[Skip to Content]
[Skip to Content Landing]
Article
September 1995

Clinical Validity of the Mattis Dementia Rating Scale in Detecting Dementia of the Alzheimer Type: A Double Cross-Validation and Application to a Community-Dwelling Sample

Arch Neurol. 1995;52(9):899-904. doi:10.1001/archneur.1995.00540330081018
Abstract

Objective:  To assess the clinical validity of the Dementia Rating Scale (DRS) in detecting patients with dementia of the Alzheimer type (DAT).

Background:  The DRS is widely used to evaluate cognitive functioning in older adults. Adequate normative data are unavailable; studies addressing the clinical validity of the DRS are limited by small sample sizes.

Design and Methods:  Administered the DRS to 254 outpatients with DAT and 105 healthy elderly subjects. Performed (1) multiple regressions of demographic factors on the DRS and its subscales; (2) derivation of optimal DRS cutoff scores using receiver operating characteristic curves; (3) double cross-validation with stepwise logistic regressions; and (4) application of results to a community-dwelling sample.

Results:  Age- and education-adjusted DRS scores were computed. The optimal DRS cutoff score for DAT of 129 or less revealed a sensitivity of 98% and a specificity of 97%. The logistic regressions resulted in a combination of the Memory and Initiation/Perseveration subscales that correctly classified 98% of all subjects, 92% of a subsample of 76 patients with mild DAT, and 100% of the 51 patients with autopsyconfirmed DAT. The resultant equation was then applied to a community-dwelling sample (238 healthy elderly subjects and 44 patients with DAT): 91% of patients and 93% of normal subjects were correctly classified. Of an additional 77 individuals with questionable DAT, 43 were classified as demented and 34 were classified as nondemented.

Conclusions:  The DRS is a clinically valid psychometric test for the detection of DAT. The Memory and Initiation/Perseveration subscales are its best discriminative indexes for an abbreviated version.

×