TY - JOUR

T1 - Mini-Mental Status Examination: mixed Rasch model item analysis derived two different cognitive dimensions of the MMSE

AU - Schultz-Larsen, Kirsten

AU - Kreiner, Svend

AU - Lomholt, Rikke Kirstine

N1 - Keywords: Age Distribution; Aged; Aged, 80 and over; Attention; Bias (Epidemiology); Cognition; Cognition Disorders; Dementia; Denmark; Female; Humans; Male; Mass Screening; Mental Recall; Mental Status Schedule; Models, Psychological; Neuropsychological Tests; Orientation; Psychometrics; Sex Distribution

PY - 2006

Y1 - 2006

N2 - OBJECTIVES: This study published in two companion papers assesses properties of the Mini-Mental State Examination (MMSE) with the purpose of improving the efficiencies of the methods of screening for cognitive impairment and dementia. STUDY DESIGN AND SETTING: An item analysis by conventional and mixed Rasch models was used to explore empirically derived cognitive dimensions of the MMSE, to assess item bias, and to construct diagnostic cut-points. The scores of 1,189 elderly residents were analyzed. RESULTS: Two dimensions of cognitive function, which are statistically and conceptually different from those obtained in previous studies, were derived. The corresponding sum scales were (1) age-correlated MMSE scale (A-MMSE scale: orientation to time, attention/calculation, naming, repetition, and three-stage command) and (2) non-age-correlated MMSE scale (B-MMSE scale: orientation to place, registration, recall, reading, and copying). The "writing" item was not included due to differential effects of age and sex. The analysis also showed that the study sample consisted of two cognitively different groups of elderly. CONCLUSIONS: The findings indicate that a two-scale solution is a stable and statistically supported framework for interpreting data obtained by means of the MMSE. Supplementary analyses are presented in the companion paper to explore the performance of this item response theory calibration as a screening test for dementia.

AB - OBJECTIVES: This study published in two companion papers assesses properties of the Mini-Mental State Examination (MMSE) with the purpose of improving the efficiencies of the methods of screening for cognitive impairment and dementia. STUDY DESIGN AND SETTING: An item analysis by conventional and mixed Rasch models was used to explore empirically derived cognitive dimensions of the MMSE, to assess item bias, and to construct diagnostic cut-points. The scores of 1,189 elderly residents were analyzed. RESULTS: Two dimensions of cognitive function, which are statistically and conceptually different from those obtained in previous studies, were derived. The corresponding sum scales were (1) age-correlated MMSE scale (A-MMSE scale: orientation to time, attention/calculation, naming, repetition, and three-stage command) and (2) non-age-correlated MMSE scale (B-MMSE scale: orientation to place, registration, recall, reading, and copying). The "writing" item was not included due to differential effects of age and sex. The analysis also showed that the study sample consisted of two cognitively different groups of elderly. CONCLUSIONS: The findings indicate that a two-scale solution is a stable and statistically supported framework for interpreting data obtained by means of the MMSE. Supplementary analyses are presented in the companion paper to explore the performance of this item response theory calibration as a screening test for dementia.

U2 - 10.1016/j.jclinepi.2006.06.007

DO - 10.1016/j.jclinepi.2006.06.007

M3 - Journal article

C2 - 17292021

VL - 60

SP - 268

EP - 279

JO - Journal of Clinical Epidemiology

JF - Journal of Clinical Epidemiology

SN - 0895-4356

IS - 3

ER -