroke risk may diminish over time, there is little or no doubt that the elimination of heavy drinking will reduce the incidence of stroke (246).

Recommendations of the American Heart Association / American Stroke Association (148).

	• Reduction or elimination of alcohol consumption in heavy drinkers through established screening and counseling strategies as described in the U.S. Preventive Services Task Force update is recommended (Class I; Level of Evidence A) (224).
	• For individuals who choose to drink alcohol, consumption of less than two drinks per day for men and less than one drink per day for nonpregnant women might be reasonable (Class IIb; Level of Evidence B).

Physical activity and stroke. Physical activity is of clear benefit in reducing the risk of stroke (107). Studies have also evaluated the association between physical activity and the risk of stroke (234; 94; 01; 80; 117; 89; 192; 131). The Honolulu Heart Program, which investigated older middle-aged men of Japanese ancestry, showed a protective effect of habitual physical activity from thromboembolic stroke only among non-smokers. The Framingham Study showed the benefits of combined leisure and work physical activities for men, but not for women. In the Oslo Study, increased leisure physical activity was related to a reduced incidence of stroke among men aged 40 to 49. For women aged 40 to 65, the Nurses’ Health Study showed an inverse association between physical activity and the incidence of any stroke. In the Northern Manhattan Stroke Study, the benefits of leisure-time physical activity were noted for all ages, sexes, and race-ethnic subgroups. The optimal amount of exercise needed to prevent stroke is unclear, particularly for the elderly. Among subjects who were free of cardiovascular disease in a case-controlled study in West Birmingham, UK, vigorous exercise was no more protective than walking. Among the older cohort of the Framingham Study, the strongest protection was detected in the medium tertile physical activity subgroup, with no benefit from additional activity. The protective effect of physical activity may be partly mediated through its role in controlling risk factors such as hypertension, diabetes, dyslipidemia, obesity, excessive alcohol consumption, and tobacco use (107). In addition, physical activity may improve endothelial function, blood rheology, and vessel collateralization (182).

Recommendations of the American Heart Association / American Stroke Association (148).

	• Physical activity is recommended because it is associated with a reduction in the risk of stroke (Class I; Level of Evidence B).
	• Healthy adults should perform at least moderate to vigorous intensity aerobic physical activity at least 40 minutes 3 to 4 days a week (Class I; Level of Evidence B).

The latest guidelines of the American College of Cardiology and American Heart Association Guideline on the Primary Prevention of Cardiovascular Disease counsel that adults should accumulate at least 150 minutes of moderate-intensity aerobic activity or 75 minutes of vigorous-intensity aerobic activity weekly to reduce the risk of atherosclerosis cardiovascular disease, including stroke (28).

Dietary factors, obesity, and stroke. Poor quality diet is an independent stroke risk factor that contributes to other stroke risk factors as well (149). Healthful nutrition is, therefore, important for stroke prevention. Diet is considered 1 of 10 potentially modifiable stroke risk factors (160). Specific foods and certain dietary patterns, rather than isolated single nutrients, have been found to be protectively associated with stroke incidence (96) A diet rich in fruits and vegetables can reduce the risk of stroke (109; 54; 259; 258). In addition to lowering blood pressure and improving lipid profile, these foods may protect against stroke through their antioxidant mechanisms (109; 258).

To date, there is no conclusive evidence showing that dietary supplementation with antioxidants such as vitamin C, vitamin D, beta carotenes, and selenium reduces stroke incidence (96; 109; 110; 247). A meta-analysis showed that vitamin E supplementation significantly reduces the incidence of ischemic stroke but not hemorrhagic stroke, whereas other evidence has shown that both higher dietary and supplemented vitamin E intakes are associated with an increased risk of hemorrhagic stroke (109; 141).

In a meta-analysis of six cohort studies and 16 randomized controlled trials, neither a daily dietary intake of 200 to 1500 mg of calcium nor supplementation of calcium, alone or combined with vitamin D, increased the risk of stroke (245). In the last few years, a meta-analysis has assessed the association between magnesium intake and stroke and found that compared with the lowest intake category, the highest intake group showed a significant 12% reduction in the risk of ischemic stroke, although this association was not significant for intracerebral and subarachnoid hemorrhage (256). Moreover, in a large case-control genetic study of ischemic stroke, high serum magnesium concentrations were associated with a low cardioembolic stroke risk (127). The effect on this ischemic stroke subtype may be due to antiarrhythmic effects and the antithrombotic properties of this essential body mineral (127). High fish intake reduces the incidence of stroke, although omega-3 polyunsaturated fatty acid supplementation has not shown any beneficial effects in reducing stroke risk (96; 222).

Mediterranean style diet was inversely associated with fatal cardiovascular disease, myocardial infarction, stroke, and pulmonary embolism after 12 years of follow-up (106; 78). A DASH (dietary approaches to stop hypertension) eating pattern (rich intake of vegetables, fruits, whole grains, and low-fat dairy along with a low intake of red meat, saturated fat, and salt) has also been associated with reduced risk of stroke and is highly recommended (109; 54). Vegetarian diet seems to reduce the risk of both ischemic and hemorrhagic stroke. Vegetarians included in two prospective cohorts had about half the risk of stroke as nonvegetarians (55).

Conversely, high consumption of red meat was related to higher all-cause mortality, and a significant association with processed meat intake was also observed for cardiovascular diseases (187; 253).

Dietary salt consumption is closely associated with the level of blood pressure; stricter salt reduction more markedly decreased blood pressure. Salt reduction is essential for the prevention and treatment of hypertension, and for the prevention of cardiovascular diseases (14; 196).

In a meta-analysis of dietary patterns and stroke risk that comprised 21 trials published from 1991 to 2014, a significant reduction in stroke risk was observed (OR 0.77, CI 0.63-0.93) in individuals with the healthiest dietary pattern (high intake of vegetables, fruit, fish, low-fat milk, and whole grains) compared to individuals with a lower intake of these items (254). Moreover, individuals with a high intake of all kinds of red or processed meats, refined grains, sweets, desserts, high-fat dairy products, and high-fat gravy (the Western-style dietary pattern) did not show a greater risk compared to individuals with the lowest intake of the same dietary pattern.

Recommendations of the American Heart Association / American Stroke Association (148).

	• Reduced intake of sodium and increased intake of potassium as indicated in the U.S. Dietary Guidelines for Americans are recommended to lower blood pressure (Class I; Level of Evidence A).
	• A diet that is rich in fruits and vegetables, and thereby high in potassium, is beneficial and may lower the risk of stroke (Class I; Level of Evidence B).
	• A Mediterranean diet supplemented with nuts may be considered in lowering the risk of stroke (Class IIa; Level of Evidence B).
	• A DASH-style diet, which emphasizes fruits, vegetables, and low-fat dairy products and reduced saturated fat, is recommended to lower blood pressure (Class I; Level of Evidence A).

Hyperhomocysteinemia and stroke. Another important dietary component is homocysteine (a sulfhydryl-containing amino acid), which is an important intermediate product of methionine and cysteine metabolism. Both acquired and genetic factors can have an impact on total plasma homocysteine levels. Plasma homocysteine levels are associated in a dose-dependent manner with ischemic stroke, particularly in young and middle-aged individuals (102). This association was found mainly for infarcts due to large atherosclerotic disease and small vessel disease (176; 112). In addition, in a meta-analysis elevated plasma homocysteine levels were also associated with increased hemorrhagic stroke risk (257). Homocysteine levels are inversely related to aging, smoking, alcohol intake, physical activity, and food supplements, principally folate and vitamin B12. In the past, increased dietary or supplemental intake of folates and vitamins B6 and B12 to reduce stroke incidence has yielded no conclusive results (217). In the last few years, meta-analysis evidence has suggested that folic acid supplementation may reduce the incidence of first stroke, particularly in populations with a high prevalence of hypertension and low baseline folate levels (255).

Recommendations of the American Heart Association / American Stroke Association (148).

• The use of the B complex vitamins, cobalamin (B12), pyridoxine (B6), and folic acid might be considered for the prevention of ischemic stroke in patients with hyperhomocysteinemia, but its effectiveness is not well established (Class IIb; Level of Evidence B).

Estrogens and stroke. In past decades, when oral contraceptive pills contained higher estrogen doses, studies showed increased ischemic stroke among women, particularly those over 35 years with additional stroke risk factors such as smoking and migraine. Modern combined oral contraceptive pills have a lower dose of estrogen and a new generation of progestins, potentially reducing stroke risk. A meta-analysis of 15 case-control and three cohort observational studies showed that oral contraceptive pill users had a 2-and-a-half-fold higher risk of first cerebral infarct compared with noncurrent users (244). This meta-analysis confirmed several facts: the risk of first ischemic stroke was even higher in current combined contraceptive users when other stroke risk factors (hypertension, smoking, and migraine with aura) were present; there was a dose-dependent relationship between estrogen dose and stroke risk; and compared with other progestin generations, the fourth generation had the lowest risk of first ischemic stroke (244). An updated meta-analysis of 12 case-control and six cohort studies showed a dose-dependent relationship between estrogen dose and risk of total stroke as well as of ischemic and hemorrhagic stroke in current combined oral contraceptive pill users. Specifically, a 20% increase in the risk of total stroke and first ischemic stroke occurred with each 10 ug increase in estrogen dosage and a 10% increase occurred in risk for hemorrhagic stroke (133). Moreover, this metaanalysis also showed that a longer duration of oral contraceptive pill use significantly contributed to an increased risk of total stroke and cerebral infarct, with a 20% increase in risk for every 5 years of oral contraceptive pill use (133).

Women with a short reproductive lifespan (the time between the ages at menarche and menopause) have been linked to a greater stroke risk. In a 20-year prospective study, women with a reproductive life span of less than 34 years had almost twice the risk of first stroke as those with a 38- to 40-year span (152).

Evidence regarding stroke risk and postmenopausal hormone replacement therapy has shown that the number of years between menopause and hormone replacement therapy initiation had a strong impact on stroke incidence. Pooled data from five population-based cohort studies with a total of 88,914 female participants found that women who began taking hormone replacement therapy in the first 5 years of menopause experienced no change in the stroke-free period compared with women who did not receive hormone replacement therapy. Likewise, commencement of hormone replacement therapy more than 5 years after the onset of menopause increased the risk of ischemic and hemorrhagic stroke (47). In addition, data from five national registries (totaling 980,003 female patients) showed that oral hormone replacement therapy posed an increased ischemic stroke risk whereas the transdermal mode of delivery did not. Interestingly, a decreased risk of stroke was found with vaginal estrogen use (142).

Based on the currently available evidence, hormone replacement therapy should not be used for primary prevention of stroke in postmenopausal women (224).

Recommendations of the American Heart Association / American Stroke Association (148).

	• Smoking cessation should be strongly recommended in women with migraine headaches with aura (Class I; Level of Evidence B).
	• Alternatives to oral contraceptives, especially those containing estrogen, might be considered in women with active migraine headaches with aura (Class IIb; Level of Evidence B).
	• Treatments to reduce migraine frequency might be reasonable to reduce the risk of stroke (Class IIb; Level of Evidence C).
	• Closure of patent foramen ovale is not indicated for preventing stroke in patients with migraine (Class III; Level of Evidence B).

Sickle cell disease and stroke. Children with sickle cell disease have a 300-fold higher risk of cerebrovascular complications, which may manifest as clinically silent cerebral infarcts, transient ischemic attacks, or overt ischemic and hemorrhagic strokes (05). Silent and overt ischemic strokes are common in children below the age of 10 and in adults over the age of 30. In contrast, intracranial hemorrhages (intraparenchymal, subdural, subarachnoid, intraventricular, or in combination) occur most commonly in young adults during the second decade of life and may account for up to a third of cerebrovascular complications in patients with sickle cell disease (144).

The latest American Society of Hematology guidelines recommend annual transcranial Doppler ultrasonography screening for children between 2 and 16 years of age who have phenotype hemoglobin SS or hemoglobin S-beta thalassemia (68). For such children with a high stroke risk on transcranial Doppler ultrasonography who live in high-income settings where the best treatment is available, American Society of Hematology recommends regular red cell transfusion for at least a year. The target is a decrease in hemoglobin S levels to below 30% and hemoglobin levels of more than 9.0 g/dL (68).

In low- and middle-income settings, where regular blood transfusion therapy and chelation therapy are less feasible, the recommended therapy for children aged 2 to 16 years with phenotype hemoglobin SS, hemoglobin S-beta thalassemia, or compound heterozygous sickle cell disease who have a transcranial Doppler ultrasonography-indicated high stroke risk is a minimum fixed daily dose of 20 mg/kg of hydroxyurea or the maximum tolerated dose per day (68).

Recommendations of the American Heart Association / American Stroke Association (148).

	• Transcranial Doppler screening for children with sickle cell disease is indicated starting at 2 years of age and continuing annually to 16 years of age (Class I; Level of Evidence B).
	• Transfusion therapy (target reduction of hemoglobin S, < 30%) is effective for reducing stroke risk in those children at elevated risk (Class I; Level of Evidence B).
	• Although the optimal screening interval has not been established, it is reasonable for younger children and those with borderline abnormal transcranial Doppler velocities to be screened more frequently to detect the development of high-risk transcranial Doppler indications for intervention (Class IIa; Level of Evidence B).
	• Pending further studies, continued transfusion, even in those whose transcranial Doppler velocities revert to normal, is probably indicated (Class IIa; Level of Evidence B).
	• In children at high risk for stroke who are unable or unwilling to be treated with periodic red cell transfusion, it might be reasonable to consider hydroxyurea or bone marrow transplantation (Class IIb; Level of Evidence B).
	• MRI and MRA criteria for selection of children for primary stroke prevention with transfusion have not been established, and these tests are not recommended in place of transcranial Doppler for this purpose (Class III; Level of Evidence B).
	• Transfusion therapy (target reduction of hemoglobin S from a baseline of > 90% to < 30%) is effective for reducing stroke risk in those children at elevated stroke risk (Class I, Level of Evidence B).

Migraine and stroke. A considerable amount of evidence has shown that migraine increases the risk of stroke, an association that is often described in young women suffering migraines with aura (49; 108; 04; 124; 146). Tzourio and colleagues found that stroke was strongly associated with migraine, both without aura (OR 3.0, 95% CI 1.5 to 5.8) and with aura (OR 6.2; 95% CI 2.1 to 1.80) (221). In another study, a history of migraine was more frequent in stroke patients than in controls (OR 1.9, 95% CI 1.1 to 1.3) (46). In the prospectively designed subgroup analyses, a history of migraine reached the highest odds ratio (3.7, 95% CI 1.5 to 9.0) and was the only significant risk factor in women below 35 years of age (46). In agreement with previous studies (219; 135), ischemic stroke was associated with both migraine with aura (OR 3.81, 95% CI 1.26 to 1.15) and migraine without aura (OR 2.97, 95% CI 0.66 to 1.35).

Available data suggest that stroke risk is higher in women migraineurs who are oral contraceptive users. The risk of ischemic stroke was substantially increased by the use of oral contraceptives (OR 13.9), for which a dose-effect relationship between risk of stroke and dose of estrogen was found (pills containing 50 µg estrogen: OR 4.8; pills with 30 to 40 µg: OR 2.7; pills with 20 µg: OR 1.7; pills with progesterone: OR 1) (221). Although the evidence is limited, stroke risk seems not to be significantly increased by today's combined hormonal contraceptives, many of which contain low estrogen concentrations (≤ 35 μg ethinyl estradiol). A nested case-control study using data from U.S. databases identified 25,887 ischemic strokes among women aged 15 to 49 years between 2006 and 2012. Compared with women with no migraines and no combined hormonal contraceptives use, the OR for ischemic stroke in women who had migraines with aura but did not use combined hormonal contraceptives was 2.7 (95% CI 1.9-3.7) but rose to 6.1 (95% CI 3.1-12.1) in those who did (48). Moreover, the OR for ischemic stroke was 1.8 (95% CI 1.1-2.9) among women who had migraines without aura and also used combined hormonal contraceptives, compared with women who did not have migraines and did not use combined hormonal contraceptives (48).

In a systematic review assessing the effect of hormonal contraceptives on women with migraines, the estimated absolute risk ratio of ischemic stroke per year was 6-fold higher in young women with migraine with aura who used hormonal contraception (25.4/100,000 per year) compared with their counterparts who did not use contraceptives (4/100,000 per year) (194).

The risk of ischemic stroke was also elevated in heavy smokers (≥20 cigarettes per day: OR 10.2). In migrainous women, coexistent use of oral contraceptives, history of high blood pressure, and smoking habit had greater than multiplicative effects on the odds ratio for ischemic stroke (OR 34.4, 95% CI 3.27 to 3.61). However, this dramatic increase was based on only nine cases and two controls (49).

The higher the frequency of migraine attacks, the higher the stroke risk. Moreover, higher stroke risk has been found in individuals with active migraine (at least one attack in the last 12 months) but not in those with severe migraine attacks (163).

Multiple hypotheses have been put forth to explain the mechanism by which migraine may increase the risk of stroke. The exact mechanism remains unclear and may be multifactorial. Migraine and stroke share genetic factors and common comorbidities (patent foramen ovale, atrial fibrillation, cervical arterial dissection, and antiphospholipid antibody syndrome) (74; 216; 92). Endothelial dysfunction and coagulation abnormalities are also common to both pathologies (169). Migraine-related cortical spreading depression may reduce cerebral blood flow, thus, potentially triggering ischemic stroke. Finally, vasoconstriction induced by migraine-specific drugs (triptans and ergot alkaloids) can theoretically increase the risk of ischemic stroke (97; 07; 138; 51; 163).

The risk of stroke in young migrainous women seems to apply not only to ischemic stroke but also to hemorrhagic stroke, although hemorrhagic stroke has not been extensively studied and the results are less consistent (220; 193; 07).

Recommendations of the American Heart Association / American Stroke Association (148).

	• Smoking cessation should be strongly recommended in women with migraine headaches with aura (Class I; Level of Evidence B).
	• Alternatives to oral contraceptives, especially those containing estrogen, might be considered in women with active migraine headaches with aura (Class IIb; Level of Evidence B).
	• Treatments to reduce migraine frequency might be reasonable to reduce the risk of stroke (Class IIb; Level of Evidence C).
	• Closure of patent foramen ovale is not indicated for preventing stroke in patients with migraine (Class III; Level of Evidence B).

Asymptomatic carotid artery disease and stroke. The role of prophylactic endarterectomy in people with asymptomatic extracranial carotid artery stenosis is controversial. A few randomized controlled trials have addressed this issue; the largest was the Asymptomatic Carotid Atherosclerosis Study (ACAS) (23). Patients (n = 1662) were randomized to surgery plus best medical therapy or to best medical therapy without carotid endarterectomy. Angiographic complications occurred in 1.2%, and the perioperative stroke risk was 2.3%. After a median follow-up of 2.7 years, the study was stopped early because a significant benefit of surgery was found. The rate of ipsilateral stroke, any perioperative stroke, or death in surgically treated patients was estimated at 5% over 5 years, whereas in medically treated patients the rate was 11% (55% risk reduction, p = 0.004). The benefit was most notable among men, and there was no relationship between benefit and the degree of carotid artery stenosis. It was also clear from the absolute risks that any perioperative complication rates greater than the 2.3% for stroke or death reported in the ACAS trial would eliminate the potential benefit of the operation. Another trial, the Asymptomatic Carotid Surgery Trial (ACST), concluded that in subjects younger than 75 years with severe carotid stenosis, carotid endarterectomy approximately halves the net 5-year risk of stroke. Stroke risk at 5 years in the immediate surgery group was 3.8%, but 11% in the deferred surgery group. The immediate, 30-day risk of stroke or death after endarterectomy was 3.1%. This does not mean that all patients should undergo surgery, but that careful selection is required, and the risk of surgery has to be below 3% to be beneficial (95).

Nonsurgical treatment of carotid disease with angioplasty and stent placement through endovascular techniques is becoming more widespread. It is now a technically feasible option and has been used in patients who are not good surgical candidates, either because of the location of the stenosis or because of a high risk for undergoing anesthesia. A meta-analysis of all the available data on long-term outcome in randomized trials of endovascular treatment versus endarterectomy for symptomatic carotid stenosis showed a significantly worse outcome after endovascular treatment (188).

In a meta-analysis of five randomized controlled trials comparing the efficacy and safety of carotid endarterectomy versus carotid artery stenting among 3019 individuals with asymptomatic carotid stenosis (92% of whom had stenosis of at least 70%), carotid artery stenting resulted in a borderline statistically significant high risk of any periprocedural stroke (1.84 relative risk), periprocedural nondisabling stroke (1.95 relative risk), and periprocedural stroke or death (1.72 relative risk) compared to patients who underwent carotid endarterectomy (154). Concerning the long-term risk of ipsilateral stroke, no relevant differences were observed between these carotid revascularization procedures (154).

Best medical management in patients with asymptomatic carotid stenosis should also consider administering antiplatelet therapy. However, there is very little evidence that antiplatelet therapy is beneficial in preventing stroke or the progression of stenosis in asymptomatic patients. Specifically, one randomized trial of aspirin versus placebo in patients with asymptomatic carotid artery stenosis failed to show any significant long-term protective effect of aspirin therapy. Additionally, a meta-analysis of four primary prevention trials in “high risk” patients (some included patients with asymptomatic carotid stenosis) has suggested that the risk-benefit ratio favors antiplatelet therapy (usually monotherapy with aspirin), if the annual baseline risk of myocardial infarction exceeds 1.5% per year. For this reason, the U.S. Preventive Services Task Force has recommended daily aspirin as cardiovascular prevention in patients with anticipated cardiac morbidity of 3%. The current AHA guidelines for the primary prevention of cardiovascular disease and stroke, in line with the U.S. Preventive Services Task Force, suggests the use of aspirin in persons at high risk, but it recommends a greater than or equal to 10% risk per 10 years rather than a greater than 3% risk over 5 years to improve the likelihood of a positive balance of coronary risk reduction over bleeding and hemorrhagic stroke, which can be caused by aspirin. There is no evidence to suggest that antiplatelet agents, other than aspirin, can improve benefit in asymptomatic patients with carotid stenosis (167).

Recommendations of the American Heart Association / American Stroke Association (148).

	• It is reasonable to consider performing carotid endarterectomy in asymptomatic patients who have greater than 70% stenosis of the internal carotid artery, if the risk of perioperative stroke, MI, and death is low (< 3%). However, its effectiveness compared with contemporary best medical management alone is not well established (Class IIa; Level of Evidence A).
	• Prophylactic carotid artery stenting might be considered in highly selected patients with asymptomatic carotid stenosis (minimum, 60% by angiography, 70% by validated Doppler ultrasound), but its effectiveness compared with medical therapy alone in this situation is not well established (Class IIb; Level of Evidence B).
	• In asymptomatic patients at high risk of complications for carotid revascularization by either carotid endarterectomy or carotid artery stenting, the effectiveness of revascularization vs. medical therapy alone is not well established (Class IIb; Level of Evidence B).
	• Based on moderate-quality evidence, the latest European guidelines recommend carotid endarterectomy for patients with asymptomatic carotid stenosis of 60% or more and who are considered to be at high risk of stroke despite optimal medical therapy (39). Currently, there are several ongoing trials investigating the benefit of carotid revascularization versus the best medical therapy alone.

Atrial fibrillation and stroke. Atrial fibrillation is the most common sustained cardiac rhythm disorder in adults. It is currently estimated that both atrial fibrillation incidence and prevalence nearly doubled globally since the 1990s (67). These estimates are unfortunately expected to increase over the next 30 years. The global burden of atrial fibrillation in 2017 was estimated at around 37.6 million individuals, with 3.05 million new cases diagnosed annually. The incidence and prevalence of atrial fibrillation increase with advancing age and are greater in men than in women (232). For both sexes worldwide in 2017, peak incident and prevalent atrial fibrillation case counts were recorded at ages 65 to 69 and 75 to 79, respectively (67). About one third of the atrial fibrillation population is asymptomatic, meaning that atrial fibrillation prevalence is most likely underestimated (71). Epidemiological data showed that atrial fibrillation incidence and prevalence in Asians and Blacks are lower than in Whites (186; 252). The overall burden of atrial fibrillation seems to be greater in countries with a high socioeconomic index, although in countries with a middle and low socioeconomic index trends are also tending toward increasing burdens (67; 140; 252).

Age is the most important risk factor for atrial fibrillation. After adjustment for age and predisposing conditions, men have a 1.5-fold greater risk of developing atrial fibrillation than women. Hypertension, diabetes mellitus, hyperthyroidism, alcohol abuse, obesity, sleep apnea, and chronic kidney disease are independent cardiovascular risk factors for atrial fibrillation (73; 129; 59). In addition, after adjusting for cardiovascular risk factors, heart failure, valvular heart disease, and myocardial infarction have been shown to increase the risk of atrial fibrillation (59). Emerging risk factors for atrial fibrillation include reduced vascular compliance, atherosclerosis, insulin resistance, environmental factors, inflammation, increased level of natriuretic peptides, and genetic predisposition (115; 59).

Atrial fibrillation is the most important independent risk factor for ischemic stroke. Overall, nonvalvular atrial fibrillation is associated with an approximate 5-fold increased risk of stroke (242). Indeed, one in every four ischemic strokes occurs in patients with atrial fibrillation. Multivariate analysis revealed age, hypertension, diabetes mellitus, prior stroke or transient ischemic attack, myocardial infarction, and congestive heart failure as significant additional risk factors for stroke in patients with atrial fibrillation (22).

Several clinical and observational studies found that the incidence of ischemic stroke in patients with paroxysmal atrial fibrillation was similar to that in patients with permanent atrial fibrillation (101; 158). Nevertheless, studies have demonstrated that thromboembolism was significantly higher in patients with persistent or permanent atrial fibrillation than in those with paroxysmal atrial fibrillation (226; 213; 183). Short asymptomatic episodes of atrial fibrillation detected by cardiac implantable electronic devices are termed atrial high-rate episodes or subclinical atrial fibrillation. A 35% detection rate for subclinical atrial fibrillation has been reported in patients with cardiac implantable electronic devices within the first few years after device implant. A meta-analysis of seven studies comprising 15,353 patients demonstrated that subclinical atrial fibrillation was associated with a 2.4-fold higher risk of stroke (145). Furthermore, patients with a high subclinical atrial fibrillation burden showed a greater risk of ischemic stroke than those with a lower burden (171).

Atrial fibrillation undoubtedly constitutes a major public health problem entailing a substantial burden to patients, social health, and the health economy because of its associated cardiovascular morbidity and mortality. A multidisciplinary approach targeting the adoption of healthy lifestyle choices and optimal risk factor control is a cornerstone of stroke prevention in atrial fibrillation patients. Besides rhythm and rate control strategies, anticoagulation is indicated to prevent stroke in most atrial fibrillation patients. Currently, oral anticoagulation, mainly with direct oral anticoagulants, is recommended for stroke prevention in patients with nonvalvular atrial fibrillation and increased thromboembolic risk (CHA₂DS₂-VASc score of 2 or higher), whereas no anticoagulation is recommended for patients with a CHA₂DS₂-VASc score of 0 (50). Due to a lack of randomized controlled trials, no clear recommendations as to optimal therapy have been issued for atrial fibrillation individuals with a CHA₂DS₂-VASc of 1, an annual thromboembolic risk of 0.6% to 1.3%, and a bleeding risk between 1.02 to 1.51. In this setting, therapeutic decisions to reduce stroke risk must be weighed against the bleeding risk. If oral anticoagulation treatment is chosen, direct oral anticoagulants with a higher clinical benefit should be preferred over warfarin (211).

Vitamin K antagonists (VKAs) with an international normalized ratio (INR) between 2.0 and 3.0 are effective for stroke prevention in patients with atrial fibrillation. A meta-analysis of six randomized trials showed that VKAs provided a 65% risk reduction of ischemic stroke in comparison to placebo (98). A similar risk reduction is seen in patients who receive VKAs for secondary prophylaxis (100). Aspirin provides a 22% reduction in the incidence of stroke versus placebo (98). A meta-analysis of five randomized trials showed that warfarin provided a 36% risk reduction for all strokes and a 46% risk reduction for ischemic strokes versus aspirin (98). VKAs prevent more severe and disabling strokes compared to aspirin (98).

In the ACTIVE W trial, warfarin was significantly more effective than aspirin plus clopidogrel for stroke prevention in patients at high risk of stroke (03). Among patients with atrial fibrillation for whom VKA therapy was considered unsuitable, the combination of clopidogrel and aspirin was associated with a reduction in the primary outcome of stroke, myocardial infarction, non-cerebral systemic embolism, or death from vascular causes compared to aspirin alone. The difference was primarily due to a reduction in the rate of stroke. Major bleeding was significantly more common in patients assigned to the combination of clopidogrel and aspirin (02).

VKAs are associated with an increased risk of bleeding, particularly intracranial hemorrhage and gastrointestinal bleeding, with respect to aspirin and no treatment (0.3%, 0.2%, and 0.1% per year, respectively) (98). VKAs also increase the odds of major extracranial hemorrhage.

In a landmark meta-analysis that included pivotal phase 3 clinical trials for stroke prevention in nonvalvular atrial fibrillation, 42,411 patients were randomly assigned to direct oral anticoagulants and 29,272 were assigned to warfarin. Compared with warfarin, direct oral anticoagulants significantly reduced the risk of systemic embolism or stroke, intracranial hemorrhage, and all-cause mortality by 19%, 52%, and 10%, respectively, although there was an increased risk of gastrointestinal bleeding (190). A systematic review of anticoagulation in atrial fibrillation individuals aged over 65 years found that direct oral anticoagulants reduced the risk of hemorrhagic strokes and intracranial bleeding compared with vitamin K antagonists but increased gastrointestinal bleeding. In addition, direct oral anticoagulants led to a nonsignificant reduction in mortality risk (205).

Clinical guidelines on atrial fibrillation consistently recommend long-term oral warfarin to prevent stroke in valvular atrial fibrillation patients. In a meta-analysis, direct oral anticoagulants were found to be at least as safe and effective as warfarin in reducing stroke risk in atrial fibrillation patients with mechanical and bioprosthetic valves (70). The results of six randomized controlled trials comprising 13,850 valvular atrial fibrillation patients included in this meta-analysis showed that compared with warfarin, direct oral anticoagulants significantly decreased the incidence of stroke or systemic embolism by 22% and intracranial hemorrhage by 49% while also lowering major bleeding risk (70). The evidence is still limited, however, and more studies are needed across a wide range of valvular atrial fibrillation patients.

Identifying patients with atrial fibrillation at risk of stroke to guide antithrombotic prophylaxis. The risk of stroke varies widely among patients with atrial fibrillation. Various risk stratification schemes initially based on clinical factors have been proposed to assess stroke risk in patients with atrial fibrillation (22; 99; 233; 03; 139). In the last few years, risk stratification schemes have incorporated clinical factors and other, previously unconsidered potential risk factors such as electrocardiographic markers, serum biomarkers, and structural and functional markers as a means to improve stroke risk stratification in patients with atrial fibrillation (177; 204; 105; 82). Together with classical clinical factors, these new atrial fibrillation markers comprise complex risk scores and may provide more accurate atrial fibrillation risk assessment, although at the expense of being more demanding and less practical for timely decision making in everyday clinical practice.

In CHADS2, a cumulative score (range 0-6) is calculated according to the presence of defined risk factors. Different risk factors are given different weightings. Two points are assigned for a previous stroke or transient ischemic attack, and 1 point is assigned for each of the following: age older than 75 years, hypertension, diabetes mellitus, and recent cardiac failure. Scores of 0, 1, and higher than 2 denote low, moderate, and moderate-to-high risk of stroke, respectively (85; 84).

CHA2DS2-VASc was developed with the aim of more accurately predicting stroke risk for patients with atrial fibrillation by taking into account some of the additional risk factors not recognized by CHADS2 (139). Like CHADS2, a cumulative scoring system is used. However, the scoring for age is stratified, and younger age (≥65 years) is recognized as a risk factor whereas CHADS2 only recognizes patients over the age of 75 years. In addition, female gender and vascular disease are included in the evaluation of stroke risk, whereas these risk factors are not recognized by CHADS2. CHA2DS2-VASc was validated in the Euro Heart Survey: prediction of stroke risk is improved compared with CHADS2 and only a small proportion of patients are categorized as being at “intermediate risk” of stroke.

Currently, the guidelines of the American Heart Association/American College of Cardiology/Heart Rhythm Society (AHA/ACC/HRS), the American College of Chest Physicians (ACCP), the Asia Pacific Heart Rhythm Society (APHRS), and the European Society of Cardiology (ESC) recommend using the CHA2DS2-VASc score for stroke risk assessment in patients with atrial fibrillation (50).

To minimize stroke and bleeding risks, determining bleeding risk is critical to proper atrial fibrillation patient management when considering oral anticoagulants. To date, there are six validated risk scores for assessing bleeding in patients with atrial fibrillation (86; 203; 178; 79; 159; 105). Current guidelines recommend using HAS-BLED to assess bleeding risk in atrial fibrillation patients. In a systematic review comparing the HAS-BLED, HEMORR2HAGES, ATRIA, and ABC-Bleeding scores, HAS-BLED was found to be the best for predicting major bleeding, although based on modest evidence (40).

Atrial fibrillation patients’ thromboembolic and bleeding risk profiles may both change over time, particularly for those considered to be at low risk. As patients with atrial fibrillation age, more risk factors and comorbidities may appear, resulting in an increase in either or both of the CHA2DS2-VASc and HAS-BLED scores, necessitating regular reassessment of both these risk scores in patients with atrial fibrillation (72; 50).

New oral anticoagulants in development for stroke prevention in atrial fibrillation. The ideal profile of a new oral anticoagulant includes the following: a predictable pharmacological profile so that INR monitoring and dose modifications are not required, rapid onset and offset of action, and fixed, oral dosing that would be most convenient for patients and could potentially improve adherence to the prescribed regimen.

Oral direct thrombin inhibitors. The SPORTIF trials (164; 06) showed that ximelagatran was at least as effective as warfarin for the prevention of stroke, with no difference in rates of major bleeding and a lower rate of total bleeding. Ximelagatran was withdrawn from the market in 2006 due to liver toxicity. Nevertheless, this drug provided the proof of concept for direct thrombin inhibition and showed that oral anticoagulation is possible without regular INR monitoring.

In the landmark phase III RE-LY trial, dabigatran was the first oral anticoagulant to show superior efficacy to warfarin for stroke prevention in atrial fibrillation (02). Dabigatran 110 mg twice daily (bid) resulted in a rate of stroke and systemic embolism similar to that in patients treated with warfarin (1.53% per year vs. 1.69% per year, p < 0.001 for non-inferiority), but with a lower rate of major hemorrhage (2.71 per year vs. 3.36 per year, p = 0.003) (63). Dabigatran 150 mg bid resulted in a lower rate of stroke and systemic embolism than warfarin (1.11% per year vs. 1.69% per year, p < 0.001 for superiority) and a similar rate of major hemorrhage to patients treated with warfarin (3.11% per year vs. 3.36% per year, p=0.31). Dabigatran was associated with higher rates of treatment discontinuation, myocardial infarction, major GI bleeding, and dyspepsia.

Oral direct factor Xa inhibitors. Several randomized phase III trials explored the use of oral direct factor Xa inhibitors for stroke prevention in patients with atrial fibrillation. Rivaroxaban and apixaban have provided cumulative evidence in favor of these novel anticoagulants. Their main advantages, apart from their treatment efficacy, include the reduced rate of intracranial hemorrhage, the lack of need for routine coagulation monitoring, the predictable anticoagulation response, and the limited interaction with food and drugs.

In the double-blind ROCKET trial, 14,264 patients with nonvalvular atrial fibrillation who were at increased risk for stroke were randomly assigned to receive either rivaroxaban (at a daily dose of 20 mg) or dose-adjusted warfarin (172). In the primary analysis, the primary endpoint (stroke or systemic embolism) occurred in 188 patients in the rivaroxaban group (1.7% per year) and in 241 in the warfarin group (2.2% per year) (HR in the rivaroxaban group 0.79, 95% CI 0.66 to 0.96, p < 0.001 for non-inferiority). In the intention-to-treat analysis, the primary endpoint occurred in 269 patients in the rivaroxaban group (2.1% per year) and in 306 patients in the warfarin group (2.4% per year) (HR 0.88, 95% CI 0.74 to 1.03, p < 0.001 for non-inferiority, p = 0.12 for superiority). Major and non-major clinically relevant bleeding occurred in 1475 patients in the rivaroxaban group (14.9% per year) and in 1449 in the warfarin group (14.5% per year) (HR 1.03, 95% CI 0.96 to 1.11, p = 0.44), with significant reductions in intracranial hemorrhage (0.5% vs. 0.7%, p = 0.02) and fatal bleeding (0.2% vs. 0.5%, p = 0.003) in the rivaroxaban group.

A randomized, double-blind trial (ARISTOTLE) compared apixaban (at a dose of 5 mg twice daily) with warfarin (target INR 2.0 to 3.0) in 18,201 patients with atrial fibrillation and at least one additional risk factor for stroke (93). The primary outcome was ischemic or hemorrhagic stroke or systemic embolism. The trial was designed to test for non-inferiority, with key secondary objectives of testing for superiority with respect to the primary outcome and to the rates of major bleeding and death from any cause. The median duration of follow-up was 1.8 years. The rate of the primary outcome was 1.27% per year in the apixaban group, as compared with 1.60% per year in the warfarin group (HR with apixaban 0.79, 95% CI 0.66 to 0.95, p < 0.001 for non-inferiority, p = 0.01 for superiority). The rate of major bleeding was 2.13% per year in the apixaban group, as compared with 3.09% per year in the warfarin group (HR 0.69, 95% CI 0.60 to 0.80, p < 0.001), and the rates of death from any cause were 3.52% and 3.94%, respectively (HR 0.89, 95% CI 0.80 to 0.99, p = 0.047). The rate of hemorrhagic stroke was 0.24% per year in the apixaban group, as compared with 0.47% per year in the warfarin group (HR 0.51, 95% CI 0.35 to 0.75, p < 0.001), and the rate of ischemic or uncertain type of stroke was 0.97% per year in the apixaban group and 1.05% per year in the warfarin group (HR 0.92, 95% CI 0.74 to 1.13, p = 0.42).

Recommendations of the American Heart Association / American Stroke Association (148).

	• For patients with valvular atrial fibrillation at high risk for stroke, defined as a CHA2DS2-VASc score of greater than or equal to 2 and acceptably low risk for hemorrhagic complications, long-term oral anticoagulant therapy with warfarin at a target INR of 2.0 to 3.0 is recommended (Class I; Level of Evidence A).
	• For patients with nonvalvular atrial fibrillation, a CHA2DS2-VASc score of greater than or equal to 2, and an acceptably low risk for hemorrhagic complications, oral anticoagulants are recommended (Class I). Options include warfarin (INR, 2.0 to 3.0) (Level of Evidence A), dabigatran (Level of Evidence B), apixaban (Level of Evidence B), and rivaroxaban (Level of Evidence B). The selection of antithrombotic agent should be individualized on the basis of patient risk factors (particularly risk for intracranial hemorrhage), cost, tolerability, patient preference, potential for drug interactions, and other clinical characteristics, including the time that the INR is in therapeutic range for patients taking warfarin.
	• For patients with nonvalvular atrial fibrillation and CHA2DS2-VASc score of 0, it is reasonable to omit antithrombotic therapy (Class IIa; Level of Evidence B).
	• For patients with nonvalvular atrial fibrillation, a CHA2DS2-VASc score of 1, and an acceptably low risk for hemorrhagic complication, no antithrombotic therapy, anticoagulant therapy, or aspirin therapy may be considered (Class IIb; Level of Evidence C). The selection of antithrombotic agent should be individualized on the basis of patient risk factors (particularly risk for intracranial hemorrhage), cost, tolerability, patient preference, potential for drug interactions, and other clinical characteristics, including the time that the INR is in the therapeutic range for patients taking warfarin.

The latest AHA/ACC/HRS guidelines recommend oral anticoagulation for primary thromboembolism prevention in patients with atrial fibrillation who have an elevated CHA2DS2-VASc score (2 or higher in men, 3 or higher in women). Warfarin has an “A” level of evidence, whereas dabigatran, rivaroxaban, and apixaban have a “B” level, and edoxaban has a lower quality of evidence compared with other non-vitamin K oral anticoagulants (NOACs). In addition, based on the results of four randomized controlled trials comparing NOACs with warfarin in atrial fibrillation patients, the guidelines recommend anticoagulation with NOACs over warfarin when there is no contraindication for their use (111).

Conclusion

Stroke prevention is a crucial issue given that stroke is a frequent and severe disorder and that acute stroke therapies that are effective at the individual level have only a limited impact on public health. Prevention of stroke and transient ischemic attack includes both conventional approaches to vascular risk factor management (blood pressure lowering, cholesterol reduction with statins, smoking cessation, and antiplatelet therapy) and more specific interventions, such as carotid revascularization or anticoagulation for atrial fibrillation.