Mini-Mental State Examination (MMSE) and the Modified MMSE (3MS): A psychometric comparison and normative data.

study, 525 community-dwelling participants, aged 65-89, were divided into 2 groups: no cognitive impairment (NCI; n - 406) and Alzheimer's disease (n = 119). Both tests yielded comparable reliability estimates. Fewer years of education decreased specificity and increased sensitivity, whereas increasing age primarily decreased specificity. It is concluded that although the 2 tests produce comparable effects, the inclusion of a verbal fluency test would increase the sensitivity of the MMSE. Normative data for the NCI group, stratified for 2 age levels (65-79 and 80-89) and 2 educational levels (0-8 and 9+ years), are presented. Folstein, Folstein, and McHugh (1975) introduced the MiniMental State Examination (MMSE) as a brief, objective assessment of cognitive functioning and as a measure of changes in cognitive status. The MMSE usually can be administered in 510 min and has been employed extensively in clinical settings, community surveys, and epidemiological studies. In a recent review of the literature, Tombaugh and Mclntyre (1992) concluded that the MMSE possessed moderate to high reliability coefficients, demonstrated high levels of sensitivity for cognitive deficits in patients suffering from moderate to severe Alzheimer's disease, and reflected the cognitive decline typical of dementia patients. Criticisms of the MMSE included (a) its failure to discriminate between people with mild dementia and those who are not demented, (b) a limited ability to detect impairment caused by focal lesions, particularly those in the right hemisphere, (c) overly simple language items that reduce sensitivity to mild linguistic deficits, and (d) a large number of false-positive errors because of its bias against individuals with low education. In response to these problems, several attempts have been made to improve the MMSE. Of these, the Modified Mini-Mental State Examination (3MS; Teng & Chui, 1987) represents the most extensive revision. Teng and Chui (1987) added four additional subtests (date and place of birth, word fluency, similarities, and delayed recall of words). The maximum score was