Cerebral lesions on magnetic resonance imaging, heart disease, and vascular risk factors in subjects without stroke. A population-based study.

BACKGROUND AND PURPOSE: To assess the prevalence of asymptomatic abnormalities on magnetic resonance imaging of the brain and their possible relation to hypertension, heart disease, and carotid artery disease, we studied 77 randomly selected subjects (mean age, 65.1 years; range, 36 to 95 years) with no history of focal brain lesions. METHODS: The study protocol included magnetic resonance imaging of the brain, transthoracic and transesophageal echocardiography, ultrasonography of the carotid arteries, and electrocardiographic recording. Deep and periventricular white matter hyperintensities on magnetic resonance imaging were assessed both separately and together. RESULTS: On magnetic resonance imaging of the brain 62.3% (95% confidence interval CI, 51.5% to 73.2%) of the subjects had white matter hyperintensities. These abnormalities increased significantly with age (chi 2 test; P = .0001), from 13.6% (95% CI, 0% to 28.0%) of subjects aged younger than 55 years to 85.2% (95% CI, 71.8% to 98.6%) of subjects aged 75 years or older. Six subjects had deep gray matter hyperintensities localized in the basal ganglia, and one had a cerebellar infarction. Stepwise logistic regression analysis identified age and a history of heart disease (but not echocardiographic findings) to be independently associated with deep and periventricular white matter hyperintensities. Hypertension was only independently associated with periventricular white matter hyperintensities. Of the 68 subjects examined with both transthoracic and transesophageal echocardiography, potential cardioembolic sources were detected in 38.2% (95% CI, 26.7% to 49.8%) of the subjects with transthoracic echocardiography and in 47.1% (95% CI, 35.2% to 58.9%) of those with transthoracic and transesophageal echocardiography combined. In subjects aged 75 years or older, a possible cardiac embolic source was detected in 64.0% on transthoracic echocardiography and in 72.0% on transthoracic and transesophageal echocardiography combined, compared with 5.3% and 15.8%, respectively, in subjects aged younger than 55 years. CONCLUSIONS: White matter hyperintensities and potential cardioembolic sources are frequently present in asymptomatic individuals, stressing the need for age-matched control subjects in studies of patients with stroke or dementia.