Incidence of Silent Stroke in the United States

Abstract

P134 Background: Recent estimates of stroke incidence in the US range from 715,000–750,000 annually. These estimates, however, do not reflect silent infarcts and hemorrhages. Since population-based studies have found that prevalence of silent stroke is 10–20 times that of symptomatic, estimates of stroke incidence based solely on symptomatic events may substantially underestimate the annual burden of stroke. Silent strokes contribute to vascular dementia, gait impairment, and other major adverse patient outcomes. Methods: Incidence of silent infarcts for different age strata were derived from two US population-based studies of the prevalence of silent infarct-like lesions on MRI, Atherosclerosis Risk In Communities and Cardiovascular Health Study. Prevalence observations in these studies and age-specific death rates from the US Census Bureau were inputted to calculate silent infarct incidence (method of Leske et al). Similarly, incidence rates of silent hemorrhage at differing ages were extrapolated from population-based prevalence observations employing MR GRE imaging in the Austrian Stroke Prevention Study. Age-specific incidence rates were projected onto age cohorts in the 1998 US population to calculate annual burden of silent stroke. Results: Derived incidence rates per 100,000 of silent infarct ranged from 6400 in the age 50–59 strata to 16400 at ages 75–79. Extrapolated incidence rates of silent hemorrhage ranged from 230 in the age 30–39 strata to 7360 at ages > 80. Incidence rates of both subclinical infarcts and hemorrhage increased exponentially with age. Overall estimated annual US occurrence of silent infarct was 9,039,000, and of silent hemorrhage 2,130,000. Conclusion: In 1998, nearly 12 million strokes occurred in the United States, of which ∼750,000 were symptomatic and over 11 million were subclinical. Among the silent strokes, ∼81% were infarcts and ∼19% hemorrhages. These findings demonstrate that the annual burden of stroke is substantially higher than suggested by estimates based solely on clinically manifest events, and suggest that greater research and clinical resources should be allocated to stroke prevention and treatment.