Correlates of leukoaraiosis and ventricular enlargement on magnetic resonance imaging: A study in normal elderly and cerebrovascular patients

Several studies have repeatedly demonstrated that leukoaraiosis as well as ventricular enlargement are common findings in normal elderly and in stroke patients, although there is no general consensus on prevalence rate as well as on their clinical correlations. It is also controversial whether white matter changes and ventricular enlargement are reciprocally related. In this study we investigated the prevalence and extent of white matter hyperintensities and the degree of ventricular enlargement on magnetic resonance imaging in 50 normal elderly individuals (mean age 62.1 +/- 7.3 years) and in 50 consecutive chronic ischemic stroke patients (mean age 66.1 +/- 7.7 years). All subjects underwent extensive clinical assessment. White matter hyperintensities were graded from 0 to 3 on a semi-quantitative scale, while bifrontal horn, bicaudate, and third ventricle ratio indices were used as measures of brain atrophy. Hypertension, diabetes, alcohol consumption, cardiac disease, carotid pathology occurred significantly more often in patients than in controls. Prevalence rates of white matter hyperintensities were 30% in controls and 82% in patients. Patients had significantly larger ventricular indices than controls. Significant univariate correlations for the extent of white matter hyperintensities were found with age, sex, hypertension, cardiac disease, carotid pathology, diabetes, history of stroke and ventricular enlargement. Age, sex, cardiac disease, alcohol habit, cerebrovascular disease and extent of white matter hyperintensities correlated with severity of ventricular enlargement. Multivariate regression analysis identified age, hypertension and history of stroke as independent predictors of white matter hyperintensities, while history of stroke, age and alcohol consumption were found as the only independent predictors of ventricular enlargement. Separate analysis between periventricular, subcortical or deep white matter hyperintensities and each of the three ventricular indices failed to show a significant association after adjustment for clinical and demographic factors. We suggest that leukoaraiosis and ventricular enlargement are independent pathological processes associated with different risk factors in addition to age and stroke disease.