Difference in MRI findings and risk factors between multiple infarction without dementia and multi-infarct dementia

Abstract

MRI findings and risk factors for vascular dementia were evaluated with multi-variate analysis in 96 multi-infarct patients without dementia and 40 multi-infarct patients with dementia (MID). Only subjects with small infarcts in the territory of the perforator artery or deep white matter were studied. The diagnosis of MID was diagnosed according to DMS-III criteria and Hachinski's ischemia score. Location and area of patchy high-intensity areas including small infarcts, the degree of periventricular high intensity (PVH), and the degree of brain atrophy were examined with MR images. Independent variables were: history of hypertension, diabetes mellitus, other complications; systolic and diastolic blood pressure, atherosclerotic index, hematocrit, history of smoking, level of education, and activities of daily life (ADL). Hayashi's quantification method II was used to analyze the data. The most significant correlation was found between history of hypertension and dementia (partial correlation coefficient: 0.39). Significant correlations were also found between ADL and dementia (0.32), between thalamic infarction and dementia (0.31), and between PVH and dementia (0.27). Age, brain atrophy index, and history of diabetes mellitus contributed little to dementia. The contribution to dementia did not differ significantly between right and left patchy high-intensity areas on MR images. Location of infarcts, except for bilateral thalamic infarcts and large PVH, contributed little to dementia. Thus it would be difficult to base a prediction of the prevalence of vascular dementia on MRI findings. However, both hypertension and ADL contribute to vascular dementia and both are treatable, which may be significant for the prevention of dementia.