Analysis of brief language tests in the detection of cognitive decline and dementia

Abstract

Abstract Lexical access difficulties are frequent in normal aging and initial stages of dementia. Verbal fluency tests are valuable to detect cognitive decline, evidencing lexico-semantic and executive dysfunction. Objectives: To establish which language tests can contribute in detecting dementia and to verify schooling influence on subject performance. Method: 74 subjects: 33 controls, 17 Clinical Dementia Rating (CDR) 0.5 and 24 (Brief Cognitive Battery - BCB e Boston Naming Test - BNT) 1 were compared in tests of semantic verbal fluency (animal and fruit), picture naming (BCB and BNT) and the language items of Mini Mental State Examination (MMSE). Results: There were significant differences between the control group and both CDR 0.5 and CDR 1 in all tests. Cut-off scores were: 11 and 10 for animal fluency, 8 for fruit fluency (in both), 8 and 9 for BCB naming. The CDR 0.5 group performed better than the CDR 1 group only in animal fluency. Stepwise multiple regression revealed fruit fluency, animal fluency and BCB naming as the best discriminators between patients and controls (specificity: 93.8%; sensitivity: 91.3%). In controls, comparison between illiterates and literates evidenced schooling influence in all tests, except for fruit fluency and BCB naming. In patients with dementia, only fruit fluency was uninfluenced by schooling. Conclusion: The combination of verbal fluency tests in two semantic categories along with a simple picture naming test is highly sensitive in detecting cognitive decline. Comparison between literate and illiterate subjects shows a lesser degree of influence of schooling on the selected tests, thus improving discrimination between low performance and incipient cognitive decline.