To develop a short cognitive test for screening mild cognitive impairment (MCI) in Hong Kong Chinese older adults.
The Chinese Abbreviated MCI (CAMCI) test was developed with a multistage process. In phase 1, a short version of the cognitive test comprising a 1-min animal fluency test and a 10-min delayed word list recall was developed and tested in 578 volunteers (community-dwelling active elderly persons). In phase 2, the CAMCI test was validated in an independent and randomly recruited sample of 459 participants in a community survey. Additionally, the predictive significance of the CAMCI test was evaluated in a group of 196 subjects assessed in phase 1 for conversion to clinical dementia at 20 months' follow-up. The discriminating power of the CAMCI test in differentiating MCI from normal control (NC) and mildly demented subjects was compared with Mini Mental State Examination (MMSE) and Alzheimer's Disease Assessment Scale-Cognitive (ADAS-Cog) subscales.
The CAMCI test was found to have high discriminating power in differentiating NC from MCI and mildly demented subjects in the phase 1 volunteer sample. The receiver operating characteristics (ROC) revealed an area under the curve (AUC) of 0.91. The ROC were further validated in the phase 2 sample. The AUC of the CAMCI test was compared with MMSE and ADAS-Cog subscales. The short MCI test was comparable to the ADAS-Cog subscale in discriminating NC from MCI and demented subjects (chi(2) test, p = n.s.). Logistic regression analysis was carried out to determine significant baseline predictors for conversion to dementia at phase 3 follow-up. Both ADAS-Cog total [Exp(B) = 1.115, p = 0.028] and CAMCI [Exp(B) = 0.88, p = 0.045] scores were significant predictors for dementia status at follow-up.
The CAMCI test is able to discriminate NC from MCI and mild dementia in Hong Kong Chinese older adults. Its potential for large-scale community screening for early detection of cognitive impairment in late life should be emphasized and explored.