Cut-off Score on MMSE-K for Screening of Dementia in Community Dwelling Old People

Kim, Jae Min; Shin, Il Seon; Yoon, Jin Sang; Kim, Jung Hun; Lee, Hyung Yung

Abstract

OBJECTIVES
This study aimed to estimate the optimal cut-off score on Korean version of Mini-Mental State Examination (MMSE-K) for screening of dementia in community dwelling old people, and to investigate possible changes of the cut-off score according to age, gender, educational level, living area, and depression of the old people.
METHODS
For the old people aged 65 or over dwelling in two areas of Kwangju, cognitive function (MMSE-K) and demographic characteristics (age, gender, educational level, living area) were investigated. For those scoring 24 or less on MMSE-K of them, clinical diagnosis of dementia (DSM-IV) and depression (Hamilton Rating Scale for Depression) were assessed. The optimal cut-off score on MMSE-K with a sensitivity and a specificity was examined by using receiver operating characteristics curve analysis served by diagnosis of dementia as a gold standard. Associations of the cut-off score on MMSE-K with demographic characteristics and depression were estimated by odds ratios. For those factors showing significant association with cut-off score on MMSE-K regardless of the diagnosis of dementia, the optimal cut-off scores on MMSE-K were measured again at each level.
RESULTS
In the old people scoring 24 or less on MMSE-K (n=341), the optimal cut-off score on MMSE-K served by diagnosis of dementia was 21/22 with a sensitivity of 81% and a specificity of 64%. Only the age among the investigated characteristics was significantly associated with the cut-off score on MMSE-K, and the optimal cut-off score in those aged 75 or over was revealed to be 20/21.
CONCLUSIONS
We suggest that the cut-off score on MMSE-K for screening of dementia investigated in this study could be used by community health care staff as a guideline for assessing old people with questionable cognitive dysfunction or for considering to consult them to an expert.