A novel neuropsychological assessment to discriminate between ischemic and nonischemic dementia

In the present study we have elaborated a multivariate-multifactor diagnostic method for diagnosis of ischemic and nonischemic dementia, and validated it using different human populations. Our purpose was to improve dementia diagnosis by means of comprehensive neuropsychological assessment and the control of intervening variables.

Data were obtained from 114 healthy volunteers to 101 patients consecutively referred for suspected dementia. On the basis of neuromorphological criteria, the patient sample was subdivided into those with ischemic lesions (IL: N=12) and without ischemic lesions (non-IL: N=89). The patient groups and the healthy subjects (HS) were matched according to age, education, sex ratio and handedness. A comprehensive neuropsychological battery was administered to all the participants.

The two patient groups differed in their Hachinski ischemic score, but not in terms of demographic variables, the Blessed rating scale, Mini Mental State, Geriatric Depression Scale, or other measures. Discriminant analysis selected fifteen neuropsychological variables from the comprehensive battery, and these were found to provide accuracy of classification in 98% of HS and 87% of the patients considered as a whole. Considering the three groups, those variables provided accuracy of classification in 98% of HS, 82% of IL patients and 82% of non-IL patients. Subtests were internally consistent (Cronbach's alpha: 0.87) and a positive association between ischemic lesions and cognitive impairment was observed when the 15 dependent variables were added.

It is possible to select a combination of cognitive tests that discriminated HS from dementia patients and, within this category, patients with and without ischemic lesions. Analysis of the 15 variables provides an improved method for diagnosis of ischemic vs. nonischemic lesion dementia in humans.