Neuropsychological and neuropsychiatric prediction of cognitive functioning and cognitive decline: A cross-cutural perspective

The goal of this study was to examine whether seven neuropsychological tests and three depression measures were associated with cognitive functioning among Hispanic elderly, and to assess whether they operated differently for Hispanic and White ethnic groups. Participants were 89 community-dwelling elderly Hispanics and 89 Whites matched on clinical and demographic characteristics, all followed longitudinally at the NYU Alzheimer’s Disease Center (NYUADC). Their cognitive functioning ranged from normal to moderate dementia. Although a large proportion of the measures evidenced an association with cognitive status, hierarchical regression analyses showed that the initial recall of the NYU-Paragraph Test and the Retardation measure (especially for Spanish-speaking Hispanics), as well as WAIS-Digit Symbol (especially for English-speaking Hispanics) were most strongly associated with cognitive status after controlling for demographic and other cognitive measures. Retardation (for the entire Hispanic group) and the total score of the Hamilton Depression Rating Scale (HDRS; for the entire White group) differentiated between normal cognition and Mild Cognitive Impairment (MCI). For the full range of cognitive status, the combined predictive usefulness of the 10 measures differed significantly between Hispanics and Whites; this difference was especially driven by the Retardation measure. For nondemented participants, only the depression measures predicted differently between the two ethnic groups. When the analyses examined the primary language of the participants, the predictive usefulness of neuropsychological and depression measures differed between Spanish-speaking Hispanics and their matched Whites, but no differences were found between English-speaking Hispanics and their White counterparts. Subsequent analyses showed that, for Hispanics, the usefulness of the 21-item HDRS-total score reflected the contributions of Work and Activities and Retardation, but not the other 19 items. The WAIS-Digit Symbol lacked specificity for Spanish-speaking Hispanics. Preliminary longitudinal analysis showed that the delayed recall of the NYU-Paragraph Test predicted cognitive decline among Spanish-speaking Hispanics. These results suggest that the NYU-Paragraph Test and Retardation may improve diagnostic accuracy and prediction of decline among Spanish-speaking Hispanics.