Characteristics of the Modified Mini-Mental State Exam among elderly persons.

This study assesses psychometric properties of the Modified Mini-Mental State Exam (3MS) and present population norms and demographic risk factors for low 3MS scores. The subjects were 885 persons aged 65 and older who took the 3MS as part of the Stirling County Study, a population-based longitudinal study of adult residents of a county in Atlantic Canada. 3MS scores were not dependent on the specific rater who scored the test; thus, the 3MS is free of rater bias. Interrater reliability was high (intraclass correlation coefficient=0.98), as was internal consistency (coefficient alpha=0.91). Test-retest reliability over 3 years was 0.78. One third of subjects tested as cognitively impaired. Risk factors for low scores include older age, less education, male gender, and examination in French. The correlation between 3MS and Mini Mental State Exam scores was 0.95. The 3MS can be used as an epidemiologic measure of global cognitive performance among elderly persons.