Cognitive predictors for five-year conversion to dementia in community-dwelling Chinese older adults

Candy H. Y. Wong*, Grace T. Y. Leung, Ada W. T. Fung, W. C. Chan, Linda C. W. Lam

*Corresponding author for this work

    Research output: Contribution to journalJournal articlepeer-review

    24 Citations (Scopus)

    Abstract

    Background: This study evaluates which cognitive measure is best for predicting incident dementia in a population-based random sample of Chinese older adults without dementia over a five-year period.

    Methods: A total of 787 community-dwelling Chinese older adults without dementia in Hong Kong were assessed at baseline (T0), at two years (T1), and at five years (T2).

    Results: The annual conversion rate to dementia was 1.6% and 6.3% for baseline normal participants and baseline mild cognitive impairment (MCI) participants, respectively. The Cantonese version of Mini-mental State Examination (CMMSE) scores declined significantly over time. In participants who progressed to dementia, Category Verbal Fluency Test (CVFT) scores dropped significantly from T0 to T1. A 1-SD drop of either CMMSE or CVFT in two years predicted dementia at five years with 91.5% sensitivity and 62.0% specificity. A stable CMMSE and CVFT at two years predicted a 91% chance of not progressing to clinical dementia at five years.

    Conclusion: In this community sample of Chinese older adults, a decline in cognitive screening tests in short term (two years) offered useful information in predicting dementia conversion over a longer period.
    Original languageEnglish
    Pages (from-to)1125-1134
    Number of pages10
    JournalInternational Psychogeriatrics
    Volume25
    Issue number7
    DOIs
    Publication statusPublished - Jul 2013

    User-Defined Keywords

    • cognitive predictors
    • dementia
    • mild cognitive impairment

    Fingerprint

    Dive into the research topics of 'Cognitive predictors for five-year conversion to dementia in community-dwelling Chinese older adults'. Together they form a unique fingerprint.

    Cite this