Ming-yue Gao1, Min Yang2, Wei-hong Kuang3, Pei-yuan Qiu2. 1. Department of Epidemiology and Biostatistics, West China School of Public Health, Sichuan University,Chengdu 610041,China. 2. Department of Epidemiology and Biostatistics, West China School of Public Health, Sichuan University,Chengdu 610041,China;West China Research Center for Rural Health Development, Chengdu 610041,China. 3. Mental Health Center,West China Hospital of Sichuan University,Chengdu 610041,China.
Abstract
OBJECTIVE: To examine factors that may have impact on the Mini-Mental State Examination (MMSE) screening validity, which could lead to further establishing the general model of the MMSE score in Chinese health elderly and to improve the screening accuracy of the existing MMSE reference. METHODS: Based on the data of the Chinese Longitudinal Healthy Longevity Survey (CLHLS), the MMSE scores of 19,117 normal elderly and 137 dementia patients who met the inclusion criteria were used for the analysis. The area under the curve (AUC) and validity indexes were used to compare the screening accuracy of various criteria. Multiple linear regression was used to identify factors that had impact on the MMSE score for both the normal and dementia elderly. Descriptive analysis was performed for differences in the MMSE scores by age trends and gender between the normal and dementia elderly. RESULTS: The AUC of MMSE was ≥0.75(P<0.05). The MMSE score of the normal elderly declined nonlinearly as the age grew older(male: R2=0.924, P<0.05; female: R2=0.951, P<0.05), and increased nonlinearly as the education level rose(male: R2=0.948, P<0.05; female: R2=0.859, P< 0.05). The females had significantly lower MMSE scores than the males, with a faster decline trend with age than the males (95%CI of female partial regression coefficient was not overlapped with 95%CI of male partial regression coefficient). The dementia elderly showed a much lower MMSE score (male: difference of Z score:-1.573, P<0.05; female: difference of Z score:-1.222, P<0.05) and tended to with a faster decline speed than that of the normal elderly (95%CI of dementia partial regression coefficient included 95%CI of normal partial regression coefficient). CONCLUSION: The screening validity of MMSE in CLHLS is not affected by educational level. The analysis of factors that may impact on the MMSE screening validity are gender, age, vision and residence which with validity identification. These four factors can be used as assist tool of MMSE in the screening of dementia to improve the screening accuracy.
OBJECTIVE: To examine factors that may have impact on the Mini-Mental State Examination (MMSE) screening validity, which could lead to further establishing the general model of the MMSE score in Chinese health elderly and to improve the screening accuracy of the existing MMSE reference. METHODS: Based on the data of the Chinese Longitudinal Healthy Longevity Survey (CLHLS), the MMSE scores of 19,117 normal elderly and 137 dementiapatients who met the inclusion criteria were used for the analysis. The area under the curve (AUC) and validity indexes were used to compare the screening accuracy of various criteria. Multiple linear regression was used to identify factors that had impact on the MMSE score for both the normal and dementia elderly. Descriptive analysis was performed for differences in the MMSE scores by age trends and gender between the normal and dementia elderly. RESULTS: The AUC of MMSE was ≥0.75(P<0.05). The MMSE score of the normal elderly declined nonlinearly as the age grew older(male: R2=0.924, P<0.05; female: R2=0.951, P<0.05), and increased nonlinearly as the education level rose(male: R2=0.948, P<0.05; female: R2=0.859, P< 0.05). The females had significantly lower MMSE scores than the males, with a faster decline trend with age than the males (95%CI of female partial regression coefficient was not overlapped with 95%CI of male partial regression coefficient). The dementia elderly showed a much lower MMSE score (male: difference of Z score:-1.573, P<0.05; female: difference of Z score:-1.222, P<0.05) and tended to with a faster decline speed than that of the normal elderly (95%CI of dementia partial regression coefficient included 95%CI of normal partial regression coefficient). CONCLUSION: The screening validity of MMSE in CLHLS is not affected by educational level. The analysis of factors that may impact on the MMSE screening validity are gender, age, vision and residence which with validity identification. These four factors can be used as assist tool of MMSE in the screening of dementia to improve the screening accuracy.