Mini-Mental State Examination (MMSE)
The Mini-Mental State Examination (MMSE) was developed by Folstein et al. (1975) as an examination of mental state for hospitalized psychiatric patients, though it is now commonly used to examine cognition in older adults. Folstein et al. (1975) noted that while there were many cognitive performance assessments available at that time, the assessments that were available were lengthy. This posed a significant problem, as individuals with dementia symptoms or delirium are only able to maintain focus for short periods of time (Halstead 1943). Alternatively, the MMSE is a structured assessment of mental status/global cognitive functioning that is very brief, taking approximately 5–10 min to administer. The MMSE measures cognitive functioning in the domains of orientation, learning and memory, attention and calculation, language, and visuospatial functioning.
The MMSE was developed because the authors identified the need for a global cognitive assessment that could be quickly and serially implemented in medical settings with patients who have varied levels of cognitive abilities. Folstein et al. (1975) described the assessment as “mini” because it focuses on the cognitive aspects of mental functioning and excludes evaluations of mood or abnormal thinking. It is separated into two sections: the first being questions that solely require verbal responses in the domains of orientation, memory, and attention and the second being questions that require the ability to follow commands, write a sentence spontaneously, and copy a semi-complex figure (Folstein et al. 1975). The maximum score is 30. The clinical cutoff score for an indication of dementia is considered a 24/30 (Lezak et al. 2012), though this has been debated as being too low (O’Bryant et al. 2008). The MMSE has been widely adopted and translated into at least 66 languages (Steis and Schrauf 2009).
MMSE scores have been shown to be impacted by age and education, such that increased age is associated with lower scores and increased education is associated with higher scores (Anthony et al. 1982). Norms stratified for age and education are available for use (Bravo and Hebert 1997; Crum et al. 1993).
The MMSE was originally validated by administration to 206 patients with affective disorders, dementia symptoms, affective disorders with cognitive impairment, mania, schizophrenia, and personality disorders, as well as in 63 individuals with normal cognitive and emotional functioning (Folstein et al. 1975). Participants in the validation sample were divided into two groups: sample A consisted of patients chosen for clinical conditions (dementia, depression with cognitive impairment, affective disorder with depression, normal) and sample B which was intended as a standardization sample of patients who were consecutively admitted to the hospital (Folstein et al. 1975). The MMSE was able to separate the diagnostic groups within sample A from each other and from the individuals who were controls (Folstein et al. 1975). The MMSE was also able to differentiate severity of impairment: out of a total of 30 points, the mean MMSE score for those with normal functioning was 27.6, depression was 25.1, depression with cognitive impairment was 19.0, and dementia was 9.7 (Folstein et al. 1975). None of the normal older adults scored below a 24, thus establishing the clinical cutoff score for differentiating cognitive impairment (Folstein et al. 1975). The authors determined concurrent validity through correlating the MMSE with the Wechsler Adult Intelligence Scale-Verbal and Performance IQ scores (Wechsler 1955). When the MMSE was compared to Verbal IQ, Pearson r was 0.776 (p < 0.0001), and for MMSE compared to Performance IQ, Pearson r was 0.660 (p < 0.001) (Folstein et al. 1975). In regard to reliability, the authors found that when administered 24 h apart, intra-rater reliability was 0.887 (p < 0.0001) and inter-rater reliability was 0.827 (p < 0.001); when administered 28 days apart, there was no statistical difference in the scores (correlation 0.98) (Folstein et al. 1975).
Key Research Findings
Since its creation in 1975, the MMSE has been placed under considerable research to determine its utility for assessing and identifying patients with various levels of cognitive impairment. In an outpatient setting with individuals 65 years of age and older, the MMSE was found to have high sensitivity and specificity when differentiating patients with dementia from those without (Harvan and Cotter 2006). Additional research found that when screening for dementia in an elderly Swedish population, a cutoff score of 23/24 out of 30 had a sensitivity of 87%, specificity of 92%, and positive predictive value of 69% (Grut et al. 1993).
While the MMSE is shown to be effective at differentiating individuals with dementia from individuals without, research has demonstrated that it may not be as effective at distinguishing subtle cognitive decline or psychiatric conditions as originally suggested. The MMSE has been shown to lack sensitivity in distinguishing amnestic syndromes such that in 1 study assessing 11 amnestic patients, 6 patients obtained scores above the clinical cutoff score of 24, and 3 patients performed the memory task without error despite documentation of a memory disorder (Benedict and Brandt 1992). In research assessing individuals with multiple sclerosis, it was found that none of the participants scored below the standard cutoff score of 24, yet 8% percent of the patients in the study demonstrated cognitive impairment on other assessments of attention and speed (Franklin et al. 1988). Despite commonly being used as a screening instrument for cognitive changes in acute stroke patients, research on the efficacy of the MMSE in this population has been mixed. While one study found that some acute stroke patients were not even able to complete the MMSE (Pederson et al. 1996), another found that the accuracy of detecting cognitive impairment has been shown to be no better than chance in this population (Nys et al. 2005).
Recently, a primary goal of research regarding the MMSE has focused on comparison of the MMSE to a newer cognitive screening tool, the Montreal Cognitive Assessment (MoCA; See “Montreal Cognitive Assessment”; Nasreddine et al. 2005). The MoCA is a 10-min cognitive screening tool that was designed to detect mild cognitive impairment, a clinical state that is often seen to progress to dementia (Nasreddine et al. 2005). When comparing the MMSE to the MoCA, it has been found that both the MMSE and MoCA are effective at detecting dementia (Nasreddine et al. 2005); however, the MoCA has higher sensitivity for the detection of mild cognitive impairment or more subtle cognitive changes (Hoops et al. 2009; Dong et al. 2010; Roalf et al. 2013). In sum, research assessing the utility of the MMSE has found that while it is effective at distinguishing individuals with normal cognition or subtle cognitive impairments from individuals with dementia, it lacks the necessary specificity and sensitivity to accurately identify individuals with subtle cognitive changes or mild cognitive impairment.
Examples of Application
Though primarily used in conjunction with other neuropsychological measures (See “NIH Toolbox: Cognitive Battery”) to establish the cognitive capabilities of older adults (Lezak et al. 2012), the MMSE has been applied to functional settings as well. In the Rotterdam Study, a prospective study of community-dwelling older adults, the MMSE was able to predict compliance with antihypertensive drug regiments in individuals who live alone (Salas et al. 2001). Furthermore, a critical role of rehabilitation psychologists when working with older adults is to be aware of cognitive functioning abilities (Ruchinskas and Curyto 2003). The MMSE has been shown to provide valuable information to the clinician regarding expected course of rehabilitation and recovery (Ruchinskas and Curyto 2003). Cognitive ability as demonstrated through MMSE performance is associated with both longer hospital stay and rehabilitation course in a population of older adults recovering from hip fracture (Huusko et al. 2000). In acute stroke patients, MMSE scores were able to predict the chance for discharge to independent facilities after rehabilitation (Pederson et al. 1996). In a Swedish population, a strong association was demonstrated between MMSE score and annual cost of care for older adults with Alzheimer’s disease such that for every 1 point decrease in MMSE score, there was an association with a 15,000 Swedish kronor increase in cost of care (equivalent to $2000 US; Jonsson et al. 1999.
Future Directions of Research
Additional research is being conducted on the utility or inutility of the MMSE to differentiate patients with suspected dementia or mild cognitive impairment and to determine appropriate trade-offs in specificity and sensitivity (Larner 2018). Future research should focus on continuing to determine which patient populations are most appropriately distinguished through the MMSE, especially given its ability to be administered in various languages making it a highly useful clinical tool.
Since its creation by Folstein et al. (1975), the Mini-Mental State Examination has been widely utilized and accepted as a global cognitive screening tool in older adults that assesses the domains of orientation, learning and memory, attention and calculation, language, and visuospatial functioning. It has been shown to effectively differentiate older adults with normal cognitive functioning from older adults with dementia, though it may not be sensitive enough to detect subtle cognitive decline.
- Lezak M, Howieson D, Bigler E, Tranel D (2012) Neuropsychological assessment, 5th edn. Oxford University Press, New YorkGoogle Scholar
- Pedersen PM, Jørgensen HS, Nakayama H, Raaschou HO, Olsen TS (1996) General cognitive function in acute stroke: the Copenhagen stroke study. J Neurol Rehabil 10(3):153–158Google Scholar
- Wechsler D. (1955). Manual for the Wechsler Adult Intelligence Scale. New York: The Psychological CorporationGoogle Scholar