OBJECTIVE: To examine the relation between performance on a dementia screening test and the demographic variables of age, education, gender, and urban vs rural residency. DESIGN: Community survey with cluster sampling. SETTING: One urban and one rural community from each of four geographic regions in Taiwan, Republic of China. PARTICIPANTS: A total of 5265 nondemented individuals approximately equally divided between men and women and between urban and rural residency with a range in age from 41 to 88 years and in education from 0 to 20 years. MAIN OUTCOME MEASURE: Score on a Chinese adaptation of the Mini-Mental State Examination. RESULTS: Lower test scores were associated with older age and less education. The decrease in score with age was faster among participants who had never attended school. Better performance by men and by urban residents was found only among participants with fewer than 6 years of schooling. In this group, the magnitudes of sex and residency differences were comparable among those subjects aged 41 to 64 years and those aged 65 to 88 years. Women who had never worked outside of the home performed poorer than those who had worked outside of the home. CONCLUSIONS: The influence of educational background on test performance is most evident in individuals with less education. Commonly used dementia screening tests may be unfair to poorly educated individuals, especially women and rural residents. Efforts should be made to develop ecologically relevant cognitive tests for the intended study populations. To help distinguish test bias from different rates of cognitive decline, the study populations should include individuals in predementia age ranges.
OBJECTIVE: To examine the relation between performance on a dementia screening test and the demographic variables of age, education, gender, and urban vs rural residency. DESIGN: Community survey with cluster sampling. SETTING: One urban and one rural community from each of four geographic regions in Taiwan, Republic of China. PARTICIPANTS: A total of 5265 nondemented individuals approximately equally divided between men and women and between urban and rural residency with a range in age from 41 to 88 years and in education from 0 to 20 years. MAIN OUTCOME MEASURE: Score on a Chinese adaptation of the Mini-Mental State Examination. RESULTS: Lower test scores were associated with older age and less education. The decrease in score with age was faster among participants who had never attended school. Better performance by men and by urban residents was found only among participants with fewer than 6 years of schooling. In this group, the magnitudes of sex and residency differences were comparable among those subjects aged 41 to 64 years and those aged 65 to 88 years. Women who had never worked outside of the home performed poorer than those who had worked outside of the home. CONCLUSIONS: The influence of educational background on test performance is most evident in individuals with less education. Commonly used dementia screening tests may be unfair to poorly educated individuals, especially women and rural residents. Efforts should be made to develop ecologically relevant cognitive tests for the intended study populations. To help distinguish test bias from different rates of cognitive decline, the study populations should include individuals in predementia age ranges.