Abstract
Abstract
Objective: To evaluate the clinical utility of the Spanish versions of memory vs. non-memory instruments in a Spanish-speaking sample; specifically, the comparative performance for the English versions in amnestic MCI (aMCI) and cognitively unimpaired (CU) elders.
Method: 255 Non-Hispanic English speakers (NH-ES; 157 CU; 98 aMCI) and 201 Hispanic Spanish speakers (H-SS; 103 CU; 98 aMCI) were classified using the Clinical Dementia Rating Scale and a validated culturally fair test. ANCOVA procedures were used to test for differences between four groups across languages adjusting for age, sex, and education for memory (e.g., Hopkins Verbal Learning Test-HVLT-R; NACC Logical Memory) and non-memory measures (e.g., Trail-Making Test A-B, Category and Phonemic Fluency).
Results: For memory measures, CU older adults were statistically differentiated from aMCI counterparts regardless of language. No significant differences between H-SS aMCI and NH-ES aMCI groups’ on HVLT-R and NACC story passages immediate/delayed conditions. No significant differences found between H-SS CU and NH-ES CU on memory indices. Conversely, Bonferroni corrected means (p<.05) revealed that even after adjustment for demographic covariates, ES CU outperformed H-SS CU counterparts on Trails A-B, and Category/Phonemic Fluencies. NH-ES aMCI groups scored better than H-SS aMCI on Trails B.
Conclusions: Results indicate that memory measures experience less bias when employed among different linguistic/cultural groups regardless of cognitive status. In contrast, CU H-SS groups scored lower on all non-memory measures. Factors that lead to potential bias, particularly among CU are discussed. Trails-B evidenced bias for both Hispanic CU and MCI, making this a measure that requires the judicious use of appropriate norms.