Psychometric Limitations of the Mini-Mental State Examination Among Nondemented Older Adults: An Evaluation of Neurocognitive and Magnetic Resonance Imaging Correlates

Abstract

Background/Study Context: Although many of the Mini-Mental State Examination's (MMSE) limitations are well accepted among geriatricians, neuropsychologists, and other interested clinicians and researchers, its continued use in psychometrically unsound ways suggests that additional investigation and dissemination of information are sorely needed. The authors aimed to describe the reliability and validity of the MMSE as a measure of cognitive function among healthy older adults.

Methods: The authors examined MMSE performance in 124 stroke- and dementia-free, community-dwelling older adults (65% male; mean age = 66.5 years). All participants were administered an extensive neuropsychological battery composed of measures of attention, executive function, memory, and visuospatial function. A subset of 99 participants also underwent magnetic resonance imaging (MRI). MMSE test-retest reliability was examined among 65 participants who underwent repeat MMSE testing over an average interval of 83.2 days.

Results: Spearman test-retest correlation for total MMSE scores was r _S = .35 (p = .004), for Serial Sevens was r _S = .40 (p = .001), and for Word Recall was r _S = −.01 (p = .96). Total MMSE performance correlated significantly with a minority of neuropsychological tests and MRI-derived indices of white matter disease and brain atrophy. A subset of 17% of participants demonstrated inappropriate intrusion of MMSE Pentagon Copy during another test of visuospatial recall.

Conclusions: Overall, MMSE scores exhibited ceiling effects, poor test-retest reliability, limited sensitivity to subtle brain abnormalities, and a high rate of intrusion elsewhere in the neuropsychological battery. Individual MMSE items demonstrated poor construct validity. These qualities illustrate the serious limitations of the MMSE in detecting individual differences in cognitive function among healthy older adults.

Acknowledgments

This work was supported by National Institutes of Health (NIH) grants R29 AG15112, R01 AG15112, and K24 AG00930; a VA Merit Grant, Bristol Myers Squibb Medical Imaging, Inc., the Department of Veterans Affairs Baltimore Geriatric Research Education and Clinical Center; and the Geriatrics and Gerontology Education and Research Program of the University of Maryland, Baltimore. The authors also wish to acknowledge Keith V. Hughitt for his contributions to preliminary statistical analyses associated with a subsection of the manuscript.

Notes

*p < .05; †p < .01.

†Range possible = 0–3; observed range for periventricular and deep white matter hyperintensities = 0–3; observed range for ventricular enlargement and sulcal widening = 0–2; see Methods for detailed scoring criteria.

‡No maximum upper limit; observed range = 0–31.

*p < .05; **p < .01.