2024 Guideline for the Primary Prevention of Stroke: A Guideline From the American Heart Association/American Stroke Association

AF (covered in the 2019 American College of Cardiology [ACC]/AHA AF focused update)¹⁴;

Congenital heart disease (covered in the 2018 ACC/AHA guideline)¹⁵;

Valvular heart disease (covered in the 2020 ACC/AHA guideline)¹⁶;

Prevention of stroke in the setting of acute coronary syndromes (covered in the 2014 ACC/AHA guideline for non–ST-segment–elevation myocardial infarction [MI], and the 2017 ACC/AHA clinical performance measures for ST-segment–elevation MI and non–ST-segment–elevation MI)^17,18;

Subarachnoid hemorrhage (covered in the 2023 AHA/ASA guideline)¹⁹;

Pediatric stroke, except as it pertains to sickle cell disease (SCD)²⁰;

Secondary prevention of stroke (covered in the 2021 AHA/ASA guideline)²¹;

Cerebral venous thrombosis (covered in the 2024 AHA scientific statement)^22,23; and

Pathways for the implementation and dissemination of guideline recommendations in clinical practice.

Patient assessment;

Life’s Essential 8¹;

Atherosclerotic and nonatherosclerotic risk factors (eg, migraine);

Special populations, including trans health (a first for stroke primary prevention), SCD (the exception for including the pediatric population), genetic stroke syndromes, coagulation and inflammatory disorders, substance use, and sex-specific risk factors (pregnancy and pregnancy complications, endometriosis, hormonal contraception, menopause, and exogenous testosterone use);

Heart disease, specifically atrial cardiopathy and left ventricular dysfunction; and

Antiplatelet use for primary prevention.

The general population of adults without established CVD;

A population of adults with CVD but without a history of stroke or transient ischemic attack (TIA); and

A population of adults with CVD, including stroke or TIA.

Lack of inclusion in stroke clinical trials;

Clinical trial feasibility given low prevalence; and

Failure to identify populations as important subgroups.

An important application for stroke risk classification is to support primary prevention decision-making on initiation of antiplatelet therapy and therapy to lower BP or cholesterol. The relative risk (RR) reduction for these therapies is similar in patients at high and low risk for first stroke. The absolute risk reduction, however, is higher for those at higher risk. This means that fewer patients at high risk need to be treated to prevent a stroke compared with patients at low risk. Therefore, patients with higher absolute risk reduction may be more willing to incur the risk and inconvenience of preventive therapy than those at lower risk. Both the US Preventive Services Task Force (USPSTF) and the AHA endorse risk assessment for decision-making in primary prevention.^67,68,74,75 Clinicians use risk prediction instruments that estimate risk for atherosclerotic CVD broadly rather than cerebrovascular disease alone for 3 reasons: (1) Risk factors and preventive therapy overlap for these 2 diseases; (2) the broader instruments perform as well as stroke-specific instruments for stroke risk⁵⁸; and (3) in patient-centered care, both diseases are important. Among instruments,^67,77–79 the Pooled Cohort Equation is widely used in the United States.^58,80 However, the AHA has developed new equations, the Predicting Risk of CVD Events equations, that are expected to replace the Pooled Cohort Equation.⁷⁷

Risk prediction instruments can facilitate patient-centered, preventive stroke care. The right instrument, however, needs to be selected for the right patient, and the resulting risk estimates need thoughtful interpretation and application. Prediction instruments are commonly applied to 2 groups of patients to guide stroke prevention therapy: those with nonvalvular AF and those with or without AF who are at risk for arterial disease. The CHA₂DS₂-VASc score is recommended by the AHA to inform risk-based anticoagulation of patients with nonvalvular AF of any duration.^14,59 However, the CHA₂DS₂-VASc instrument is imperfect in that risk of stroke varies among populations with the same score. For this reason, the AHA recommends that a patient’s risk estimate may also consider other factors related to stroke risk such as burden of AF. With this understanding, oral anticoagulation is recommended for patients with an annual stroke risk ≥2% (generally a CHA₂DS₂-VASC score of ≥2 in men or ≥3 in women).^14,82 Our recommendation to apply the CHA₂DS₂-VASc score for assessment of patients with AF is consistent with other AHA guidelines.^82,83 These guidelines also recognize that risk estimation is just one factor in the decision for anticoagulation in patients with nonvalvular AF. Another is bleeding risk.⁸⁴ We agree with the ACC and AHA that treatment decisions should be individualized in the context of shared decision-making.⁸³

Several modifiable behaviors and medical conditions have been associated with increased stroke risk in observational research. Except for AF, no high-quality randomized trials have tested the effect of screening for these behaviors and conditions on stroke risk. Our recommendation to screen for modifiable behaviors and medical conditions is based on (1) randomized trials that demonstrate the benefit of treating risk factors, however identified, to reduce risk for stroke; (2) trials that demonstrate the benefits of treating risk factors to reduce the factors themselves; and (3) nonrandomized studies. In subsequent sections of this guideline that deal with specific behaviors and conditions, we summarize the evidence behind screening recommendations. For convenience, Table 4 lists modifiable behaviors or medical conditions for which we recommend screening. Some of these risk factors are associated. For example, hypertension, obesity, and obstructive sleep apnea (OSA) are risk factors for AF; modifying them could reduce risk for AF.⁸⁵ Table 4 also includes a recommendation to screen for SDOH, discussed in the Synopsis. One consequence of few screening trials is that the optimal screening interval is uncertain. The USPSTF recommends screening for hypertension yearly for adults >40 years of age and every 3 to 5 years for adults 18 to 39 years of age.⁶⁰ The AHA recommends screening for traditional risk factors every 4 to 6 years in adults 20 to 79 years of age.⁶¹ More frequent screening for modifiable behaviors and medical conditions may be warranted after a person is found to have borderline values on initial or subsequent testing.

Economic, environmental, and social factors modify risk for atherosclerotic CVD⁶² and stroke specifically.^35,63–65 Together, these nonclinical and nonbiological factors are referred to as SDOH.⁵⁷ They include fundamental factors such as exposure to structural racism, income, wealth, employment opportunity, and educational attainment; intermediate factors such as neighborhood safety, social environment (including isolation), and access to care; and proximate factors such as access to transportation, access to communication technology, and health literacy. The result of exposure to adverse SDOH is decreased detection and control of stroke risk.²⁶ In a recent scientific statement,²⁶ the AHA pointed out that fundamental causes are best addressed by policy and social movements. Intermediate and proximate factors can be addressed by local and individual interventions. At the clinic level, these include assistance with housing, food access, transportation to medical care, special efforts to build trust with health care professionals, health education, and assistance with medication adherence. No trials have tested the effect of screening for SDOH on stroke incidence, but we recommend screening for actionable determinants (eg, transportation, health knowledge, access to healthy food, health insurance, housing, transportation, communication technology, access to safe walking space) as a logical prerequisite to helping patients overcome barriers to control of their stroke risk.

Validation and testing of various tools to screen for SDOH are needed.

Testing of strategies is needed to address SDOH as an approach to decreasing stroke risk.

Development of improved instruments to predict risk of stroke alone is needed. These might include polygenic risk scoring to capture the familial aspect of risk.

In a systematic review and network meta-analysis of RCTs, Mediterranean dietary programs were superior to minimal intervention for the prevention of stroke for individuals at intermediate CVD risk (5%–10% 5-year cardiovascular event risk) and high CVD risk (20%–30% 5-year cardiovascular event risk) according to moderate-certainty evidence from 12 trials: pooled risk difference −7 per 1000 (95% CI, −11 to −1) and −16 per 1000 (95% CI, −25 to −3), respectively.⁸⁷ Subgroup analyses revealed that results did not vary according to the presence of CVD at baseline. The Mediterranean-style eating pattern has been endorsed by the AHA as part of Life’s Essential 8.¹

In an RCT of 20 995 adult men and women from rural China who had a history of stroke or were ≥60 years of age and had uncontrolled BP (SBP ≥140 mm Hg if receiving antihypertensive medication or ≥160 mm Hg if not), the use of a salt substitute (75% sodium chloride and 25% potassium chloride) compared with the use of regular salt (100% sodium chloride) reduced the risk of stroke.⁸⁸ Among the 5732 participants without stroke at baseline, the rate ratio for stroke was 0.78 (95% CI, 0.63–0.98). The use of the salt substitute was not associated with any serious adverse effects. However, because potassium levels were not measured in this trial, caution is warranted for individuals taking a potassium-sparing diuretic or a potassium supplement and those who have known kidney disease. Furthermore, the implementation of salt substitution with potassium may be difficult in countries where sodium is already added to processed food. Implementation would require food processing to be changed at the factory level.

In a systematic review of meta-analyses and single RCTs, B-complex vitamins (defined as a combination of 2 or more of the following: vitamin B₆, folic acid, vitamin B₁₂) reduced the risk of stroke in 9 of 12 trials.⁹⁰ The pooled rate ratio from the 12 trials was 0.90 (95% CI, 0.81–1.00; P=0.04) based on moderate-quality evidence. These results confirmed an earlier systematic review of meta-analyses and single RCTs.⁹¹ The studies included populations who were healthy and those with medical comorbidities such as end-stage renal disease, suspected or prevalent coronary artery disease, and prevalent CVD risk factors. None of the systematic reviews included the proportion of patients who were vitamin deficient, nor did they present results solely for the primary prevention of stroke.

REDUCE-IT (Reduction of Cardiovascular Events With Icosapent Ethyl–Intervention Trial)¹⁰³ was a randomized trial of 8179 individuals with established CVD or with diabetes and other risk factors who had been receiving statins and had a fasting triglyceride level of 135 to 499 mg/dL and a low-density lipoprotein (LDL) level of 41 to 100 mg/dL. Compared with placebo, icosapent ethyl, a highly purified form of eicosapentaenoic acid ethyl ester, which some investigators do not consider to be a dietary supplement,⁹¹ reduced the primary composite end point of cardiovascular death, nonfatal MI, nonfatal stroke, coronary revascularization, or unstable angina (hazard ratio [HR], 0.75 [95% CI, 0.68–0.83]). Although there was a reduction in stroke (HR, 0.72 [95% CI, 0.55–0.93]), the study did not report stroke outcomes among participants without baseline CVD. Meta-analyses have demonstrated that long-chain fatty acids do not reduce the risk of stroke according to moderate-quality^89,92 or low- or very low–quality^91,93,95 evidence. Another meta-analysis reported that long-chain fatty acids do not reduce the risk of stroke, with 6 of the 7 identified trials identified as good quality.⁹⁶ Study quality was not described in another systematic review and meta-analysis, which showed that long-chain fatty acids do not reduce the risk of stroke.⁹ In 3 meta-analyses, there was no effect on stroke in primary prevention CVD studies.^92–95

In 1 RCT, 14 641 male US physicians initially ≥50 years of age were enrolled and randomized to multivitamins (Centrum Silver) or placebo. Of the 14 641 participants, 754 (5.1%) had a history of CVD at baseline. At a median of 11.2 years of follow-up, multivitamins did not reduce the risk of stroke, including among those without CVD at baseline.⁹⁸ In another RCT of 1708 individuals ≥50 years of age who were ≥6 weeks after an MI, multivitamins (a 28-component high-dose multivitamin and multimineral mixture) did not reduce the risk of stroke compared with placebo.⁹⁹ Of the study population, 17% and 18% were female in the multivitamin and placebo arms, respectively. Systematic reviews and meta-analyses, which included these 2 studies, described the evidence examining stroke risk as very low quality.^90,91

Whether sodium substitution with potassium reduces the risk of incident stroke in the United States, where processed food products are common;

Whether folic acid or B-complex vitamins reduce the risk of incident stroke;

The effect of icosapent ethyl on incident stroke among those without baseline CVD;

The effect of saturated fat reduction on the risk of incident stroke;

The effect of the DASH diet on the risk of incident stroke;

The challenges of adherence to a recommended diet in socioeconomically oppressed communities;

How patients with conditions that affect dietary intake should be counseled on nutrition; and

The impact of plant-based diets on primary stroke prevention.

The US Centers for Disease Control and Prevention estimates that only ≈20% of US adults meet the 2018 guidelines for aerobic activity (ie, at least 150 minutes of moderate activity, 75 minutes of vigorous activity, or an equivalent combination) and engage in strength training at least 2 d/wk.¹³⁹ To identify patients who need help becoming more active, health care professionals can obtain data from self-report or wearable activity monitors such as pedometers.¹³¹ A recent systemic review identified 4 self-report questionnaires that scored highest on an objective rating scale, including the 9-item Rapid Assessment of Physical Activity and the 2-question Exercise Vital Sign (“On average, how many days per week do you engage in moderate to strenuous exercise [like a brisk walk]?” and “On average, how many minutes do you engage in exercise at this level?”).^105,131,140 In a large, single health system, implementation of Exercise Vital Sign surveys was associated with greater clinical documentation of physical activity participation rates, physician counseling on physical activity, more lifestyle-related referrals, and modest reductions in body weight and hemoglobin A1c.¹³² Wearable activity monitors are best at measuring step count; although this metric can be a useful measure of walking, it does not inform adherence to the US Department of Health and Human Services guideline for moderate and vigorous physical activity, misses some aerobic activity (eg, swimming, cycling), and does not reflect strength training.¹³¹ In addition, monitors vary in accuracy.¹⁴¹ No method for classifying physical activity is perfectly accurate,^{128,142–144} and no unbiased research shows that screening and classification lead to sustained physical activity change in individual patients or reduced stroke incidence. Despite these shortcomings in classification and effectiveness research, we recommend screening on the rationale that it can identify individuals with low activity who may benefit from counseling.

The AHA and the USPSTF recommend that health care professionals offer or refer adults for counseling to promote physical activity.^131,145,146 The target for physical aerobic activity according to these organizations and the World Health Organization¹³⁵ is at least 150 minutes of moderate-intensity, at least 75 minutes of vigorous-intensity physical activity, or some equivalent of moderate and vigorous activity each week. In addition, most organizations recommend regular strength training, usually ≥2 d/wk. A systematic review conducted for the USPSTF reported that, on average, behavioral counseling increased physical activity by 33 min/wk among adults without CVD risk factors.¹⁴⁶ Individuals assigned to a physical activity intervention were more likely to meet physical activity recommendations at 6 to 12 months compared with individuals assigned to control interventions (pooled odds ratio [OR], 1.41 [95% CI, 1.18–1.67]). Evidence-based methods to promote physical activity include repeated, individual counseling (minimum, 3–5 sessions) or group meetings based on models of behavior change that typically involve goal setting, monitoring, problem-solving, and feedback.^{110,111,147–149} There is evidence that wearable activity trackers, as a primary or secondary intervention, may improve participation in physical activity.¹⁵⁰ When physicians cannot provide intensive counseling themselves, brief counseling followed by referral to an exercise coach may be effective.^{110,111,151,152} Most studies of interventions to improve physical activity have had short durations; studies with longer follow-up (eg, >1 year) show that treatment effects commonly do not persist without ongoing coaching.^152,153 The USPSTF recommends offering or referring adults with cardiovascular risk factors for behavioral counseling to promote physical activity (Grade B recommendation from the USPSTF).¹⁴⁵

Sedentary behavior has been associated with a statistically significant increased risk for CVDs, including stroke in several^115,154 but not all studies.¹⁵⁵ In a recent systematic review and meta-analysis, stroke risk began to increase in a nearly linear pattern after 3.7 h/d sedentary behavior. Above 6.5 h/d, risk of stroke increased 6% for each additional hour. Above 11 h/d, risk increased by 21% for each hour.¹¹⁴ The association of sedentary behavior and risk for stroke may be modified by physical activity. As an example, in 1 study, nonoccupational sedentary behavior (≥8 hours) was significantly associated with increased stroke risk only in individuals with low physical activity.¹⁰⁷ Other researchers have reported a similar effect modification by physical activity level.^154–156 One meta-analysis, however, reported that greater time in sedentary behavior was associated with higher risk for CVD mortality even in individuals in the more-than-low levels of physical activity, suggesting that physical activity and sedentary behavior may have independent effects.^112,115,157 No unbiased research has tested the effect of interventions to reduce sedentary time on stroke risk. Our recommendation is based on epidemiological associations that suggest a benefit of reducing sedentary time and the low harm of acting on this data.

Further research would help to confirm that physical activity modifies the effect of sedentary behavior and to further test whether physical activity and sedentary behavior have independent effects on stroke risk or whether they are simply colinear variables.

There is a great need for both public health and clinical interventions that are effective for helping individuals reduce time in sedentary behavior and increase time in moderate-to-vigorous physical activity.

BMI is associated with an increased risk of stroke globally.^181–186 For each 5 kg/m² increment of BMI, stroke risk increases by 10%, as shown in a meta-analysis of 44 prospective cohort studies, >4 million participants, and >100 000 cases of incident stroke.¹⁶⁹ WC, waist-to-hip ratio (WHR), and waist-to-height ratio have been associated with incident stroke independently of BMI.^{158–160,183} WHR was associated with stroke in the Northern Manhattan Study after adjustment for BMI and other stroke risk factors (OR, 3.0 [95% CI, 1.8–4.8], fourth versus first quartile). Men with WHR >0.93 cm had a 3.8-fold (95% CI, 1.8–5.0) and women with WHR >0.86 cm had a 2.5-fold (95% CI, 1.6–4.0) increased risk of stroke, with similar magnitude in White individuals, Black individuals, and Hispanic individuals.¹⁶² For waist-to-height ratio, a meta-analysis of 7 studies showed that this measure was associated with an RR of 1.55 (95% CI, 1.37–1.76) for ischemic stroke and trended toward significance with intracerebral hemorrhage (ICH; RR, 1.30 [95% CI, 0.99–1.73]) when the highest and lowest quartiles were compared. The same meta-analysis showed that for a 10-cm higher WC, the relative stroke risk was 10% higher; for a 0.1-unit increase in WHR, the risk was 16% higher; and for a 0.05-unit increase in waist-to-height ratio, the risk was 13% higher.¹⁶¹ Therefore, screening for and recognizing obesity and adiposity are important first steps in the evaluation of stroke risk in primary care.

Bariatric surgery such as gastric bypass or sleeve gastrectomy can lead to significant weight loss in patients with obesity. Although not all studies have been shown to reduce the risk of incident stroke,^187–190 several high-quality nonrandomized studies do show reduced risk. An observational cohort study of 2287 patients with obesity who received bariatric surgery compared with controls with obesity showed that the risk of cerebrovascular disease (including ischemic stroke, hemorrhagic stroke, and carotid interventions) was reduced by 33% (HR, 0.67 [95% CI, 0.48–0.94]; P=0.02).¹⁶³ A retrospective propensity score–matched cohort of patients with severe obesity (BMI >35 kg/m² with comorbidities or BMI >40 kg/m²) showed a lower risk of cerebrovascular events in the surgical compared with the matched nonsurgical group (HR, 0.162 [95% CI, 0.073–0.360]).¹⁶⁴ A prospective cohort study of 2010 patients with obesity with bariatric surgery compared with 2037 matched nonsurgical controls reported a 34% decrease in total stroke events (adjusted HR, 0.66 [95% CI, 0.49–0.90]; P=0.008).¹⁶⁶ One meta-analysis reported significant reductions in relative odds of stroke, a 51% relative reduction in cumulative odds of stroke (4 studies; OR, 0.49 [95% CI, 0.32–0.75]),¹⁶⁵ and another reported a 36% relative reduction in stroke (14 studies; OR, 0.64 [95% CI, 0.53–0.77]; P<0.001).¹⁶⁷ The meta-analyses are therefore consistent in showing a benefit in stroke reduction.

Medications for both diabetes and obesity such as the glucagon-like protein-1 (GLP-1) receptor agonists, dual GLP-1 receptor agonists/glucose-dependent insulinotropic polypeptide agonist, and combination GLP-1 receptor agonists/long-acting amylin analog and sodium-glucose cotransporter-2 inhibitors, as well as new multitarget drugs demonstrating dramatic weight loss in phase II trials, should be tested in trials and other research to determine their safety and efficacy for primary stroke prevention.

Visceral adiposity index is a measure of abdominal adipose volume, calculated from WC, BMI, and triglyceride and high-density lipoprotein cholesterol levels with some adjustments that differ between men and women. This has been associated with incident stroke in China but should be assessed in other studies in diverse populations. Further research is needed before it is incorporated into clinical practice.

Additional studies of bariatric surgery with or without GLP-1 receptor agonists and sodium-glucose cotransporter-2 inhibitors and new derivatives to maintain weight loss are needed to determine the best strategy to lower cardiovascular risk, including stroke.

Direct evidence that screening for OSA reduces stroke risk is lacking. A 2022 review of the evidence on screening for OSA included 86 studies, and none addressed whether screening for OSA in adults improves health outcomes.¹⁹¹ No study compared OSA screening with no screening directly.¹⁹¹ The authors and the USPSTF concluded that the evidence is insufficient to recommend screening for OSA in the general population of adults with no signs or symptoms of OSA.¹⁹¹ They noted that none of the screening instruments for OSA (eg, the Berlin Questionnaire, STOP-BANG [Snoring, Tiredness during daytime, Observed apnea, high blood Pressure-BMI, Age, Neck circumference, Gender] questionnaire, Epworth Sleepiness Scale) have been adequately validated in general populations such as people in primary care settings.

Nonrandomized observational studies suggest CPAP treatment might reduce stroke risk in patients with OSA. A meta-analysis of 4 cohort studies including 2681 participants with or without stroke suggested that CPAP treatment for OSA was associated with lower stroke risk (pooled OR, 0.59 [95% CI, 0.35–0.99]; P=0.047) with modest heterogeneity across studies (I²=21%).¹⁹² In a separate meta-analysis of 3 cohort studies including 912 participants, CPAP treatment was associated with lower stroke risk (RR, 0.27 [95% CI, 0.14–0.53]; P≤0.001); however, 1 RCT and 2 studies using administrative data did not reproduce this result.¹⁹³ A propensity score–matched analysis from the SAVE (Sleep Apnea Cardiovascular Endpoints) trial showed that patients who were adherent to CPAP therapy had a lower stroke risk than those in the usual-care arm (HR, 0.56 [95% CI, 0.32–1.00]; P=0.05¹⁹⁴). Despite this observational evidence, individual and aggregated RCTs have not shown that CPAP treatment for OSA reduces stroke. For example, a meta-analysis of 9 RCTs including 4698 participants found that CPAP treatment did not reduce stroke risk (pooled OR, 0.94 [95% CI, 0.70–1.24]; P=0.64) with no evidence of heterogeneity across trials (I²=0%).¹⁹² Most trials enrolled participants with stroke or CVD.

There are no clinical trial data on the effect of interventions for other sleep disorders on stroke risk. In particular, health care professionals and patients need clinical trial evidence on the impact of interventions optimizing sleep duration (7–9 hours of sleep daily) to reduce stroke risk.

Clinical trial evidence that CPAP or mandibular advancement devices for OSA reduce stroke risk is lacking. Most CPAP and mandibular advancement device trials included middle-aged men who were overweight or obese and had moderate-to-severe OSA. Trials in women, in younger and older adults, and in those with normal weight or mild OSA are needed.

OSA trials of participants free of CVD, including stroke, are needed. Such trials will need a larger sample size, longer duration, or both compared with the trials of participants with CVD to have sufficient stroke events and to determine the efficacy of OSA on stroke risk. Most trials of mandibular advancement devices were ≤12 weeks and did not measure stroke events. More extended studies of the effect of the devices on stroke are needed.

Adherence to CPAP treatment is challenging for patients with OSA. The minimum duration of CPAP use (hours per night) required to reduce stroke risk is unclear.

Direct evidence that screening for diabetes or prediabetes in asymptomatic adults or those with unrecognized diabetes symptoms reduces stroke risk is lacking. A 2021 review of the evidence on screening for diabetes or prediabetes found inadequate direct evidence that screening for diabetes or prediabetes leads to improvements in mortality or cardiovascular morbidity.²⁰² No trials assessed initial screening with hemoglobin A1c or fasting glucose, and none assessed screening for prediabetes.²⁰³ However, the authors and the USPSTF recommend screening for diabetes and prediabetes in adults 35 to 70 years of age who have overweight or obesity on the basis of evidence that interventions for newly diagnosed diabetes and prediabetes have net health benefits,²⁰² but they do not provide screening intervals. The American Diabetes Association recommends screening for diabetes or prediabetes for all adults ≥35 years of age at a minimum of 3-year intervals.²⁰⁴ They also recommend considering screening for diabetes or prediabetes in asymptomatic adults of any age with overweight or obesity and another risk factor: high-risk race or ethnicity (eg, African American, Latino, Native American, Asian American, Pacific Islander), CVD, hypertension, high-density lipoprotein cholesterol level <35 mg/dL, triglyceride level >250 mg/dL, polycystic ovary syndrome, physical inactivity, and insulin resistance–associated conditions (eg, severe obesity, acanthosis nigricans).²⁰⁴

Individual and aggregated RCTs have demonstrated that GLP-1 receptor agonists reduce stroke risk. SUSTAIN-6 (Trial to Evaluate Cardiovascular and Other Long-Term Outcomes With Semaglutide in Subjects With Diabetes) showed that semaglutide reduced stroke risk more than placebo (event rate, 1.6% versus 2.7%; HR, 0.61 [95% CI, 0.38–0.99]; P=0.04) in patients with diabetes.²⁰⁵ An exploratory analysis of the REWIND trial (Researching Cardiovascular Events With a Weekly Incretin in Diabetes) suggested that dulaglutide might reduce ischemic stroke in patients with diabetes (HR, 0.75 [95% CI, 0.59–0.94]; P=0.01).²⁰⁶ A meta-analysis of 12 RCTs involving 20 867 patients found that GLP-1 receptor agonist use compared with placebo/active comparator in adults with diabetes was associated with lower stroke risk (RR, 0.73 [95% CI, 0.60–0.89]; P≤0.01) with no evidence of heterogeneity across trials (I²=0%).¹⁸⁰ Most trials enrolled patients with a hemoglobin A1c of ≥7% and CVD or high cardiovascular risk. Individual and aggregated RCTs have not established that sodium-glucose cotransporter-2 inhibitors reduce stroke risk.²⁰⁷ A post hoc trial analysis suggested that sotagliflozin might reduce stroke risk more than placebo (HR, 0.66 [95% CI, 0.48–0.91]; P value not provided) in patients with diabetes and chronic kidney disease.²⁰⁸ Confirmatory trials are needed.

RCTs have not demonstrated that intensive glycemic control (targeting hemoglobin A1c ≤6.5%) reduces stroke risk more than looser glycemic control (targeting hemoglobin A1c of 7%–8%) in patients with type 1 or type 2 diabetes.^209–214 A review of the evidence on intensive compared with conventional glycemic targets in patients with type 1 diabetes, including 12 trials and 2230 patients, found low-quality evidence for stroke with small numbers of the outcome.²¹¹ The authors concluded that the evidence of the effect of intensive glycemic control on stroke was insufficient.²¹¹ Furthermore, tight glycemic control increases treatment burden, costs, and side-effect risks and might cause patient harm (eg, severe hypoglycemia).^211,215 In summary, for individuals with diabetes, both the potential benefit of more intensive control (eg, to hemoglobin A1c ≤6.5%) for prevention of microvascular disease and the potential risk of hypoglycemia should be considered in decision-making.

Health care professionals and patients need more evidence on the effect of screening for diabetes or prediabetes on stroke risk, particularly in racial groups with higher risk of diabetes and stroke than White people.

The optimal ages and frequency of screening for diabetes or prediabetes are unclear. Research is needed to better understand the optimal frequency and ages of screening.

Referring patients with prediabetes to effective interventions (lifestyle or metformin) to prevent stroke has uncertain benefits.

Health care professionals and patients need clinical trial evidence on the effects of lifestyle interventions and medical treatments for screen-detected prediabetes and diabetes on stroke over long follow-up periods, particularly in racial and ethnic groups with high diabetes and stroke risk.

Optimal hyperglycemia control (ie, hemoglobin A1c levels) to lower stroke risk in people with type 1 or type 2 diabetes is unclear. Most trials of medication treatments included middle-aged men, White people or Asian people, and participants with CVD or high CVD risk. Trials in women; people who are African American, Latino, Native American, or Pacific Islander; younger and older adults; and participants free of CVD are needed.

In primary prevention, the majority of participants in GLP1 receptor agonist trials had prevalent CVD (including stroke). Although diabetes is rarely isolated, more longitudinal research is needed to better understand the primary prevention impact of these and other diabetes drugs on stroke.

Only a few studies have assessed the value of screening for hypertension to facilitate stroke prevention. One cluster RCT randomized 39 communities in Canada to a program of pharmacy-based screening and education or no intervention.²²¹ Participants were volunteers who responded to direct solicitation, advertising, or referral from a health care professional. The intervention consisted of 3-hour weekday BP measurement, CVD risk factor assessment, and educational sessions held concurrently in all 20 intervention communities during a 10-week period in 2006. Participants received pharmacy-based BP screenings with an automated instrument and completed a standardized risk profile. The primary outcome was relative change in the mean annual rate of hospital admissions for acute MI, congestive heart failure, or stroke in the cluster populations. After 1 year of follow-up, the intervention arm had 3.02 fewer annual hospital admissions for CVD per 1000 individuals compared with the control group (rate ratio, 0.91 [95% CI, 0.86–0.97]; P≤0.01). For a stroke-specific outcome, the rate ratio was 0.99 (95% CI, 0.88–1.12; P=0.89). Design features of this and other screening RCTs do not permit accurate estimation of the effect of screening for hypertension to prevent stroke in individuals. However, compelling data linking hypertension to stroke risk and compelling data on treatment effect were considered when this recommendation was crafted. The 2021 USPSTF recommended annual screening for adults ≥40 years of age and those at increased risk for hypertension.

Abundant high-quality RCTs^{222–224,229} and systematic reviews/meta-analyses^225–228 support this recommendation for primary (and secondary) prevention of stroke. Individual primary prevention–oriented RCTs include a 2-arm cluster-designed community health care worker–led implementation trial conducted in rural China,²²² which resulted in an SBP/DBP treatment group difference of 23.1/9.9 mm Hg and a stroke (secondary outcome) HR of 0.66 (95% CI, 0.60–0.73). More traditional 2-arm RCTs include an early trial from the United States with a 5-year SBP/DBP difference of 12/4 mm Hg and a stroke (primary outcome) HR of 0.64 (95% CI, 0.50–0.82) and a more recent RCT from China with an SBP/DBP difference of 9.3/2.8 mm Hg and a stroke (secondary outcome) HR of 0.67 (95% CI, 0.47–0.97).^223,229 The ACCORD BP trial (Action to Control Cardiovascular Risk in Diabetes), which was restricted to adults with diabetes, yielded a stroke (secondary outcome) HR of 0.59 (95% CI, 0.39–0.89).²²⁴ Individual participant data,²²⁵ group data,²²⁶ and network meta-analyses²²⁷ have demonstrated a stroke benefit of more compared with less intensive antihypertensive treatment, including an SBP <130 mm Hg versus higher BP treatment targets benefit. A network meta-analysis restricted to adults with type 2 diabetes also demonstrated significant stroke reductions in those randomized to more versus less intensive antihypertensive treatment, including an HR for an achieved SBP of 130 to 134 mm Hg versus 140 to 144 mm Hg of 0.76 (95% CI, 0.54–0.99).²²⁸

In a meta-analysis of first-step treatments, diuretic, β-blocker, calcium channel blocker, angiotensin-converting enzyme inhibitor, and angiotensin receptor blocker agents were more effective than placebo for prevention of stroke.²⁴² In subsequent meta-analyses, first-step antihypertensive therapy with β-blockers was shown to be inferior to diuretic, calcium channel blocker, angiotensin-converting enzyme inhibitor, or angiotensin receptor blocker therapy, and calcium channel blocker treatment was the most consistently beneficial for prevention of stroke.^230,231 In the most scientifically rigorous RCT comparison of first-step antihypertensive medications, there was no superiority for prevention of stroke (secondary outcome) in those randomized to the calcium channel blocker amlodipine or the angiotensin-converting enzyme inhibitor lisinopril compared with the thiazide-type diuretic chlorthalidone.²³² Chlorthalidone was superior for the prevention of heart failure, especially compared with amlodipine. The ACC/AHA BP guideline recommends the use of diuretics, calcium channel blockers, angiotensin-converting enzyme inhibitors, and angiotensin receptor blockers for first-step antihypertensive drug therapy.⁶⁸

RCTs and surveys document the need for ≥2 antihypertensive medications in most adults with hypertension who require BP-lowering therapy. In trials that have used an SBP/DBP target of <140/90 mm Hg, the average number of antihypertensive medications required has been ≈2.^229,234,235 In ALLHAT (Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial), the SBP/DBP <140/90 mm Hg goal was achieved in only 29.6% of the participants treated with drug monotherapy.²³⁵ The corresponding percentages for BP goal achievement with up to 2, 3, and 4 antihypertensive medications were 54.2%, 65.6%, and 70.4%, respectively. For the trials with an SBP target <120 mm Hg, the average number of antihypertensive drugs required has been ≈3.^224,236 A National Health and Nutrition Examination Survey 2013 to 2016 report documented treatment with 1, 2, 3, or 4 antihypertensive medications in 40.1%, 38.4%, 16.6%, and 4.3% of US adults, respectively, and ≥2 antihypertensive medications were being used in almost 60%.²³⁷

It is intuitive to screen for hypertension to initiate treatment in those at a higher risk for stroke. However, no data exist to guide clinicians on BP screening frequency or interval for effective primary prevention of stroke. Future research should investigate the frequency and interval of hypertension screening for prevention of stroke based on sex, race and ethnicity, and risk of stroke.

SPRINT (Systolic Blood Pressure Intervention Trial) and post-hoc analyses of the ACCORD BP trial have reported a CVD prevention benefit and, in the case of the ACCORD BP trial, a stroke (secondary outcome) prevention benefit for treatment to an SBP target <120 mm Hg compared with <140 mm Hg. Ongoing SPRINT-like trials in Brazil and China are using the same SBP treatment targets as the SPRINT and ACCORD BP trial. Experience in these and other trials will be helpful in confirming and quantifying the potential for an incremental stroke primary prevention benefit with an SBP target <120 mm Hg compared with <140 mm Hg.

Third-generation β-blockers such as carvedilol and nebivolol are highly cardio-selective and result in vasodilation in addition to β1-receptor blockade. RCT meta-analysis indicates that third-generation β-blockers reduce central BP more than the earlier nonvasodilating second-generation β-blockers such as atenolol and metoprolol. RCTs with sufficient power to recognize stroke differences are needed to resolve the role of third-generation β-blockers for first-step treatment of hypertension in adults without a compelling indication for selection of a β-blocker.

Additional RCTs are needed to compare the effectiveness for stroke prevention of a policy of initial combination pill with a policy of sequential or stepped addition of separate antihypertensive agents.

Antihypertensive drugs with complementary mechanisms of action are recommended for initial combination drug therapy. The ACCOMPLISH RCT (Avoiding Cardiovascular Events Through Combination Therapy in Patients Living with Systolic Hypertension) investigators reported that initial benazepril plus amlodipine was superior to benazepril plus hydrochlorothiazide for a CVD composite outcome (HR, 0.80 [95% CI, 0.72–0.90]). However, the stroke difference was nonsignificant (HR, 0.84 [95% CI, 0.65–1.08]). In addition, the choice and dosage of the diuretic used have been criticized. Additional single-pill combination antihypertensive drug therapy comparisons are needed.

According to 2 meta-analyses, statin therapy reduces the risk for first stroke in adults who qualify for lipid-lowering therapy (Table 8) and are at high cardiovascular risk (risk ratio, 0.78 [95% CI, 0.68–0.89]; 10 trials, n=40 295;²⁴³ RR, 0.81 [95% CI, 0.75–0.87]; 24 trials, n=165 972).²⁴⁴ A secondary analysis of data from the HOPE-3 trial, a 2×2 factorial study of antihypertensives and rosuvastatin 10 mg in patients with vascular risk factors and no CVD, also found that treatment with a statin was associated with an overall reduction in stroke (HR, 0.70 [95% CI, 0.52–0.95]).²⁴⁸ The reduction in stroke with statin treatment in those at high cardiovascular risk can be included in patient discussions related to the overall benefit of statins for lowering their risk of vascular events.

Effective management of hyperlipidemia in adults includes potent proprotein convertase subtilisin/kexin 9 (PCSK9) inhibitors to reach treatment goals. No trials of PCSK9 inhibitors have been designed to test their effect in primary prevention of stroke. A meta-analysis compared the effects of PCSK9 inhibitors with placebo or active treatment (ie, statins, ezetimibe, or both) for the prevention of adverse vascular outcomes.²⁴⁵ The study combined data from trials of patients with and without CVD, including stroke; no separate analysis was done to test the effect of PCSK9 inhibitors on first stroke. There were no statistically significant treatment effects of alirocumab compared with another active lipid treatment on risk for any stroke (an analysis of evolucumab versus active therapy for stroke prevention was not provided).²⁴⁵ Both alirocumab and evolocumab were effective in preventing any stroke compared with placebo.²⁴⁵ The meta-analysis did not report the effect of PCSK9 inhibitors in primary prevention populations separately. In summary, the effects of PCSK9 inhibitors for primary stroke prevention are uncertain.

The CLEAR Outcomes trial (Cholesterol Lowering via Bempedoic Acid, an ACL-Inhibiting Regimen) assessed the effect of bempedoic acid on cardiovascular outcomes for statin-intolerant individuals. Patients were eligible if they were statin intolerant and had LDL-C >100 mg/dL and elevated cardiovascular risk based on an elevated Reynolds risk score, a coronary artery calcium score >400 Agatston units, or presence of either type 1 or 2 diabetes in women >65 years of age or men >60 years of age. In a prespecified subgroup analysis of the 4206 primary prevention participants, although the point estimate favored treatment, the benefit of bempedoic acid for the reduction of a first stroke was not clear (HR, 0.76 [95% CI, 0.46–1.26]).²⁴⁶ Participants had a mean age of 67.9 years; 59% were female; >90% identified as White; and >80% identified as non-Hispanic. The trial was not powered to detect an independent effect on stroke, which was a secondary outcome.²⁴⁶

There is no evidence that supplementation with long-chain omega-3 fatty acid in adults taking moderate or low amounts of long-chain omega-3 fatty acids reduces the risk of a first stroke as shown by 1 meta-analysis (RR, 0.98 [95% CI, 0.9–1.07]).⁹²

There are no prospective, randomized, primary prevention trials of lipid management in which stroke is a primary end point. Because it is unlikely that such trials will be conducted in the future, trials might stratify enrollment according to the level of stroke risk to allow more robust assessments of therapies on stroke-related outcomes.

Additional studies are needed to determine whether lipid-lowering therapies are similarly effective and have similar risks across racial and ethnic groups and in men and women.

Additional studies of PCSK9 inhibitors and bempedoic acid are needed to establish whether these therapies reduce the risk of a first stroke.

Cigarette smoking is associated with stroke regardless of age,^254,261 sex,^261–263 race,^255,279 and smoking intensity.^259,260 A contemporary meta-analysis showed that although there is heterogeneity, cigarette smoking is associated with ischemic stroke (RR, 2.17 [95% CI, 2.06–2.28]), ICH (RR, 1.77 [95% CI, 1.60–1.95]), and subarachnoid hemorrhage (RR, 3.46 [95% CI, 3.15–3.80]).²⁶³ Mendelian randomization studies support causality for ischemic stroke and subarachnoid hemorrhage.^256,264 In addition, an updated meta-analysis found that environmental tobacco smoke exposure was associated with an increased risk of stroke (RR, 1.23 [95% CI, 1.16–1.31]).²⁶⁷ New data pertain to additional tobacco and related products. Water pipe smoking was associated with an increased odds of ischemic stroke in 2 case-control studies, with a stronger association among active cigarette smokers.^265,269 Smokeless (oral) tobacco use was associated with stroke in a meta-analysis even after the exclusion of studies that did not adjust for cigarette smoking (RR, 1.18 [95% CI, 1.04–1.32]).²⁶⁶ The association of sole e-cigarette use with stroke (OR, 1.13 [95% CI, 0.99–1.29]) in a meta-analysis was inconclusive, with the confounding effect of combustible tobacco smoking noted.^257,258 Moreover, these e-cigarette data do not reflect the long-term risk of stroke incurred by transitioning from sole e-cigarette use to combustible cigarette use and dual use.^268,270

The preponderance of data demonstrating an association of cigarette smoking with stroke support discouraging smoking initiation among nonsmokers.^{254,255,259–263,279} The data on stroke risk attributable to water pipe smoking, smokeless tobacco, and electronic nicotine delivery systems such as e-cigarettes remain comparatively limited.^{257,258,265,266,269} However, users of such products frequently also use combustible cigarettes, and there is a clear association of sole e-cigarette use with higher odds of subsequent combustible smoking initiation.^268,269 Therefore, initiation of water pipe smoking, smokeless tobacco use, and use of electronic nicotine delivery systems such as e-cigarettes should also be discouraged among nonsmokers.

No randomized trials have compared individual smoking cessation strategies specifically for the outcome of stroke risk reduction. However, epidemiological analyses show that former smokers have a reduced risk of stroke compared with active smokers; smoking cessation is associated with a reduction in stroke risk.^260,278,279 For active smokers, high-quality data from meta-analyses of randomized trials demonstrate that a range of counseling and advice interventions are effective for increasing smoking cessation compared with no or minimal interventions.^{271,280,281,284} Typically, in studies included in the meta-analysis, the intervention consisted of repeated contact by telephone or face-to-face contact outside of routine clinical care,²⁸⁰ but even brief interventions have been shown to be effective, including a 30-second physician intervention.²⁸⁵ Thus, at minimum, counseling and advice should be provided to all smokers to increase the odds of smoking cessation, consistent with USPSTF recommendations.³⁰⁴ Water pipe, smokeless tobacco, and e-cigarette use is variably associated with stroke risk, and users of such products frequently have concomitant and future combustible cigarette use. Therefore, patients using these products may also benefit from counseling interventions, for which there is variable, low-certainty evidence supporting the use of face-to-face, telephone-based, and text message–based interventions.^282,283,286

Smoking cessation pharmacotherapy combined with behavioral counseling is superior to behavioral interventions alone. Pharmacotherapy options include varenicline, bupropion, and nicotine replacement therapy. In RCTs²⁷² and in large meta-analyses,^274,275,277 use of pharmacotherapy was associated with a substantial increase in smoking cessation compared with behavioral interventions alone. In a comparison of pharmacotherapy options, varenicline was found to be superior to bupropion (RR, 1.36 [95% CI, 1.25–1.49]) and nicotine replacement monotherapy (RR, 1.25 [95% CI, 1.14–1.37]) for achieving smoking abstinence.²⁷⁵ However, combination nicotine replacement therapy (fast acting plus transdermal) is superior to monotherapy (RR, 1.27 [95% CI, 1.17–1.37]),²⁷⁶ with evidence suggesting that varenicline and combination nicotine replacement therapy are similar in efficacy (RR, 1.02 [95% CI, 0.87–1.20]).²⁷⁵ These data are reflected in the USPSTF recommendation to initiate pharmacotherapy in combination with behavioral interventions for smoking cessation.³⁰⁴ However, behavioral support added to pharmacotherapy increases quit rates modestly²⁷¹; the unavailability of specialized behavioral counseling does not necessarily preclude initiation of pharmacotherapy. In addition, a recent trial found that varenicline was superior to behavioral interventions alone among light smokers (≤10 cigarettes/d), demonstrating that pharmacotherapy should be used even among light smokers, which constitute 65% of Black Americans who smoke.²⁷³

Recent research has investigated strategies to increase implementation of smoking cessation interventions. The “opt-out” approach is based on the concept that changing the default treatment may affect engagement in smoking cessation. In a randomized trial of 1000 smokers encountered in the hospital, delivery of an opt-out intervention to all patients was compared with the standard “opt-in” approach whereby patients reporting willingness to attempt smoking cessation were provided interventions.²⁸⁷ For both groups, intervention components included pharmacotherapy provision before discharge and up to 4 postdischarge telephone counseling visits. The opt-out intervention was associated with greater smoking cessation at 1 month (Bayesian posterior probability, 0.97) but not at 6 months (Bayesian posterior probability, 0.57). Patients in the opt-out group were approximately twice as likely to be engaged in smoking cessation after discharge, as reflected by smoking cessation medication and telephone counseling use. In an observational study, implementation of opt-out smoking cessation programs at the hospital level was feasible, and patients who received opt-out bedside interventions were more likely to use smoking cessation medications and quit at 1 month.²⁸⁸ Consistent with these and other data, the American Thoracic Society clinical practice guideline recommends varenicline initiation even in smokers not immediately ready to quit.³⁰⁵

According to a continuously updated systematic review and meta-analysis, there was high certainty based on 6 RCTs with 2378 participants that smoking cessation rates were higher in people randomized to nicotine e-cigarettes than in those randomized to nicotine replacement therapy (RR, 1.63 [95% CI, 1.30–2.04]).²⁸⁹ A key trial randomized 886 smokers in the UK National Health Service to nicotine replacement therapy, including combination therapy, compared with e-cigarettes, with both groups receiving behavioral counseling support.²⁹⁰ The 1-year abstinence rate was 18.0% in the e-cigarette group compared with 9.9% in the nicotine replacement group (RR, 1.83 [95% CI, 1.30–2.58]). However, among participants with 1-year smoking abstinence, 80% in the e-cigarette group had continued use of e-cigarettes, whereas only 9% in the nicotine replacement therapy group had continued use of nicotine replacement. Although e-cigarettes may increase smoking cessation, the long-term health implications are unclear because there are no data on the long-term implications of e-cigarette use for stroke risk and other health outcomes²⁹⁴ and because switching to e-cigarettes may not, in the long term, prevent relapse to cigarette smoking or dual use.²⁹¹

Longitudinal studies are required to understand the long-term clinical cerebrovascular implications of e-cigarette use.

Longitudinal studies are needed to understand the long-term impact of e-cigarette use on dual use and transitioning to use of combustible cigarettes.

Future smoking cessation intervention studies should seek to identify pragmatic smoking cessation interventions that leverage both behavioral counseling and pharmacotherapy.

Future studies should identify effective strategies to facilitate cessation of other tobacco products and use of electronic nicotine delivery systems such as e-cigarettes and vapes.

Implementation studies should elucidate strategies for increasing delivery of effective smoking cessation strategies across various practice environments, resource settings, and patient populations.

Present data strongly recommend against screening the general population for ACS.^316,322 The reasoning stems from multiple factors: effects of false-positive results, inadequate direct evidence that screening for ACS leads to reduction in stroke or death, and the likelihood of small to moderate harms of screening for treatment of ACS.^316,323 A carotid bruit can reflect an underlying stenosis; however, sensitivity for detecting it is low. In NOMAS (Northern Manhattan Study), auscultation had a sensitivity of 56% and a positive predictive value of 25%. Some consideration must be given to specific high-risk asymptomatic populations such as those with atherosclerotic risk factors. Prior guidelines have proposed that screening of highly selected populations might be of benefit.^306,324,325 Other reports have analyzed and identified patients at high risk of >50% carotid artery stenosis; risk factors identified were hypertension, current tobacco use, coronary artery disease, or first-degree family member with a history of stroke.^326,327

Multidisciplinary decision-making plays a pivotal role in identifying the best course of management for asymptomatic carotid stenosis. With evolving medical management and growing options for surgical interventions, the health care team must carefully weigh the risks and benefits associated with each approach to effectively mitigate the risk of stroke. Although it is crucial for all patients to receive optimal medical management, it is essential to identify the subset of patients who remain at high risk of stroke despite optimal medical treatment. Despite trials supporting both revascularization and emerging evidence for best medical therapy, a consensus is still lacking. This is precisely why shared decision-making assumes paramount importance in advocating for the best course of treatment. By involving patients in the decision-making process, empowering them with a sense of agency over their health, enabling them to actively participate in making informed choices that align with their unique circumstances, a collaborative approach draws on the expertise of specialists dedicated to reducing the risk of stroke in patients with asymptomatic stenosis; this collaborative effort ensures a comprehensive evaluation and patient-centered care.

Most people in this population are already on statin therapy for other diagnoses; however, in those who are not, initiation of statin therapy provides stroke risk reduction. In a meta-analysis of 14 randomized trials of statins, there was a >15% decrease in the rate of stroke for every 10% reduction in serum LDL-C.³¹⁰ In the MESA study (Multi-Ethnic Study of Atherosclerosis), carotid plaque lipid core detected by magnetic resonance imaging (MRI) was strongly associated with total cholesterol.³²⁸ In the Framingham Heart Study, the relative increased risk of carotid artery stenosis was 10% for every 10 mg/dL increase in total cholesterol.³²⁹ In ACST (Asymptomatic Carotid Surgery Trial), there was an absolute stroke risk reduction attributed to lipid-lowering therapy in patients undergoing carotid endarterectomy (CEA) for carotid stenosis (>70%).³¹³ Others have reported that the annual risk of ACS is lower than reported in CEA trials, presumably because of improved medical therapy.³¹¹ High-intensity statin therapy is appropriate for patients with ACS, regardless of revascularization. All patients with carotid artery atherosclerosis must be treated with optimal medical treatment and risk factor modification.³²⁵ No large, randomized trials have directly analyzed the effects of statin therapy in asymptomatic patients; however, the indirect evidence supports the use of statins for stroke risk reduction, as detailed in Section 4.7 Lipids.

This is a highly debated topic. Over the past 2 decades, medical management has evolved from when ACAS (Asymptomatic Carotid Atherosclerosis Study) highlighted the benefits of CEA in stroke risk reduction.³¹² ACST-1 underscored the merits of carotid revascularization (CEA or transfemoral stenting) for net 5-year risk stroke reduction benefit for patients undergoing immediate compared with deferred surgery.³¹³ However, <40% of patients received statins, which is not consistent with current practices or recommendations. A 10-year analysis showed 4.5% absolute risk reduction for CEA compared with medical therapy alone.³⁰⁹ Although it was stopped prematurely; the SPACE-2 trial (Stent-Protected Angioplasty in Asymptomatic Carotid Artery Stenosis versus Endarterectomy: Two Two-Arm Clinical Trials) did not show clear superiority of carotid revascularization over medical therapy alone in patients with >70% to 99% of asymptomatic carotid stenosis.³¹⁴ With more interventional options to treat carotid stenosis, comparative studies became essential to determine their role in stroke risk reduction.^330–332 CREST 1 was the first large prospective randomized trial comparing stenting and endarterectomy. When revascularization was considered, CEA and carotid stenting had similar rates of perioperative stroke, MI, death, and subsequent ipsilateral stroke.³³³ Notably, there was an interaction between age and intervention, with better outcomes from coronary artery stenting for patients <70 years of age and better outcomes from CEA for patients >70 years of age. Likewise, transcarotid artery revascularization is a newer technique for carotid revascularization and could be considered on an individual basis. These trials showed greater effects for men than women.³²⁵ Without a large clinical trial directly comparing revascularization with contemporary medical management, a definitive conclusion on whether there is benefit from carotid revascularization in asymptomatic patients remains elusive. Results from the anticipated CREST 2 could provide much-needed clarity on this contentious subject.

Duplex ultrasound, the method of choice for screening patients with known atherosclerotic carotid artery stenosis, has the lowest cost/risk. Both the severity of carotid stenosis and the progression of disease are associated with increased risk of stroke.³¹⁸ A systematic review and meta-analysis stratifying by degree of stenosis showed that stroke risk increased according to the progression of stenosis. Ipsilateral stroke risk was highly dependent on degree of stenosis: <5% after 5 years for moderate stenosis and 15% with severe stenosis.³¹⁵ Other studies support the association of stenosis severity and progression with independent stroke risk and adverse outcomes in high-risk cardiovascular patients.^318,334,335 The association of ipsilateral neurological events with the yearly rate of change in luminal narrowing was highly statistically significant (P≤0.001) for patients who had progression of >2 categories in 1 year, considered to be at high risk of ipsilateral ischemic event relative to nonprogressors.³¹⁷ Independent predictors of progression are male sex, high creatinine, not taking lipid-lowering therapy, less severe low grade of stenosis, and increased plaque area.³¹⁸ Once patients have been diagnosed with >50% carotid stenosis, annual follow-up with carotid duplex may be reasonable to identify the progression of stenosis and subsequent risk of stroke.

Much has been reported on the benefit of carotid intervention for asymptomatic patients to reduce the risk of stroke, particularly in the population considered at high risk. This has been detailed in several studies. High-risk patients are those with the following characteristics: decreased life expectancy of 3 to 5 years, cardiovascular comorbidities (clinically significant cardiac disease, recent MI, congestive heart failure, ejection fraction <30%, abnormal stress test, or need for coronary artery bypass graft), severe pulmonary disease, and perioperative stroke/death rates >3%.^307,319,320 Within this high-risk population, the 3-year outcome from the SAPPHIRE study (Stenting and Angioplasty With Protection in Patients at High Risk for Endarterectomy) revealed that patients undergoing carotid artery stenting had a notably higher death rate (20.0%) than stroke rate (10.1%). This brings up questions about the efficacy of reducing the risk of stroke in this high-risk cohort. In addition, there was no medically treated control group, and the complication rates in both treatment arms were high enough to raise questions about the benefit of either intervention over medical therapy alone.³¹⁹ In this vulnerable group of patients, careful consideration for management strategies has to be discussed in a multidisciplinary setting, and shared decision-making is recommended.

The efficacy of best medical management alone compared with carotid revascularization to reduce risk of stroke in patients with asymptomatic atherosclerotic carotid artery stenosis >70% should be assessed.

The role of carotid revascularization in patients considered high-risk surgical candidates should be defined. Guidance should be developed to identify this vulnerable group of patients and include shared decision-making to optimize stroke risk reduction.

Enhanced imaging modalities could provide detailed insights into plaque morphology, stability, and risk of rupture, which may ultimately play a role in risk stratification to aid decision-making on carotid interventions and medical management. Identification of patients with asymptomatic stenosis at high risk of stroke is an important priority.

Sex-specific studies evaluating the efficacy and outcomes of carotid revascularization to reduce risk of stroke in women with asymptomatic atherosclerotic carotid artery stenosis are needed.

Early evidence suggests that transcarotid artery revascularization might have a role in the treatment of patients with ACS who fulfill specific criteria. However, no RCT data exist comparing its outcomes with either transfemoral carotid stenting or CEA. There is a pressing need for more comprehensive evidence and validated data to thoroughly understand the long-term advantages and potential risks associated with the broader adoption of this surgical technique. Long-term trials to establish a clear understanding of its efficacy, safety, and overall impact on patient outcomes for ACS are still needed.

Ultimately, shared decision-making is paramount in determining the best treatment approach for ACS, including carotid revascularization with CEA, stenting (transfemoral/transcarotid artery revascularization), or contemporary medical management.

Currently, there are no validated tools to precisely determine the optimal frequency of carotid duplex surveillance for patients exhibiting a more severe degree of stenosis or rapid progression of stenosis or presenting with high-risk carotid plaque characteristics. There is a critical need for additional research aimed at pinpointing which patient groups would benefit from more frequent surveillance beyond standard annual monitoring. This will ensure tailored, evidence-based care strategies that can potentially enhance patient outcomes by closely monitoring and effectively managing the progression of carotid stenosis.

Addressing the seemingly contradictory notions that carotid revascularization might be reasonable for some patients with ACS yet screening of the general population is not recommended requires a nuanced understanding of high-risk population, risk/benefit ratios, individual patient considerations, and the evolving landscape of medical evidence.

Observational studies and meta-analyses have revealed multiple risk factors for CSVD, CSVD disease progression, and risk for subsequent stroke.^{336–341,354} Modifiable risk factors for CSVD include hypertension, diabetes, tobacco use (smoking), dyslipidemia, OSA, and other factors such as excessive dietary salt intake with inconsistent data.^{336,339,354–357} Given the relationship between these risk factors and stroke risk in the general population and elevated stroke risk in individuals with asymptomatic CSVD, incidental identification of the described radiographic CSVD should prompt clinicians to identify and manage common vascular risk factors to reduce the risk of stroke.

SCIs (majority subcortical) are more common than acute ischemic stroke, with an estimated adult prevalence of 10% to 20%.^337,338 Limited studies have investigated the role of statin therapy for the reduction of stroke risk in individuals with CSVD, including silent subcortical infarcts. Some trials support the role of statin therapy in lowering the risk of CSVD progression compared with placebo; however, the impact on stroke risk was not evaluated.^343–345 In 1 study, investigators used data from a prospective population-based cohort study and a randomized, double-blind placebo-controlled clinical trial to determine the impact of low-dose statin therapy on CSVD progression in adults ≥75 years of age.³⁴² For both the cohort and clinical trial groups, the risk of CVSD progression (primary outcome) was significantly lower with statin treatment without an increase in cerebral microbleeds. For both the cohort study and the clinical trial, the risk of incident stroke was significantly lower in the statin group compared with the nonstatin group (HR, 0.60 [95% CI, 0.37–0.97]; and HR, 0.56 [95% CI, 0.37–0.84], respectively). Although this study demonstrates the potential for benefit, study design (specific population, stroke as secondary outcome) may affect the reproducibility and generalizability of findings.

Given the benefit of antiplatelet therapy in clinical ischemic stroke clinicians may be inclined to treat patients with subcortical infarcts identified incidentally with aspirin or other antiplatelet therapy.²¹ Research to inform this practice is limited, and contemporary practice guidelines are equivocal.^358,359 In SILENCE (Longitudinal Study on Low-Dose Aspirin Versus Placebo Administration in Silent Brain Infarcts), investigators randomized 83 participants ≥45 years of age with silent infarcts to aspirin compared with placebo and followed them up for 4 years.³⁴⁶ The primary outcome was a composite of silent brain infarcts, clinical ischemic stroke and TIA. The aspirin group had 2 events and the placebo group had 9 events at follow-up (P=0.10). The small sample size was a major limitation of this study. Although this study is not sufficient to support or refute the practice of using antiplatelet therapy for management of silent brain infarcts, the high prevalence of CSVD in the older population and evidence of increased harm with aspirin for primary prevention in populations at risk for CSVD suggest the need for increased caution and consideration of patient-specific risk and benefits.^360,361 More data are needed to determine which groups, if any, would benefit from antiplatelet therapy to reduce the risk of ischemic stroke in the setting of incidentally identified subcortical brain infarcts.

Clinical tools are needed to calculate patient-specific risk of ischemic or hemorrhagic stroke in the presence of asymptomatic CSVD, including silent subcortical infarcts.

Although limited data suggest the benefit of BP lowering for the progression of radiographic CSVD, more data are needed to determine the contribution of BP lowering to stroke risk reduction.

There remains unclear benefit of statin therapy (including for specific populations) to reduce the risk of stroke in the presence of asymptomatic CSVD is still unclear.

There remains unclear benefit of antiplatelet therapy (including for specific populations) to decrease the risk of stroke in the presence of asymptomatic CSVD and SCI.

It is unclear whether the presence of cerebral microbleeds should alter the management of other specific risk factors (eg, AF) to balance benefit for ischemic risk reduction with risk of hemorrhagic stroke; this requires further study.

Four recent systematic reviews and meta-analyses including a combined total of 49 observational studies (28 cohort studies, 21 case-control studies) showed a consistent association between migraine and stroke, with RRs ranging from 1.3 to 1.6 for any stroke,^362–364 up to 1.7 for ischemic stroke³⁶² and up to 1.5 for hemorrhagic stroke.³⁶⁵ Migraine with aura confers higher risk for ischemic stroke with a pooled RR of 2.17 (95% CI, 1.78–2.64).³⁶² Cardiovascular risk factors are prevalent in patients with migraine, even at younger ages. A study using insurance claims data from 1.2 million US adults with migraine 18 to 64 years of age found that 18% of them were considered to be at either medium or high risk of vascular events according to comorbidities.³⁶⁸ Incidence rates of ischemic stroke were 3-fold higher in the medium-risk compared with the low-risk group and >16-fold higher in the high-risk compared with the low-risk group.³⁶⁸ Migraine prevalence peaks in the fourth decade of life,³⁷⁴ an age when standard risk calculators may underestimate cardiovascular risk. Thus, although no randomized trials have specifically investigated vascular risk factor modification for primary stroke prevention in individuals with migraine, particular attention to vascular risk factor screening and modification is recommended in these patients.

A systematic review of 12 observational studies found a consistent association between migraine, combined hormonal contraceptive (CHC) use, and ischemic stroke with an absolute risk of 36.9 per 100 000 in women 20 to 44 years with migraine with aura who used CHC compared with 2.5 per 100 000 in women with neither migraine nor CHC use.³⁷¹ Increased risk was also seen in those with migraine without aura. Studies were highly heterogeneous in terms of estrogen dose, reporting of aura, and covariates included, with especially low agreement between studies of migraine without aura. A 2020 systematic review of 4 case-control studies including 12 256 women included only studies of low-dose CHC and found a consistent association of migraine and low-dose CHC use with ischemic stroke, albeit with high heterogeneity between studies.³⁷² In a case-control study of US female patients taking low-dose CHC (1884 cases, 7536 controls), those with migraine with aura who took CHC had 6-fold higher risk of ischemic stroke compared with those without either risk factor.³⁷⁰ Among those with migraine without aura, use of low-dose CHC did not increase stroke risk.³⁷⁰ No increased risk of stroke has been identified in individuals with migraine using progestin-only forms of contraception.³⁸⁰

Determining whether migraine is associated with increased stroke risk in those prescribed estrogen-based therapy for reasons other than contraception such as gender-affirming hormone therapy (HT) or menopausal HT and what estradiol dose is considered safe for these indications;

Determining the relationship among PFO, migraine with aura, and cryptogenic ischemic stroke, including RCTs to test whether specific primary prevention therapy is warranted in this population;

Determining whether preventive treatment of migraine to reduce attack frequency and intensity reduces stroke risk in individuals with migraine; and

Determining whether specific subgroups among those with migraine (eg, those with brainstem aura or persistent aura, pregnant patients, those with other vascular risk factors such as AF and female-specific factors, patients with antiphospholipid antibodies [aPLs]) would benefit from treatment with antiplatelets, anticoagulants, statins, or other medications aimed at primary stroke prevention.

Elevated TCD blood flow velocities are a strong predictor of stroke risk.³⁸¹ In a study of 315 patients who were 3 to 18 years of age with SCD followed up for a mean of 64.4 months, the incidence of stroke in those with nonimaging time-averaged mean of the maximum velocities <170 cm/s was 2%, with velocities 170 to 200 cm/s was 7%, and with velocities >200 cm/s was 40%.^381,382 In 209 children in the STOP-1 and -2 trials (Stroke Prevention Trial in Sickle Cell Anemia), there were 20 strokes, and the last TCD examination before the stroke showed abnormal velocities in all cases.^383,384 TCD has become part of usual care to evaluate stroke risk in children 2 to 16 years of age with SCD. The utility of TCD screening for patients >16 years of age has not been established.³⁹⁵ Notably, abnormally low TCD velocities (eg, <50 to 70 cm/s) in any major vessel may indicate cerebral vasculopathy and warrants additional workup. As shown in Table 9, earlier RCTs implemented broader flow-velocity paradigms with 3 categories, with more recent guidelines^396,397 now implementing 4 categories when considering TCD re-evaluation timing. As outlined, TCD should be repeated at an interval based on the highest flow velocity in the terminal internal carotid artery or proximal middle cerebral artery.

CRCTs for patients 2 to 16 years of age with high-risk TCD is the mainstay of usual care. In STOP-1, 130 children with SCD (mean age, 8 years) and high-risk TCD (velocity >200 cm/s in 2 repeated studies) were randomized to CRCT with a goal hemoglobin S of <30% of total hemoglobin or to observation/standard care.^382,383 The trial was terminated prematurely at a mean follow-up of 20 months because of a marked reduction in strokes in the prophylactic CRCT group (1 stroke in the transfusion arm [2%] versus 10 strokes and 1 ICH in the observation arm [16%]).

SCIs are the most common neurologic injury in SCD, occurring in 39% of children (by 18 years of age)⁴⁰¹ and 53% of adults.⁴⁰² Patients with SCI are at increased risk for recurrent SCI, overt stroke, and cognitive impairment.⁴⁰³ In the SIT Trial (Silent Cerebral Infarct Transfusion), 196 children without high-risk TCD velocities who were found to have SCI on brain MRI were randomized to CRCT or observation. Over a median follow-up of 3 years, CRCT led to a 58% RR reduction of stroke and new/enlarged SCIs.³⁸⁶ Because this is substantially less than the 92% RR reduction with CRCT seen in STOP,³⁸³ the decision to pursue CRCT for children with SCI remains individualized as a result of the significant risks associated with CRCT. Young children may not remain stationary for brain MRI. Sedating agents should be avoided because of the risk of vaso-occlusive complications (eg, through hypoxia/respiratory depression). Therefore, it is advisable to wait until a child is developmentally appropriate before obtaining an unsedated brain MRI. It is important to note that TCD and MRI studies are often discordant in patients with SCD³⁸⁵; therefore, MRI should be done in addition to TCD and should not supplant it as a screening mechanism.

Once a patient has initiated a prophylactic transfusion program, in most cases, this should usually continue indefinitely through 16 years of age with a reduced need for periodic TCD screening. This position is supported by data from the STOP-2 trial, in which children with SCD who had a high risk of stroke according to TCD measurements and who had received transfusions for ≥30 months with normalization of TCD readings were randomized to continue transfusion or not.³⁸⁴ Children with severe stenotic lesions on cerebral magnetic resonance angiography were excluded. The study was stopped early because of safety concerns after 79 children of a planned enrollment of 100 underwent randomization. Among the 41 children in the transfusion-halted group, high-risk TCD velocities redeveloped in 14 patients and stroke occurred in 2 others within a mean±SD of 4.5±2.6 months (range, 2.1–10.1 months) of the last transfusion. Neither stroke nor reversion to high risk of stroke Doppler results occurred in the 38 children who continued to receive transfusions.

Although CRCT is optimal, evidence indicates that hydroxyurea can be used in patients with normalized velocities and no stenosis posttransfusion therapy.³⁸⁷ In the Créteil newborn cohort, 45 patients with SCD whose TCD velocities normalized on CRCT were switched to hydroxyurea and followed up with quarterly TCDs for a mean of 3.4 years. Reversion to abnormal TCD velocities on hydroxyurea occurred in 13 of 45 patients (28.9%); transfusion was quickly reinitiated in these patients. No patient developed stroke or SCI while on hydroxyurea.³⁸⁷ In the TWiTCH trial (TCD With Transfusions Changing to Hydroxyurea), 121 children with SCD on CRCT (1-year minimum) with continued abnormal TCD flow velocities (≥200 cm/s) but no severe vasculopathy were randomized to continue transfusions (n=61) or transition to hydroxyurea after 1 year of CRCT (n=60).³⁸⁸ The primary end point was the 24-month TCD velocity; however, at the first scheduled interim analysis, noninferiority was demonstrated (P=8.82×10⁻¹⁶), and the study was terminated early. Iron burden (ie, ferritin), an established side effect of CRCT, was also beneficially decreased (P≤0.001). Ideally, transition to hydroxyurea should occur after transfusion therapy for the longest possible interval and with transfusion continued during escalation to the maximum tolerated dose of hydroxyurea.

Not all patients with SCD at risk for stroke will have access to or will be able to tolerate indefinite CRCT. The recent trials of children with SCD in Nigeria (SPRING [Stroke Prevention Trial in Nigeria]) and in Tanzania (SPHERE [Stroke Prevention With Hydroxyurea Enabled Through Research and Education]) demonstrated that, in such situations, daily hydroxyurea is an effective alternative for primary stroke prevention.^389,390 In both studies, TCD velocities decreased after initiation of hydroxyurea, and the stroke incidence rate was comparable to historical controls. The BABY HUG trial (Pediatric Hydroxyurea Phase III Clinical Trial)⁴⁰⁴ and SCATE trial (Sparing Conversion to Abnormal TCD)⁴⁰⁵ further suggest that hydroxyurea may mitigate rises in TCD velocities in low-risk children. In the Cre Créteil newborn cohort,³⁸⁷ 24 patients with SCD who had normalized velocities and no stenosis posttransfusion therapy were treated with bone marrow transplantation from a genoidentical sibling donor. During the mean posttransplantation follow-up of 3.5 years (range, 0.3–11.0 years), no patient experienced a stroke or an SCI. Additional supporting data for bone marrow transplantation are derived from a study in which 67 children with elevated TCD velocities were treated with sibling-donor bone marrow transplantation or standard care (CRCT with the option to switch to hydroxyurea after the first year).⁴⁰⁶ Follow-up at 3 years showed no strokes or deaths in either group, no differences in cognitive performance, fewer SCIs, and a greater reduction in TCD velocities with bone marrow transplantation.

Although TCD can be used to identify children who are at high risk of stroke and would benefit from transfusion therapy, improved prediction algorithms incorporating additional parameters such as anterior cerebral artery velocity, blood biomarkers, genetic variations, and nocturnal oxygen saturation should be developed and evaluated.

Studies evaluating how antithrombotics, antihypertensive agents, and statins may influence primary stroke prevention in SCD are lacking.

Most studies for primary (and secondary) stroke prevention in SCD were conducted in pediatric populations. Hence, a knowledge gap exists concerning whether transfusion or hydroxyurea recommendations in the pediatric population remain appropriate in the adult SCD population.

As seen in STOP-2, even those whose risk of stroke decreases with transfusion therapy on the basis of TCD criteria have an ≈50% probability of reverting to high-risk TCD velocities or having a stroke if transfusion therapy is discontinued. Hydroxyurea may be an appropriate maintenance therapy. Further studies on optimal dosing and duration are required.

Although there is some benefit of long-term transfusion therapy for children with SCI, the number needed to treat to see reduction of strokes is large, and the risk and cost of regular transfusions are high. Investigation of patient and SCI characteristics that portend the highest risk of stroke may help to determine which patients are most likely to benefit from long-term transfusions. Studies evaluating MRI frequency in pediatric and adult populations are lacking.

The role of newer therapies for SCD, including voxelotor, crizanlizumab, and L-glutamine, in stroke prevention warrants further investigation.

RCTs for primary stroke prevention studies in adults with SCD and within specific populations (eg, pregnancy) are lacking.

Sickle cell trait has been inconsistently associated with ischemic stroke, although hypercoagulability and chronic kidney disease, among other factors, may modify this risk; additional studies are warranted.

Moyamoya vasculopathy as a late complication of SCD requires further study.

The US Food and Drug Administration recently approved exagamglogene autotemcel and lovotibeglogene autotemcel, the first gene therapies for the treatment of SCD in patients ≥12 years of age. Studies to determine their effects on stroke risk are required.

CADASIL is an inherited CSVD caused by mutations in the NOTCH3 gene. Although there is no cure, epidemiological data suggest that stroke risk in CADASIL is associated with modifiable stroke risk factors. Prospective cohort studies had demonstrated that hypertension, active cigarette smoking, and total pack-years of exposure are each associated with increased stroke risk in patients with CADASIL.^407,408 Smoking may also be associated with earlier onset of stroke or TIA.⁴¹⁰ A large prospective cohort study of 973 NOTCH3 carriers showed that cardiovascular risk burden was associated with increased risk for stroke.⁴⁰⁹ Together, these data suggest indirectly that cardiovascular risk factor control may decrease stroke risk in patients with CADASIL. However, no randomized trials have evaluated the effect of vascular risk factor modification on stroke risk in individuals with CADASIL. Despite the lack of high-quality data on the effect of interventions such as BP control and smoking cessation, these measures are nonetheless recommended, at minimum to mitigate their deleterious effects on atherosclerotic vascular disease, which likely compounds stroke risk in those with CADASIL.⁴¹² Preventive therapies of proven benefit in other patients with stroke, including antiplatelet agents, have not been shown to affect the incidence of stroke in individuals with CADASIL.^411,412

HHT is an autosomal dominant vascular dysplasia that typically manifests as epistaxis, telangiectasias, and vascular malformations of the brain, lungs, and liver. PAVMs are present in nearly half of patients with HHT and are associated with embolic complications, including ischemic stroke and brain abscesses. In a cross-sectional study of 108 patients with HHT, those with PAVM had more total embolic complications and more stroke/TIA, with a 7-fold increased odds for embolic events in an adjusted model.⁴¹³ In a multivariate analysis of a retrospective cohort study of 353 patients with HHT, PAVMs were independently associated with SCI.⁴¹⁴ In a cross-sectional study of 75 individuals with PAVMs, those with multiple PAVMs had greater odds of having stroke compared with those with a single PAVM.⁴¹⁵ Data from observational research have demonstrated an association between PAVM embolization and reduced rate of stroke, but the findings have not been confirmed in clinical trials.^416–421 Therefore, a multidisciplinary team with HHT expertise, consisting of at minimum specialists in pulmonology, interventional radiology, and neurology/neurosurgery, should weigh the risks and benefits of intervention for patients with PAVMs.

Fabry disease is an X-linked lysosomal storage disorder resulting in α-galactosidase deficiency and glycosphingolipid accumulation in various tissues, including vascular endothelium. ERT with recombinant α-galactosidase A (agalsidase β) has been shown in an RCT to reduce microvascular endothelial globotriaosylceramide deposits in the kidneys, heart, and skin in patients with Fabry disease.⁴²² However, the impact of ERT on stroke in Fabry disease remains unclear. In an open-label trial extension of the aforementioned trial in which 58 patients with Fabry disease received ERT, 8.6% had a stroke during the 30- to 36-month follow-up, which was lower than the historical rate of 13.7%.⁴²⁴ In a subsequent randomized placebo-controlled trial, ERT delayed the first clinical event (renal, cardiac, or cerebrovascular events or death), but low stroke incidence (n=2) limited assessment of the impact of ERT on stroke.⁴²³ A meta-analysis of 77 cohort studies showed that ERT is associated with fewer composite complications (renal, cardiovascular, cerebrovascular).⁴²⁵ In contrast, in a prospective observational study, stroke rates did not differ between patients receiving ERT and matched patients with Fabry disease not on ERT.⁴²⁶ Another prospective cohort study of 362 patients with Fabry disease similarly failed to show a difference in stroke rate based on type or timing of ERT initiation.⁴²⁷

Disease-modifying therapies, including gene therapies, for genetic syndromes that place affected individuals at high risk of stroke need to be developed, and stroke-specific outcomes must be evaluated. The roles of vascular risk factor modification, antiplatelet therapy, and calcium channel blockers in stroke prevention in patients with CADASIL are unknown.

In patients with HHT, the optimal screening method for PAVM and the age at which screening should begin are unknown.

In individuals with HHT and PAVMs, PAVM treatment selection criteria and optimal approach need refinement.

Further research on iron-deficiency anemia as a modifiable stroke risk factor in individuals with HHT and PAVMs may lead to strategies to reduce stroke.

The role of screening and treatment for brain arteriovenous malformations in individuals with HHT remains undefined.

Future studies that evaluate the risk of ICH related to female HT in patients with familial cerebral cavernous malformations should stratify by type of HT and account for risks of withholding contraception.

Research on the role of antithrombotic agents, propranolol, statins, and rhoA kinase inhibitors in individuals with familial cerebral cavernous malformations is needed.

Patient selection for and timing of administration of tumor necrosis factor inhibitors in individuals with deficiency of adenosine deaminase 2 warrant further investigation.

Given the risk of hemorrhagic stroke in patients with deficiency of adenosine deaminase 2 and type IV collagen–related disorders, the risks of antithrombotic agents in these individuals should be systematically evaluated.

Additional research is needed to better understand how genetic drivers influence stroke risk through vascular risk factors, common subtypes of stroke, and uncommon familial stroke syndromes.

Further research is needed in pharmacogenomics to improve and personalize stroke prevention strategies.

Inflammation plays a key role in atherosclerosis,⁴⁵⁹ and colchicine is an anti-inflammatory medication that has been tested in a number of CVD trials. In COLCOT (Colchicine Cardiovascular Outcomes Trial),⁴⁶⁰ patients with recent MI (30 days) and planned intensive statin therapy were randomized to low-dose colchicine (0.5 mg daily) versus placebo. Colchicine was associated with a significant reduction in the primary end point (cardiovascular death, cardiac arrest, MI, stroke, angina leading to revascularization) and in the planned secondary end point of stroke (HR, 0.26 [95% CI, 0.10–0.70]). The absolute risk difference after a mean follow-up of 23 months was 0.6%. The potential benefit of colchicine for disease prevention with chronic coronary artery disease was evaluated in the low-dose colchicine trial LoDoCo2 (Low-Dose Colchicine vs. Placebo in Patients With Chronic Coronary Disease).⁴⁶¹ In this study, colchicine was associated with a significant reduction in risk for the primary end point (ischemia-driven coronary revascularization, MI, ischemic stroke, cardiovascular death) but a nonsignificant reduction in risk for ischemic stroke (HR, 0.66 [95% CI, 0.35–1.25]).

Asymptomatic aPL carriers can be defined as individuals with ≥1 aPLs without history of thrombosis, obstetrical APS, or SLE. Use of aspirin (75–100 mg daily) for primary stroke prophylaxis was first supported by results of a meta-analysis of 7 observational studies of 460 asymptomatic aPL carriers showing that the risk of first thrombosis (arterial and venous combined) was reduced by half in those who used aspirin compared with those who did not use aspirin (OR, 0.50 [95% CI, 0.25–0.99]). Most patients had high-risk aPL profiles (see definition in Table 11) and a paucity of traditional cardiovascular risk factors.⁴⁶² Associations of similar magnitude were present in a smaller 5-study meta-analysis of individual patient data derived from these same studies.⁴⁶³ Here, subgroup analysis revealed a greater protective effect of aspirin against arterial thrombosis in asymptomatic aPL carriers (HR, 0.43 [95% CI, 0.20–0.93]) compared with venous thrombosis (HR, 0.49 [95% CI, 0.22–1.11]). Both meta-analyses revealed consistent effects and clear benefits. Although this evidence is largely from observational studies, other professional societies have drawn similar conclusions given the likelihood of benefit and low risk of adverse events of this intervention.⁴⁷²

In patients with SLE and high-risk aPL (see definition in Table 11), a subanalysis of 8 mostly observational studies from a larger meta-analysis provides supportive evidence for the use of aspirin (75–100 mg daily) for primary stroke prevention. In this analysis, the risk of first thrombosis (arterial and venous combined) was reduced by almost half among patients treated with aspirin compared with patients not treated with aspirin (OR, 0.55 [95% CI, 0.31–0.98]) without major bleeding events.⁴⁶² In a follow-up study evaluating individual patient data in this clinical setting, there was greater benefit of aspirin use in arterial thrombosis prevention (HR, 0.43 [95% CI, 0.20–0.94]) compared with venous thrombosis (HR, 0.49 [95% CI, 0.21–1.10]).⁴⁶³ It is notable that this association was independent of the use of hydroxychloroquine, suggesting that aspirin offers additional benefit in this patient group. Patients with a high-risk aPL profile made up the majority (but not all) of patients in these studies. Although there was heterogeneity between the studies, the direction of effect was consistent, favoring aspirin use.^463,463

In patients diagnosed with APS after an initial unprovoked venous thrombosis, recurrence rates without VKA therapy are high, so anticoagulation should be continued lifelong. Data from an RCT⁴⁶⁴ reporting exclusively patients with venous events and pooled data from several studies showed no additional benefit of a target international normalized ratio of 3 to 4 (versus 2–3).^457,464 Notably, the target international normalized ratio in the high-intensity groups was inconsistently achieved, potentially inferring the nonstatistically increased bleeding risk in those assigned to receive high-intensity warfarin (HR, 2.18 [95% CI, 0.92–5.15]).⁴⁵⁷ When anticoagulation has been compared with aspirin, an international normalized ratio of 2 to 3 has been shown to be effective.^458,464,474 Use of direct oral anticoagulants has been suggested in patients not able to achieve a target international normalized ratio despite good adherence to VKA or in those with contraindications to VKA (eg, allergy or intolerance to VKA). Unfortunately, there is limited evidence on their safety and effectiveness in patients with APS.⁴⁷⁶ A recent meta-analysis of 4 open-label RCTs demonstrated that patients with thrombotic APS randomized to direct oral anticoagulants compared with VKAs had increased odds of arterial thrombosis (OR, 5.43 [95% CI, 1.87–15.75]; P≤0.001), especially stroke,⁴⁶⁵ but without significant differences in subsequent VTE risk or major bleeding.

Rheumatoid arthritis is a chronic autoimmune disorder that induces a globally elevated inflammatory response that affects primarily joints, also causing increased atherosclerosis that contributes to elevated MI and stroke risks. Treatment of rheumatoid arthritis focuses on reducing inflammation systemically. Some evidence supports a potential beneficial impact of statins on rheumatoid arthritis disease activity, attributable to their anti-inflammatory and immunomodulatory properties. This premise was recently explored in a large systematic review and meta-analysis of 40 307 patients.⁴⁶⁶ Using a total of 6 studies, including 1 double-blind, placebo-controlled RCT,⁴⁶⁷ 4 propensity score–matched cohorts, and 1 observational study, that study showed that the rate of major adverse cardiovascular events (nonfatal MI, nonfatal presumed ischemic stroke, TIA, any coronary or noncoronary revascularization, or cardiovascular death) was lower in patients with rheumatoid arthritis receiving statin therapy compared with those not on statin (OR, 0.67 [95% CI, 0.51–0.89]; P=0.005).⁴⁶⁶

The primary prevention of thrombosis with aspirin (75–100 mg daily) in women with a history of obstetric APS without SLE was addressed in a meta-analysis including 5 observational studies.⁴⁶² The pooled OR for first thrombosis (arterial and venous) associated with the use of aspirin was 0.25 (95% CI, 0.10–0.62).⁴⁶² A retrospective study evaluating the subgroup of female patients with APS diagnosed with APS solely as a result of pregnancy morbidity demonstrated the protective effect of aspirin (75–100 mg daily), with 59% of non–aspirin-treated and 10% of aspirin-treated patients experiencing further aspirin-related clinical events (P=0.00006).⁴⁶⁸ Although the data remain limited, prophylactic treatment with aspirin (75–100 mg daily) in nonpregnant women with a history of obstetric APS can be considered but only after adequate risk/benefit evaluation (ie, aPL profile, coexistent traditional cardiovascular risk factors, intolerance/contraindication to aspirin).

Less evidence is available on the use of aspirin (75–100 mg daily) in patients with a history of obstetric APS without SLE and low-risk aPL profile (see definition in Table 11); however, data from 2 cohort studies indicate that the use of aspirin was associated with a lower risk of arterial (and venous) thrombosis in this clinical setting.^469,470 In aggregate, the results of these studies and others^462,463 demonstrate that up to one-third of patients with SLE with aPL (inclusive of all profiles) develop thrombotic complications at sometime during the disease course. The prevalence of clinical thrombotic events was significantly higher when all 3 types of aPL were present compared with only anticardiolipin–positive cases.^469,470

Whether, in the absence of venous thrombosis, there is an indication for antithrombotic therapy for primary stroke prevention;

Whether, in the absence of venous thrombosis, the presence of a PFO influences the risk of primary stroke and what the optimal primary prevention strategy is; and

Adequately powered studies to study each trait individually; if a heightened risk of primary stroke is confirmed, clinical trials to evaluate the optimal antithrombotic treatment to reduce risk.

Further clarification of whether direct oral anticoagulants (all or some) are less effective than warfarin to reduce the risk of stroke in this population and

The role of dual antiplatelets in APS, with reports suggesting a possible preventive role or in addition to anticoagulation.

Patients with cancer are variably predisposed to hypercoagulability but may also be at a higher risk for bleeding with antithrombotic therapies. In these patients, the benefit of antithrombotic therapies for primary stroke prevention is not well established, and further research is needed.

The need for anticoagulation to prevent stroke in different types of cancer is not known and should be a focus of further research.

Biomarkers (eg, D-dimer) and mRNA expression profiles to predict future stroke risk in the setting of cancer hold promise, but additional research is required.

Although low-molecular-weight-heparin is often used in patients with cancer and stroke to prevent thromboembolic complications, the potential benefit for primary cancer-related stroke prevention is unknown.

Although there is some evidence that direct oral anticoagulants may be used as an alternative to low-molecular-weight-heparin, further studies are needed to clarify indications, risks, and benefits.

A prospective nationwide population-based study assessed tooth scaling and CVD.⁴⁸⁷ In this study, 10 887 subjects ≥50 years of age without prior MI or ischemic stroke who had received full-mouth or localized tooth scaling were compared with 10 989 subjects who did not have tooth scaling. During an average follow-up of 7 years, the tooth scaling group had lower stroke risk (HR, 0.85 [95% CI, 0.78–0.93]). A retrospective study included 510 762 subjects ≥20 years of age with PD and 208 674 subjects without PD from 2000 to 2010.⁴⁸⁸ Those who received dental prophylaxis or intensive treatment for PD had significantly lower stroke risk than the non-PD group (HR, 0.78 [95% CI, 0.75–0.81]; HR, 0.95 [95% CI, 0.91–0.99], respectively). The individuals with PD without treatment had the highest risk of stroke (HR, 2.17 [95% CI, 1.64–2.87]). The ancillary dental ARIC study (Atherosclerosis Risk in Communities) assessed dental care use.⁴⁸⁶ In a total of 6736 subjects classified into 7 periodontal profile classes, all 7 levels showed a trend toward an increased stroke risk. PD was significantly associated with ischemic stroke (HR, 2.2 [95% CI, 1.3–3.8]).

A meta-analysis of 7 RCTs compared treatment with escalated-dose (intermediate- or full-dose) anticoagulation versus prophylactic-dose anticoagulation in 5154 patients with COVID-19 requiring hospitalization.⁴⁸⁹ The primary outcome was all-cause death (RR, 0.96 [95% CI, 0.78–1.18]). There were 4 RCTs with the secondary outcome of stroke. The incidence of stroke was 0.5% (11/2296) in the escalated-dose group and 0.5% (12/2195) in the standard-dose group, resulting in no significant differences between regimens (RR, 0.94 [95% CI, 0.43–2.09]; I²=0%). Stone et al⁴⁹⁰ conducted a randomized, 3-arm, open-label, active-controlled, multinational multicenter trial between August 26, 2020, and September 19, 2022. There were 3398 noncritically ill patients hospitalized with COVID-19 randomized to prophylactic-dose enoxaparin (n=1141), full-dose enoxaparin (n=1136), or full-dose apixaban (n=1121). The 30-day primary composite outcome of all-cause mortality, requirement for intensive care unit level of care, systemic thromboembolism confirmed by imaging or requiring surgical intervention, or ischemic stroke occurred in 13.2% of patients in the prophylactic-dose group and 11.3% of patients in the combined therapeutic-dose groups (HR, 0.85 [95% CI, 0.69–1.04]; P=0.11). There were no stroke cases in the prophylactic-dose group and only 1 stroke case in the full-dose group.

The use of the influenza vaccine for primary stroke prevention in adults remains unclear. Systematic reviews and meta-analyses of observation studies (cohort, case-control, and nested case studies) suggested a benefit of influenza vaccination on lowered stroke risk. However, other meta-analyses and RCTs reported no benefit. Further studies and clinical trials are needed to determine whether influenza vaccination could be used as a public health strategy for primary stroke prevention.

It has been challenging to study the effects of SARS-CoV-2 on stroke risk because of the progressive decline of the pandemic, resulting in several studies ending early with inadequate sample size. Further research is needed to elucidate whether stroke is a complication specifically of SARS-CoV-2 infection or if it is an effect as seen with other infections and whether treatment of SARS-CoV-2 would lead to a lower risk of future stroke.

Heavy alcohol use, with a dose-dependent response, is consistently associated with increased stroke risk (ischemic and hemorrhagic).^506–510 Although smoking or ingestion routes are not specified, hospitalizations for stroke in younger (18–38 years of age) cannabis users is increasing,⁵¹³ often in the absence of identifiable cardiovascular risk.⁵¹⁴ Similarly, when stroke-related covariates, including tobacco smoking, are controlled for, stroke/TIA risk is elevated with more frequent than once-weekly cannabis use across age ranges (adjusted incidence rate ratio, 4.7 [95% CI, 2.1–10.7])⁵³⁵ as supported by systematic review evidence⁵¹¹ and representative surveillance data confirming the dose-dependent (risk behavior–adjusted) relationship.⁵¹² Meta-analyses identify stroke as associated with opioid misuse (adjusted OR, 2.27).⁵¹⁶ A systematic review examining misuse of prescription and other amphetamine-type stimulants identifies increased stroke risk (2 cohort studies showing adjusted RRs of 1.6 and 3.4) and increased hemorrhagic stroke risk in former users (adjusted RR, 2.3)].⁵¹⁷ Longitudinal data show that individuals who use methamphetamines have higher stroke incidence, particularly hemorrhagic stroke (HR, 2.09).⁵¹⁸ Registry data (>3 million hospital discharges) identify hemorrhagic stroke risk associated with amphetamine (OR, 4.95) and cocaine (OR, 2.33) use and cocaine-associated ischemic stroke risk (OR, 2.03).⁵¹⁵ Observational data show cocaine use or Khat chewing (common in Middle Eastern and African cultures) is associated with increased stroke risk in existing cardiac conditions.^519,520,537

Although current evidence cannot directly support interventions addressing substance misuse or substance use disorders for stroke prevention, mitigating their known risks for premature atherosclerosis⁵⁴⁷ and untimely death^548–550 alongside associated stroke risk is reasonable. Screening, Brief Intervention and Referral to Treatment (SBIRT)⁵²⁵ programs show lower prevalence of heavy drinking (43.4%) and illicit drug use (75.8%) at 6 months. Systematic reviews provide moderate evidence^522–524 supporting brief interventions for harmful drinking in men and women⁵²² whereby multicontact primary care interventions (10–15 minutes) can result in 11% more adults drinking within the recommended limits.⁵²³ Behavioral techniques can include feedback on alcohol-related harm, clarification of low-risk consumption levels, benefits of intake reduction, motivational enhancement, analysis of high-risk drinking situations, and coping strategies and personal reduction planning. In individuals with an alcohol use disorder, evidence supports pharmacotherapies (acamprosate, disulfiram, baclofen, and oral naltrexone) or combined pharmacotherapy and evidence-based behavioral therapy (eg, cognitive behavioral therapy, motivational enhancement therapy)^526–528 for improving abstinence and heavy drinking. Similarly, in noncoerced adults with an alcohol use disorder, peer-led treatments (eg, Alcoholics Anonymous) and professionally delivered treatments with peer-led involvement (eg, Twelve-Steps Facilitation) can improve rates of continuous abstinence at up to 36 months.⁵²¹ Evidence-based agonist treatment strategies for maintenance or suppression of opioid use disorders include buprenorphine and methadone.^529–531

Additional studies are required to understand the effects of the recent INTERSTROKE study association between moderate alcohol consumption and ICH and regional differences identified in low alcohol intake and overall stroke risk.

Longitudinal studies are required to understand the effects of moderation/cessation of alcohol misuse on future stroke risk.

Implementation studies are required to identify optimal strategies and age range targets to increase the delivery of alcohol misuse screening and brief counseling or intervention across various practice and resource settings.

A standardized dose for cannabis use is required (similar to the standard drink or cigarette) to allow nonmedical cannabis use to be routinely recorded in national and local surveillance systems to monitor the effects of use and use patterns on CVD risk, including stroke.

Longitudinal studies addressing use frequency and tetrahydrocannabinol concentrations and duration of exposure are required to allow a more nuanced understanding of the long-term clinical cerebrovascular implications of cannabis use.

Fundamental and clinical research on the potential short- and long-term health consequences of synthetic cannabinoid products on cerebrovascular function is required.

The current association between substance (cocaine, heroin, methamphetamines or prescription medications [stimulants or opioids]) misuse and stroke risk relies on hospital-based stroke prevalence data, which are subject to significant biases. Well-conceptualized, prospective longitudinal studies are required that control for confounding factors, including multiple substance use.

Large, well-conceptualized studies are required to separately examine stroke types (ischemic, intracerebral hemorrhagic, subarachnoid hemorrhagic) and their association with the misuse of different substances.

No data directly link reduced alcohol and substance use with reduction in stroke risk, and this gap in knowledge should be filled.

Acute, severely elevated BP in pregnancy or the puerperium constitutes a neurological emergency because of its high risk of hemorrhagic stroke. Observational data support prompt BP-lowering treatment for pregnant and postpartum patients with severe hypertension (see Table 12 for definitions) to prevent fatal maternal stroke. A 30-year retrospective study of 347 fatal maternal strokes in a UK death registry, half of which were due to ICH, found that delayed control of systolic hypertension was a major contributing factor.⁵⁵¹ Four other retrospective case series including a total of 157 maternal strokes, the majority of which were hemorrhagic, consistently found that delayed or ineffective treatment of severe hypertension contributed significantly to maternal morbidity and mortality.^552–555 A 2013 Cochrane review of treatment for acute, severe hypertension elevations postpartum found no significant difference in efficacy among fast-acting oral nifedipine, intravenous labetalol, or intravenous hydralazine, and all are considered reasonable options.⁵⁵⁶ A 2014 systematic review of 16 fair-quality RCTs found that oral nifedipine, labetalol, and methyldopa are suitable options for treatment of severe hypertension in pregnancy or postpartum.^557,558

HDP, including chronic hypertension, increase maternal stroke risk.⁵⁶⁰ Those with chronic hypertension who develop superimposed preeclampsia and those with sustained BP >140/90 mm Hg are at higher risk.^559,560,564 In a retrospective study of 44 million US deliveries, stroke occurred in 1 in 370 deliveries among people with superimposed preeclampsia compared with 1 in 5000 deliveries in those without hypertension.⁵⁶⁰ Limited data suggest that tighter BP control could reduce maternal stroke risk without increasing fetal risk. A meta-analysis of 12 studies including 251 172 pregnant individuals found that SBP >140 mm Hg or DBP >90 mm Hg was associated with a pooled risk ratio of 2.64 (95% CI, 1.08–6.46) for maternal stroke compared with BP <120/80 mm Hg.⁵⁶³ The randomized CHAP trial (Cardiovascular Health Awareness Program) found that among pregnant women with chronic hypertension, tighter BP control (goal <140/90 mm Hg) reduced the risk of superimposed preeclampsia compared with standard care (goal <160/110 mm Hg) with no increased fetal risk.⁵⁶⁵ Of note, the optimal lower BP limit during pregnancy for fetal well-being is unknown.⁵⁶⁶ A retrospective study of 239 454 US patients hospitalized with severe preeclampsia found that an increase over a 10-year period in the proportion of patients receiving any antihypertensive correlated with a >50% reduction in stroke during delivery hospitalizations over the same time period.⁵⁶¹ A French nationwide registry-based study found that treatment with any antihypertensive medication (compared with no treatment) was associated with decreased maternal stroke risk in women with chronic hypertension.⁵⁶²

Research to define the optimal target BP during pregnancy and postpartum to prevent pregnancy-associated stroke in individuals with chronic hypertension;

Mechanistic and translational research to understand the unique pathophysiology of pregnancy-associated stroke, including investigation of the relationship between pregnancy and acute vasculopathies such as reversible cerebral vasoconstriction syndrome and cervical artery dissection;

Creation of national registries for pregnancy-related stroke to aid in research into risk factors for maternal stroke and early clinical warning signs;

Investigation of predictors of maternal stroke, including blood-based or imaging biomarkers, BP trajectories, or other clinical characteristics, as well as potential protective factors such as the use of low-dose aspirin (currently recommended by the USPSTF during pregnancy to reduce the risk of preeclampsia in high-risk patients);

Clinical trials to test optimal BP management strategies for peripartum and postpartum stroke prevention in those with HDP, including chronic hypertension; because of the rarity of the stroke outcome, potential surrogate end points (eg, neuroimaging) could be considered; and

Further investigation of the management of unruptured intracranial vascular lesions in patients who are or plan to become pregnant, which constitutes an area of uncertainty.

APOs affect 1 in 5 US pregnancies^634,638 and are associated with increased risk of developing cardiovascular risk factors and symptomatic CVD,⁶³⁹ including stroke specifically. Eleven recent systematic reviews and meta-analyses investigated the impact of APOs individually and collectively on long-term maternal stroke risk.^{609,614,615,617,619,640–645} A strong, consistent association between APO history and stroke risk was identified, with risk estimates from 20% to 30% higher in those with a history of gestational hypertension or small-for-gestational-age infant to nearly 4-fold higher in those with a history of superimposed preeclampsia (Table 13). Long-term stroke risk increases when multiple pregnancies are affected,^616,618,635 suggesting a dose response. It is unknown whether APOs directly increase stroke risk or indicate an accelerated trajectory of cerebrovascular disease due to genetic or environmental factors.^646,647 Prepregnancy health influences the development of APOs,⁶⁴⁸ and shared risk factors likely account for some of the association between APOs and long-term stroke risk. Thus, despite the fact that no randomized trials have specifically evaluated the effect of vascular risk factor modification on stroke risk in individuals with a history of APOs, it is recommended that clinicians, including obstetrician-gynecologists and other primary care clinicians, screen parous adults for a history of APOs and, if present, identify and treat modifiable stroke risk factors using the Life’s Essential 8 approach (see Sections 4.1–4.8).

A retrospective cohort study of 2 227 711 women without preexisting chronic hypertension found, using French health registry data, that those with preeclampsia or gestational hypertension had nearly 8 or 9 times more chronic hypertension, respectively, by 10 years postpartum compared with those without HDP. After adjustments for other cardiovascular risk factors, 10-year risk for stroke doubled after preeclampsia (adjusted HR, 2.0 [95% CI, 1.8–2.2]) and was 70% higher after gestational hypertension (adjusted HR, 1.7 [95% CI, 1.5–1.9]).⁶¹³ In a prospective cohort of 4484 US women followed up from early in their first pregnancies, those with any APO had double the risk of incident hypertension within 2 to 7 years of the index pregnancy.⁶³⁸ In a Finnish cohort of 144 306 parous women with a median follow-up time of 31 years, those with a history of recurrent APOs (>1 pregnancy affected) had more than double the odds of stroke before 45 years of age compared with those with no APO history.⁶¹⁸ No randomized trials have specifically evaluated BP reduction as a strategy to reduce long-term stroke risk in individuals who experience APOs. However, given the consistently observed associations between APOs and early-onset hypertension and the strong association between untreated hypertension and stroke risk (see Section 4.6, Blood Pressure), screening of BP and identification and treatment of hypertension are recommended for adults with a history of APOs, regardless of age, to reduce the risk of stroke in midlife and later.

Mechanistic, translational, and high-quality population-based research to determine the effects of normal pregnancy, age at first pregnancy, and number of pregnancies on long-term maternal cerebrovascular risk;

Studies to determine whether the relationship between APOs and future stroke risk is directly or indirectly causal;

Randomized clinical trials to determine whether specific primary prevention strategies such as antiplatelets, statins, or targeted BP management can reduce long-term stroke risk in those who have experienced HDP and other APOs;

Research to understand the relationship between infertility, including treatment for infertility, and long-term stroke risk;

High-quality population-based research to investigate the relationship between lactation and long-term maternal cerebrovascular risk; and

Research to characterize stroke risk in the offspring of pregnancies complicated by APOs.

Recent high-quality observational studies and meta-analyses have consistently found an increased risk of stroke among those with endometriosis. In a population-based study in Taiwan, those with endometriosis had a 16% increased risk (95% CI, 2%–31%) of total stroke.⁶⁵⁰ A population-based matched cohort study in the United Kingdom found an increased risk of cerebrovascular disease among those with endometriosis (HR, 1.19 [95% CI, 1.04–1.36]).⁶⁵² The Nurses’ Health Study in the United Stated found an increased risk of stroke (HR, 1.34 [95% CI, 1.10–1.62]) among individuals with laparoscopically confirmed endometriosis⁶⁵¹; 39% of the associated stroke risk was mediated by hysterectomy/oophorectomy and 16% by HT.⁶⁵¹ A meta-analysis of first incident stroke data from 3 longitudinal cohort studies of stroke^650–652 found an HR of 1.17 (95% CI, 1.7–1.29) for cerebrovascular disease.⁶⁵³ A recent Canadian population-based cohort study found an increased risk of stroke among those with endometriosis (RR, 1.11 [95% CI, 1.02–1.20]).⁶⁴⁹ Mendelian randomization analyses further support a causal relationship between endometriosis and stroke.⁶⁵⁷ Thus, asking about a history of endometriosis can be useful in evaluating stroke risk.

High-quality observational studies have consistently found a 16% to 34% increased risk of stroke among those with endometriosis.^649–653 Patients with endometriosis are at higher risk of hypertension, elevated cholesterol, and inflammation.^654,655 Diagnosis of endometriosis usually occurs during the reproductive years,⁶⁵⁴ identifying an at-risk group in early adulthood in whom prevention, identification, and modification of vascular risk factors may reduce the risk of stroke; however, definitive data are lacking.

Whether adding endometriosis to risk prediction models improves cardiovascular prediction;

Whether interventions to treat cardiovascular risk factors reduce risk among those with endometriosis;

Whether early treatment for endometriosis (eg, medications or surgical removal) works as a cardiovascular preventive strategy; and

Whether polycystic ovarian syndrome is a potentially important risk factor for stroke in individuals with ovaries, which should be a future area of research.

Research has identified a direct, linear relationship between a higher dose of estrogen in CHC and an increased risk of stroke. Conversely, lower estrogen content is associated with reduced risk.^{658,659,661,662,665,670} Women who used a CHC with <50 μg estrogen have a lower risk of stroke compared with women who use preparations with a higher estrogen content (RR, 2.08 [95% CI, 1.55–2.8] versus RR, 4.53 [95% CI, 2.17–9.5]; P=0.01).⁶⁶⁸ More recently, 1 group performed a meta-analysis using 6 cohorts and 12 case-control studies to evaluate stroke risk associated with every 10 μg estrogen use (OR, 1.19 [95% CI, 1.16–1.23]). These risks were consistent for both ischemic stroke (OR, 1.24 [95% CI, 1.17–1.22]) and hemorrhagic stroke (OR, 1.10 [95% CI, 1.04–1.16]).⁶⁵⁹ Duration of use increases stroke risk after 1 year⁶⁵⁸ for every 5 years of use.

Contraceptive choice is affected by many medical and personal factors for the patient. Shared decision-making is recommended to weigh the benefits and risks of these choices. For example, the absolute stroke risk when evaluating the 6-fold increase stroke risk associated with migraine with aura using CHC is different for a person 18 years of age than for a person 45 years of age (20.4 versus 386.4 events per 100 000 person-years).^370,672 Consideration of the risk of stroke from contraceptive choice should also be balanced by consideration of the effectiveness of each contraception option (Table 14) and the stroke risk associated with pregnancy. Similar risk factors increasing the risk of stroke from CHC (ie, hypertension, smoking, migraine with aura) also increase the stroke risk from pregnancy. A thoughtful discussion of absolute risk between the health care professional and patient can improve patient-centered care, patient engagement, and stroke reduction.

CHC can synergistically increase existing stroke risk factors. Lidegaard et al⁶⁷² demonstrated an exponential stroke risk increase with age using CHC, growing from 3.4 to 64.4 events per 100 000 person-years for individuals 15 to 49 years of age. Further studies have confirmed an increased risk related to CHC use with increased age, especially when 35 years of age is used as a cutoff.^660,665,670 CHC synergistically increases the risk of hypertension up to 4-fold, smoking up to 3-fold,⁶⁷⁰ and migraine with aura up to 6-fold.³⁷⁰ Higher estrogen doses significantly increased the risk of stroke in those who smoke.⁶⁶⁸ Progestin-only contraception in all routes is not associated with an overall increased stroke risk.^{663,664,666,667,669} However, progestin-only contraception use in the setting of hypertension can increase stroke risk.⁶⁶⁹ The RATIO study (Risk of Arterial Thrombosis in Relation to Oral Contraceptives) demonstrated a higher risk of stroke in women with cardiovascular risk factors than women without in combinations containing 30 µg of estrogen plus second or third generation progestins.⁶⁷⁰ However, the different generations of progestin stroke risk did not differ in women without stroke risk factors.⁶⁷²

There is a lack of new cohort data to reflect the changes rates of stroke with CHC in our current population.

There are gaps in direct comparison of stroke risks related to CHC versus pregnancy.

Approximately 5% of women experience a natural menopause before 45 years of age, and many others have bilateral oophorectomy before 45 years of age. The preponderance of data support that early menopause (before 45 years of age) is a risk factor for stroke specifically. Multiple prospective studies have consistently shown an increased risk of stroke among those with premature or early menopause,^674,675,677 with other analyses showing similar trends.^685,692 In a meta-analysis of individual data from 15 prospective studies, those with menopause before 40 years of age had a 32% increased risk of stroke (95% CI, 1.43–2.07) and those with menopause between 40 and 44 years of age had an HR of 1.09 (95% CI, 1.18–1.43).⁶⁷⁶ Findings for stroke mortality have been less consistent.^692,693 The type of menopause, natural or surgical, does not appear to modify the association with stroke.

During the decline of estradiol levels in the menopausal transition, LDL levels generally rise and high-density lipoprotein levels decline.⁶⁹⁴ In many women, menopause contributes to a rise in BP, warranting monitoring during the menopausal transition.⁶⁹⁵ Premature menopause and early menopause have been associated with a substantially increased risk of stroke and advance the time of onset of changes in lipids and BP.^{674,675,677,679} Data are lacking on whether hormone replacement therapy, at least until the average age of menopause, might modify this risk. The ACC/AHA 2018 guideline on cholesterol management included premature menopause as a risk-enhancing factor to be considered in cholesterol management decisions.⁷⁵ Primary prevention of CVD and screening for and managing risk factors are warranted for those with a history of premature and early menopause.

The excess risk of stroke with the use of estrogen-containing HT is well established, but the majority of the RCT data come from the Women’s Health Initiative, in which the mean age was 67 years. The USPSTF in 2022 estimated that use of estrogen-only formulations results in 79 (95% CI, 15–159) more strokes per 10 000 women treated, and estrogen/progestin formulation results in 52 (95% CI, 12–104) more strokes per 10 000 women treated.⁶⁹⁶ A synthesis of systematic reviews of 17 RCTs reported a cumulative RR of 1.17 (95% CI, 1.05–1.29) and an RR of 1.35 (95% CI, 1.08–1.69) for nonfatal stroke with HT use.⁶⁸⁰ Another meta-analysis showed a 32% risk during HT (HR, 1.32 [95% CI, 1.12–1.56]) but among those with past use (HR, 1.00 [95% CI, 0.85–1.28]).⁶⁸² Observational data from the UK Biobank showed a small but significant increase in the risk of ischemic stroke with HT and a 33% increased risk of subarachnoid hemorrhage (HR, 1.33 [95% CI, 0.14–1.71]).⁶⁵⁸ According to the Menopause Society, settings where estrogen-based HT may be most appropriate include the use of the lowest effective dose of estrogen in women <60 years of age with low cardiovascular, thromboembolic, and breast cancer risk profiles who do not have unexplained vaginal bleeding or liver disease⁶⁹⁷ (Figure 6). Transdermal formulations of estrogen (especially low dose) were not associated with a clear risk of stroke.^691,698,699

The presence of vasomotor symptoms but, more important, features such as severity and frequency and the association with stroke risk require more focused studies. There may ultimately be a clearer association identified with continued follow-up in cohorts such as SWAN (Study of Women’s Health Across the Nation) and other pooled analyses/meta-analyses.

Additional research on the impact of age at menarche and reproductive life span on cardiovascular risk is needed.

Whether those with premature menopause benefit from hormone replacement therapy or early cardiovascular risk factor management to reduce long-term cardiovascular risk is unknown.

Lower doses of estrogen and transdermal routes have been studied for their impact on subclinical CVD but not in RCTs, and stroke incidence in healthy users has yet to be studied.

A cohort study of 2842 transfeminine people and 2118 transmasculine people receiving care in the Kaiser Permanente health care systems found that transfeminine people using gender-affirming HT had a higher incidence of ischemic stroke compared with cisgender women (adjusted HR, 1.9 [95% CI, 1.3–2.6]) but not compared with cisgender men (adjusted HR, 1.2 [95% CI, 0.9–1.7]).⁷⁰⁰ Stroke incidence in the transmasculine cohort was similar to that in cisgender populations, although the evidence was insufficient to allow conclusions about risk among transmasculine participants.⁷⁰⁰ These findings were consistent with a study from a single center in the Netherlands that found a higher age-adjusted standardized incidence ratio for stroke in transgender women (median age, 30 years) using gender-affirming HT compared with cisgender women (standardized incidence ratio, 2.42 [95% CI, 1.65–3.42]) and cisgender men (standardized incidence ratio, 1.80 [95% CI, 1.23–2.56]), whereas no difference was found for transgender men (median age, 23 years).⁷⁰¹ Data from the 2014 to 2017 Behavioral Risk Factor Surveillance System⁷⁰² and a 2019 systematic review⁷⁰³ similarly suggest a higher prevalence of stroke in transgender women. The observational design of the published studies indicating a potentially increased risk of stroke in transfeminine individuals taking gender-affirming HT should be interpreted with caution because these studies lack important details on the hormone regimens, hormone levels, lifestyle factors (eg, tobacco use), and external minority stress risk factors in the studied populations. In addition, major limitations of the studies are the young age of this population and the limited follow-up time. It is reasonable to assume that interventions on known vascular risk factors such as tobacco use and hypertension would be effective at reducing stroke risk in this population; however, specific intervention studies inclusive of transgender people are lacking.

Systematically collecting inclusive, participant-centered gender identity separate from sex assigned at birth in population health studies, cohort studies, and electronic health records;

Assessing the cerebrovascular health of transgender and gender-diverse people who are ≥50 years of age or not using gender-affirming HT and of those who identify outside of binary gender (nonbinary people, agender people);

Identifying the mechanisms driving disparities in stroke incidence, including but not limited to specific hormone regimens, hormone duration, levels and route of administration, assessment of gender minority stress and resilience, and SDOH; and

Assessing the efficacy of interventions to reduce stroke risk in transgender and gender-diverse people.

Two systematic reviews of randomized clinical trials and observational studies published before 2017 found no evidence for an increased risk of stroke in men using exogenous testosterone therapy. However, they noted that the low level of evidence limited definitive conclusions.^705,706 The 2023 TRAVERSE study (Testosterone Replacement Therapy for Assessment of Long-Term Vascular Events and Efficacy Response in Hypogonadal Men) sought to provide higher-quality evidence through a multicenter, randomized, double-blind, placebo-controlled noninferiority trial.⁷⁰⁴ The study enrolled 5246 men 45 to 80 years of age with confirmed hypogonadism (defined as reporting associated symptoms and having 2 fasting testosterone levels <300 ng/dL) to receive daily transdermal 1.62% testosterone gel with dose adjustments to target levels between 350 and 750 ng/dL or placebo gel for a mean of 21.7±14.1 months. The participants either had preexisting CVD (defined as coronary artery disease, cerebrovascular disease, or peripheral arterial disease) or had high risk of CVD (defined as ≥3 risk factors, including hypertension, dyslipidemia, current smoking, stage 3 chronic kidney disease, diabetes, elevated C-reactive protein level, age ≥65 years, or Agatston coronary calcium score >75th percentile for age and race). Investigators found no significant difference in the incidence of the primary composite end point of major adverse cardiac events or in the component end point of nonfatal stroke.

The 2023 TRAVERSE study provides reassurance that testosterone gel does not increase the risk of stroke in men with risks for stroke and who have confirmed hypogonadism. The risks of stroke for other populations of men who use testosterone off-label are not known.

The effects of other testosterone formulations and routes of administration on the risk of stroke are not well studied.

Future research may include investigation into other populations who may use testosterone, testosterone preparations and routes of administration, and longer durations of testosterone use.

Based on the WARCEF trial (Warfarin Versus Aspirin in Reduced Cardiac Ejection Fraction)⁷¹¹ in which there was no difference in the primary composite outcome of ischemic stroke, ICH, or death but a significant reduction in ischemic stroke, the 2014 guideline on primary stroke prevention¹³ recommended anticoagulant or antiplatelet agents as reasonable for patients with heart failure and no AF. Since publication, 2 subanalyses from WARCEF^708,711 and an RCT of rivaroxaban versus placebo⁷²⁵ also have demonstrated no difference in the primary composite outcome, modestly reduced stroke events, and increased risk of bleeding. In COMMANDER HF (A Study to Assess the Effectiveness and Safety of Rivaroxaban in Reducing the Risk of Death, Myocardial Infarction or Stroke in Participants With Heart Failure and Coronary Artery Disease Following an Episode of Decompensated Heart Failure),⁷²⁵ compared with placebo, there was no difference in major cardiovascular events (25.0% versus 26.2; HR, 0.94 [95% CI, 0.84–1.05]; P=0.27), rates of stroke were 2.0% versus 3.0% (HR, 0.66 [95% CI, 0.47–0.95]), and major bleeding occurred in 3.3% versus 2.0% (HR, 1.68 [95% CI,1.18–2.39]), respectively. Taken together, there is no clear net benefit and possible harm in prescribing anticoagulation for patients with cardiomyopathy and reduced left ventricular ejection fraction who have no other indication for anticoagulation.^710,726

Atrial cardiopathy without clinical AF is a risk factor for stroke, yet gaps remain in how to reduce this risk. Studies of anticoagulation in patients with atrial high-rate events, a precursor to stroke, and in patients with cryptogenic stroke, which may be secondary to subclinical AF, have not been effective in preventing stroke. Although these studies did not enroll patients on the basis of echocardiographic or echocardiographic criteria of atrial cardiopathy, they suggest that subclinical AF and other tachyarrhythmias with subsequent thromboembolism may not be the only mechanism responsible for increased stroke risk. More studies are needed to understand the cause and management of stroke risk in patients with atrial cardiopathy without AF.

Among patients with cardiomyopathy and reduced left ventricular ejection fraction, the 2 major clinical trials show a signal for reduced ischemic strokes with anticoagulation; however, the primary outcomes of these studies are negative, stroke events are low, and there is a significantly higher bleeding risk. Therefore, the net benefit for any one patient remains unknown. Future trials of anticoagulation in patients with cardiomyopathy and reduced left ventricular systolic function should select people at highest risk for stroke and consider more individualized estimation of risk/benefit thresholds.

There are no data assessing the use of antiplatelet therapy (versus placebo) for primary stroke prevention in patients with cardiomyopathy, which may be an area for future research.

In patients with a single risk factor or multiple risk factors, the use of aspirin to prevent major vascular events, including stroke, has been tested. In the ARRIVE trial (Use of Aspirin to Reduce Risk of Initial Vascular Events in Patients at Moderate Risk of Cardiovascular Disease), 12 546 patients judged to be at moderate risk for vascular events were randomized to aspirin 100 mg/d versus placebo.⁷²⁷ The primary composite end point did not differ between the 2 groups, and there was no difference in the rate of fatal or nonfatal stroke (aspirin, 1.20%; placebo, 1.07%). In addition, this trial included an intermediate-risk sample with risk factors treated more proportionately than in prior aspirin trials (75% on antihypertensive medications and 43% on statins). The JPPP trial (Japanese Primary Prevention Project) enrolled participants between 60 and 85 years of age with a history of hypertension, diabetes, or hyperlipidemia.⁷²⁸ The primary end point was stroke, MI, and vascular death. The study was stopped for futility after 60 months of follow-up. The primary end point was 2.77% with aspirin and 2.96% with placebo.

For patients with established coronary artery disease, the use of ticagrelor in addition to background therapy with aspirin can modestly reduce stroke. In the PEGASUS-TIMI 54 trial (Prevention of Cardiovascular Events in Patients With Prior Heart Attack Using Ticagrelor Compared to Placebo on a Background of Aspirin–Thrombolysis in Myocardial Infarction 54), patients were enrolled with MI 1 to 3 years previously along with at least 1 additional vascular risk factor. With 33 months of follow-up and stroke as an exploratory outcome, the addition of ticagrelor 60 mg was associated with a reduction in any stroke (HR, 0.75 [95% CI, 0.57–0.98]; P=0.03), with a trend toward reduction in ischemic stroke (HR, 0.76 [95% CI, 0.56–1.02]; P=0.06).⁷³² The annualized rate of stroke was 0.65 in the placebo group and 0.49 in the ticagrelor group, although this follow-up may be shorter than in many primary prevention trials. The rates of major bleeding were also higher in the 90-mg and 60-mg ticagrelor groups (2.3% and 2.6%, respectively) compared with placebo (1.06%).

Advanced age is a vascular risk factor. The ASPREE trial (Aspirin in Reducing Events in the Elderly) enrolled patients ≥70 years of age to compare aspirin 100 mg/day with placebo.³⁶⁰ The primary end point was fatal coronary heart disease, nonfatal MI, fatal or nonfatal stroke, or hospitalization for heart failure. With a median age of 74 years and a median of 4.7 years of follow-up, there was no reduction in the primary end point or stroke with aspirin therapy. A prespecified secondary analysis of ASPREE found a small increase in intracranial bleeding with aspirin use (0.7% absolute increase).⁷³⁵ In a subanalysis of JPPP focused on stroke, there was no reduction in stroke associated with aspirin use among Japanese patients with a mean age of 71 years.⁷³³

For patients with chronic kidney disease, a meta-analysis compared major vascular event outcomes with aspirin or placebo.⁷³⁴ Among 4468 study participants, no clear benefit was seen with aspirin for major vascular events (HR, 0.92 [95% CI, 0.49–1.73]) or stroke (HR, 0.86 [95% CI, 0.48–1.56]).

With several large clinical trials demonstrating lack of benefit for aspirin in the primary prevention setting, the question arises whether a patient profile exists that could benefit from aspirin use. Additional studies could be focused on patient groups with a potentially favorable risk/benefit equation such as the following:

Further information is also needed on prophylactic use of aspirin in higher-risk communities such as individuals who are Black, Hispanic, Native American, or Asian.