Downloaded from pn.bmj.

com on 17 January 2007

The Mini Mental State Examination

Basil Ridha and Martin Rossor

Practical Neurology 2005;5;298-303

Updated information and services can be found at:

http://pn.bmj.com/cgi/content/abstract/5/5/298

These include:
Rapid responses You can respond to this article at:
http://pn.bmj.com/cgi/eletter-submit/5/5/298

Email alerting Receive free email alerts when new articles cite this article - sign up in the box at
service the top right corner of the article

Notes

To order reprints of this article go to:

http://www.bmjjournals.com/cgi/reprintform
To subscribe to Practical Neurology go to:
http://www.bmjjournals.com/subscriptions/
 Downloaded from pn.bmj.com on 17 January 2007
298 PRACTICAL NEUROLOGY
HOW TO DO IT
The Mini Mental S
Basil Ridha* and Martin Rossor†
*Clinical Research fellow, †Professor of Clinical Neurology, Dementia
Research Centre, National Hospital for Neurology and Neurosurgery, 8
Queen Square, London WC1N 3BG; E-mail: bridha@dementia.ion.ucl.ac.uk
Practical Neurology, 2005, 5, 298–303
INTRODUCTION
Folstein and colleagues published the Mini
Mental State examination (MMSE) for brief
quantitative assessment of cognitive function
30 years ago (Folstein et al. 1975). Since then
it has been translated into many languages and
has become the most widely used brief test of
cognition in clinical and research settings, with
adequate validation. The aim of the examina-
tion is to:
• screen for cognitive impairment
• assess the severity of any impairment
• monitor change by serial testing
It relies heavily on verbal cognitive function
at the expense of nondominant hemisphere
skills and is vulnerable to the vagaries of scor-
ing by different observers. Nevertheless, it has
stood the test of time and provides a common
and widely understood tool to measure global
cognitive function.
STRUCTURE
The MMSE is a collection of questions that test
various cognitive domains including orienta-
tion to time and place, repetition, verbal recall,
attention and calculation, language and visual
construction. The test takes 5–10 min to ad-
minister and the total test score ranges from 0
(impaired) to 30 (normal).
© 2005 Blackwell Publishing Ltd
REQUIREMENTS FOR
PERFORMING THE MMSE
• Privacy in a quiet, well-lit room with no dis-
tractions.
• Pen or pencil, watch, and a blank piece of
paper.
• The subject is conscious and alert.
• The subject has sufficient knowledge of Eng-
lish and literacy skills to be able to cooperate
with test items (a translated version of the
MMSE administered by a trained translator
may be used where available).
• The subject has adequate hearing, vision,
articulation and dominant limb ability to
perform the MMSE tasks other than those
thought to be related to cognitive impair-
ment. Hearing and visual aids should be worn
if needed.
• Consistency in the examiner’s way of admin-
istering and scoring the MMSE to minimize
variability.
HOW TO CONDUCT THE MMSE
The tester needs to establish rapport with the
subject in order to put him or her at ease, and
obtain cooperation. A simple explanation of
the purpose and structure of the test is provided
first. The subject is then asked to answer the
questions to the best of their ability and to guess
 Downloaded from pn.bmj.com on 17 January 2007
State Examination
if unsure. If next of kin are present, it is impor-
tant to ask them not to assist the subject, unless
for help with translation when required. The test
is not timed, but it is sensible to assume a score
of zero if the subject does not come up with an
answer after 10–20 s. Whenever possible, record
the patient’s exact response, and why he or she
was not able to perform a task.
MMSE SUBTESTS
The MMSE test items are classified under the
following headings:
Orientation to time
Ask the subject in turn, ‘what is the (year) (sea-
son) (month) (date) (day)?’ One should allow
for seasonal transitions in the 2-week period
spanning any seasonal change. In countries
where other seasonal divisions are used, these
can be accepted as alternatives, e.g. monsoon
season.
Give one point for each correct answer to a
maximum score of 5.
Orientation to place
Ask the subject in turn about the state/country,
city/county, area of city/town, building/street
name, floor of building/house number or type
of room if in subject’s house. The set of ques-
tions will vary depending on whether the test
is conducted in a large city or a small town, and
whether in a hospital or the subject’s home.
Give one point for each correct answer to a
maximum score of 5.
Repetition/registration
Ask the subject to listen carefully because you
will ask him or her to repeat after you three
words. Speak loudly and clearly and pause for 1
OCTOBER 2005 299
second after each word. Three unrelated words
are used, most commonly ‘apple, penny, table’.
Give one point for each correct answer to a
maximum score of 3, irrespective of recalled
word order. Phonemic or paraphasic errors
should be penalized. If the subject is not able
to repeat all three words on the first trial, repeat
them again up to five times until he or she is able
to say them all back to you. But the performance
on only the first repetition is included in the total
MMSE score.
Attention and calculation
Ask the subject to take away 7 from 100. Ask him
or her to take away another 7 and continue to a
total of 5 subtractions. An answer is considered
correct if it is exactly 7 less than the previous
answer.
Give one point for each correct answer to a
maximum score of 5.
If a subject does not attempt the test or does
not get a full score, then he or she is asked to spell
WORLD, first forwards correcting any mistakes
and then backwards. One point is given for each
letter spelled in the correct order backwards to
a maximum score of 5. If more or less then 5
letters are given, then one point is deducted for
each extra or missing letter.
Only use the best score of either the serial
7 s or spelling WORLD backwards in the total
MMSE score.
Recall
Ask the subject to recall the three words you
earlier asked him or her to remember, without
prompting or giving any clues.
Give one point for each correct answer to a
maximum score of 3, irrespective of recalled
word order.
© 2005 Blackwell Publishing Ltd
 Downloaded from pn.bmj.com on 17 January 2007
300 PRACTICAL NEUROLOGY
Naming
Point to a pen or pencil, and a watch, in turn and ask the subject
to name them. Glasses or keys are possible substitutes.
Give one point for each correct answer to a maximum score
of 2.
Repetition
Ask the subject to listen carefully and repeat what you are about to
say. Say ‘no ifs ands or buts’ making sure all the ‘s’ letters are clearly
articulated and pausing for a second after each word.
Score zero unless all the letters are clearly articulated in the cor-
rect order. A maximum score of one point is given.
Comprehension
Ask the subject to listen carefully and do what you are about to
ask. Say clearly and loudly ‘ take this paper with your right hand,
fold it in half and put it on the floor’ (Fig. 1). Do not remind the
subject what to do next if he or she forgets or does not understand
what is required.
Score one point for each step performed correctly. If the right
hand is disabled, then ask the subject to take the paper with his
or her left hand.
Reading
Show the subject a blank piece of paper with ‘CLOSE YOUR
EYES’ clearly written. Ask the subject to read and then do what the
Figure 1 A patient performing the MMSE.
© 2005 Blackwell Publishing Ltd
sentence says. Make sure that there is sufficient lighting and the
subject is wearing his or her reading glasses if required. Provide
the command in the subject’s first language if he or she cannot
read English.
Only give one point if the patient closes his or her eyes.
Writing
Ask the subject to write a sentence on a blank piece of paper. Again
make sure there is sufficient lighting, and reading glasses are worn
if needed. If the subject does not respond then prompt him or her
to write something about the weather or a pet.
Only give a score of one point if he or she writes a compre-
hensible sentence containing a subject and verb. Ignore minor
spelling, grammatical or punctuation errors. Ignore quality of
handwriting including tremor, letter size and linearity. Examples
of acceptable and unacceptable sentences are shown in Fig. 2.
Drawing
Show the subject two clearly drawn intersecting pentagons, with
each side about 2 cm long. Ask him or her to copy the diagram
on a blank piece of paper.
A score of one point is given only if two five-sided figures are
drawn overlapping to form a four-sided embedded figure. Scal-
ing, side length, rotation, and tremor are ignored. Examples of
acceptable and unacceptable drawings are shown in Fig. 3.
ALLOWANCE FOR SCORING
It is important to remember that the purpose of the MMSE is to
test cognitive function, not any noncognitive disability. This has
an impact on whether to modify the test items or allow for any
inability to cooperate despite intact cognitive function. Exam-
ples include visual or hearing impairment, aphonia or dysarthria,
and dominant limb disability. However, if the task cannot be per-
formed due to possible cognitive dysfunction then the subject
must be scored accordingly.
One should also consider the exact domain of cognition that
is being tested. An aphasic subject who fails on tests of repetition,
may still be orientated in time and place and be able to recall
the 3 words if allowed to write his or her responses. In this case
it is more valid to document the written rather than the verbal
response.
Overall, it is important to document why a subject cannot per-
form a task, particularly if the cause is thought not to be due to an
underlying cognitive deficit, and to record any minor modifica-
tions to the test items used. Although cut-off scores for dementia
may not be applicable in such subjects, one could still monitor
change with time on repeat testing – provided consistency in test-
ing and scoring is maintained.
FACTORS THAT INFLUENCE MMSE SCORE
Age
MMSE scores tend to decline with increasing age, particularly
over the age of 65 years – even after controlling for education. In
addition, the range of scores tends to get wider with age (Crum
 Downloaded from pn.bmj.com on 17 January 2007
(a)
(b)
Figure 2 (a) Acceptable sentences. (b) Unacceptable sentences.
et al. 1993). Although decline in performance on the MMSE may
reflect normal ageing, it is worth noting that increasing age is of
course associated with increasing risk of dementia.
Education
With increasing educational level, the MMSE scores tend to in-
crease and the range of scores to narrow (Crum et al. 1993). Thus,
education introduces a bias resulting in a false positive diagnosis
of dementia in less educated subjects. Conversely, high education
level may mask mild cognitive impairment and result in a false
negative diagnosis. Education does not seem to affect the rate of
OCTOBER 2005 301
(a)
(b)
Figure 3 (a) Acceptable copying of the intersecting pentagons. (b)
Unacceptable copying of the intersecting pentagons.
change of MMSE scores in either cognitively normal or dementia
subjects, and so repeat MMSE testing is not confounded (Helkala
et al. 2002).
Gender
There seems to be no significant difference in MMSE scores be-
tween men and women. However, in one study men tended to
make more mistakes on spelling and other language tasks where-
as women were more likely to err on serial subtractions (Jones
& Gallo 2002). Nonetheless, the consensus is that gender is not
a determinant of MMSE performance (Tombaugh & McIntyre
1992).
Socioeconomic status
Subjects from lower socioeconomic classes tend to have a lower
MMSE score (Brayne & Calloway 1990; O’Connor et al. 1989).
This may be in part be because of and so confounded by poor
education. Other possible explanations could be developmental
© 2005 Blackwell Publishing Ltd
 Downloaded from pn.bmj.com on 17 January 2007
302 PRACTICAL NEUROLOGY
cognitive limitations predisposing to lower socioeconomic at-
tainment, or vice versa, lower socioeconomic status associated
with a higher risk of developing dementia.
Culture, language and ethnicity
There is some suggestion that MMSE scores are affected by race
and ethnicity. MMSE scores are reported to be lower in older
black and other ethnic minority groups (Parker & Philp 2004;
Espino et al. 2001). However, most of the difference is thought
to reflect lower educational level, socioeconomic status, type of
neighbourhood, and reduced English language fluency in immi-
grants compared to the indigenous white population.
Test location
Subjects tend to perform better at home than in hospital, prob-
ably due to better orientation and reduced anxiety. Therefore, it is
important to maintain consistency in the test location on repeat
testing in order to reduce variability in MMSE performance.
Test repetition
Non-demented subjects typically improve their performance if the
test is re-administered within a two-month interval. However, this
effect is less prominent in patients with dementia where learning
is defective and counterbalanced by disease progression with time
(Helkala et al. 2002; Galasko et al. 1993). Repeat presentation of the
MMSE within a short interval (under two months) improves the
accuracy of the MMSE as a screening tool for early dementia (Hel-
kala et al. 2002). In order to reduce the bias of learning, one can use
different words for the recall task, and different common objects
to name on each testing occasion. However, there are no parallel
forms for the other components of the MMSE.
TEST PERFORMANCE
Internal consistency between the test domains
There seems to be a good degree of homogeneity (internal consist-
ency) between the various test items of the MMSE which suggests
that the test is reliable. However, as the MMSE is meant to meas-
ure different cognitive domains, some degree of heterogeneity
would be desirable. The degree of internal consistency depends
on several factors. It tends to be higher in hospital-based studies
compared with community-based studies, and among subjects
with lower education than those with higher education. The use
of serial sevens alone as opposed to the combination with spelling
WORLD backwards yields a higher level of internal consistency
(Espino et al. 2004).
Test-retest reliability
For a test-retest interval of 2 months or less, the correlation coef-
ficient between each subject’s baseline and repeat MMSE scores
is about 0.9 (Folstein et al. 1975; Tombaugh & McIntyre 1992) in
both normal and demented subjects, suggesting good test-retest
reliability. The change in score is partly due to improvement in
performance as a result of practice. With a test-retest interval of
about 2 years, the correlation coefficient is less but still acceptable
© 2005 Blackwell Publishing Ltd
at 0.8 in nondemented subjects (Hopp et al. 1997). Repeat MMSE
testing over a 5-year period in normal subjects over the age of
65 years shows a slight improvement in the first year, probably
reflecting a practice effect, followed by slight decline over the fol-
lowing 4 years (Jacqmin-Gadda et al. 1997).
Inter-rater reliability
Folstein et al. found a high inter-rater reliability coefficient of 0.8
(Folstein et al. 1975). However, this has subsequently been found
to be significantly lower (Bowie et al. 1999), mostly due to vari-
ability in the scoring criteria and the version of the attention test
used, i.e., whether serial 7 s alone, spelling WORLD backwards
alone, or both (Davey & Jamieson 2004).
Validity
The sensitivity and specificity of the MMSE in differentiating de-
mentia from normality depends on the cut off score, pattern of
cognitive impairment present, disease severity, and subject charac-
teristics such as age, education, cultural influence, language fluency
and test location. Using a cut-off score of below 24, the MMSE
is 87% sensitive and 82% specific in detecting dementia and de-
lirium among inpatients on a general medical ward (Anthony et al.
1982). A higher cut-off score increases sensitivity of detecting mild
dementia, particularly in young educated individuals, but reduces
the specificity and increases the false positive diagnosis of dementia
in older uneducated individuals and in those belonging to ethnic
minority groups. The MMSE is more specific and sensitive in de-
tecting moderate to severe dementia than mild forms of dementia,
and in detecting cognitive deficits related to language and memory
rather than frontal executive dysfunction.
Several approaches have been used to improve the validity of
the MMSE. One is to generate population-based normative data
to provide different cut-off scores, stratified according to age
and education level (Crum et al. 1993). However, it is doubtful
if such adjustments improve the accuracy of the MMSE (Jones
& Gallo 2002). Another approach is to use modified forms of the
MMSE to minimize educational and cultural bias (Xu et al. 2003).
Modifications to the MMSE have also been suggested in order to
improve the accuracy of differentiating Alzheimer’s disease from
normality, including the use of differential weighting of different
test items (Galasko et al. 1990), adding extra test items such as
word fluency (Galasko et al. 1990), date and place of birth, simi-
larities and delayed recall of words (Teng & Chui 1987), and tak-
ing into account an informant rating score (Tierney et al. 2003).
Although these modifications tend to produce subtle improve-
ments, they prolong the testing time and are not as popular as the
original MMSE.
PATTERNS OF COGNITIVE DECLINE IN
DEMENTIA SYNDROMES
Patients with Alzheimer’s disease tend to perform worst on re-
call of three words and orientation to time (Galasko et al. 1990;
Jefferson et al. 2002), whereas patients with vascular dementia,
Parkinson’s disease and dementia with Lewy bodies tend to per-
form worst on attention, obeying a three step command, writing a
 Downloaded from pn.bmj.com on 17 January 2007
SUMMARY
• Despite its limitations, the MMSE is the most widely used
brief test of cognitive function
• Clinically, it provides a baseline framework to structure the
bedside cognitive assessment
• It provides an objective record to monitor disease progres-
sion and response to treatment
• It gives an estimate of disease severity that is meaningful for
communication among different health care workers
• Although modified versions of the MMSE do exist, the
original MMSE is still the most popular form
• The utility of the MMSE in clinical practice can be improved
by bearing in mind patient characteristics that may affect
performance (age, education, culture, ethnicity, hearing or
visual impairment and any other relevant physical disabili-
ties), and by using a standardized scoring scheme within
and between physicians
sentence and copying intersecting pentagons, suggesting impair-
ment of working memory and motor, constructional and visuo-
perceptual skills (Jefferson et al. 2002; Ala et al. 2002). Patients
with fronto-temporal dementia tend to have higher MMSE scores
in comparison to Alzheimer patients at initial presentation, and
slower annual rates of decline (0.9 points vs. 2.0, respectively).
This reflects the fact that the MMSE is geared more towards de-
tecting impairments of orientation, memory and language than
the executive dysfunction typical of fronto-temporal dementia.
MMSE AND THE USE OF
CHOLINESTERASE INHIBITORS
Cholinesterase inhibitors such as donepezil, rivastigmine and
galantamine have emerged as modestly useful agents for the
symptomatic treatment of mild to moderate Alzheimer’s disease
(Lanctot et al. 2003; Seltzer et al. 2004). Their cost implications,
and the high disease prevalence in the elderly, have reinforced
the need for brief and reliable tools for assessing eligibility for
treatment, monitoring drug efficacy, and withdrawing treatment
– such as the MMSE.
REFERENCES
Ala TA, Hughes LF, Kyrouac GA, Ghobrial MW & Elble RJ (2002) The Mini-
Mental State exam may help in the differentiation of dementia with Lewy
bodies and Alzheimer’s disease. International Journal of Geriatric Psychia-
try, 17, 503–9
Anthony JC, LeResche L, Niaz U, von Korff MR & Folstein MF (1982) Limits
of the ‘Mini-Mental State’ as a screening test for dementia and delirium
among hospital patients. Psychological Medicine, 12, 397–408
Bowie P, Branton T & Holmes J (1999) Should the Mini Mental State Exami-
nation be used to monitor dementia treatments? Lancet, 354, 1527–8
Brayne C & Calloway P (1990) The association of education and socioeco-
nomic status with the Mini Mental State Examination and the clinical
diagnosis of dementia in elderly people. Age Ageing, 19, 91–6
Crum RM, Anthony JC, Bassett SS & Folstein MF (1993) Population-based
norms for the Mini-Mental State Examination by age and educational
level. Journal of the American Medical Association, 269, 2386–91
OCTOBER 2005 303
Davey RJ & Jamieson S (2004) The validity of using the mini mental state
examination in NICE dementia guidelines. Journal of Neurology, Neuro-
surgery and Psychiatry, 75, 343–4
Espino DV, Lichtenstein MJ, Palmer RF & Hazuda HP (2001) Ethnic differ-
ences in mini-mental state examination (MMSE) scores: where you live
makes a difference. Journal of the American Geriatrics Society, 49, 538–48
Espino DV, Lichtenstein MJ, Palmer RF & Hazuda HP (2004) Evaluation of
the mini-mental state examination’s internal consistency in a commu-
nity-based sample of Mexican-American and European-American elders:
results from the San Antonio Longitudinal Study of Aging. Journal of the
American Geriatrics Society, 52, 822–7
Folstein MF, Folstein SE & McHugh PR (1975) ‘Mini-mental state’. A practi-
cal method for grading the cognitive state of patients for the clinician.
Journal of Psychiatric Research, 12, 189–98
Galasko D, Abramson I, Corey-Bloom J & Thal LJ (1993) Repeated exposure
to the Mini-Mental State Examination and the Information-Memory-
Concentration Test results in a practice effect in Alzheimer’s disease. Neu-
rology, 43, 1559–63
Galasko D, Klauber MR, Hofstetter CR, Salmon DP, Lasker B & Thal LJ (1990)
The Mini-Mental State Examination in the early diagnosis of Alzheimer’s
disease. Archives of Neurology, 47, 49–52
Helkala EL, Kivipelto M & Hallikainen M et al. (2002) Usefulness of repeated
presentation of Mini-Mental State Examination as a diagnostic procedure
– a population-based study. Acta Neurologica Scandinavica, 106, 341–6
Hopp GA, Dixon RA, Grut M & Backman L (1997) Longitudinal and psy-
chometric profiles of two cognitive status tests in very old adults. Journal
of Clinical Psychology, 53, 673–86
Jacqmin-Gadda H, Fabrigoule C, Commenges D & Dartigues JF (1997) A 5-
year longitudinal study of the Mini-Mental State Examination in normal
aging. American Journal of Epidemiology, 145, 498–506
Jefferson AL, Cosentino SA & Ball SK et al. (2002) Errors produced on the
mini-mental state examination and test performance in Alzheimer’s
disease, ischemic vascular dementia, and Parkinson’s disease. Journal of
Neuropsychiatry and Clinical Neurosciences, 14, 311–20
Jones RN & Gallo JJ (2002) Education and sex differences in the mini-mental
state examination: effects of differential item functioning. Journal of Ger-
ontological B Psychological Science Society of Science, 57, 548–58
Lanctot KL, Herrmann N & Yau KKet al. (2003) Efficacy and safety of
cholinesterase inhibitors in Alzheimer’s disease: a meta-analysis. Cana-
dian Medical Association Journal, 169, 557–64
O’Connor DW, Pollitt PA, Treasure FP, Brook CP & Reiss BB (1989) The
influence of education, social class and sex on Mini-Mental State scores.
Psychological Medicine, 19, 771–6
Parker C & Philp I (2004) Screening for cognitive impairment among older
people in black and minority ethnic groups. Age Ageing, 33, 447–52
Seltzer B, Zolnouni P & Nunez Met al. (2004) Efficacy of donepezil in early-
stage Alzheimer disease: a randomized placebo-controlled trial. Archives
of Neurology, 61, 1852–6
Teng EL & Chui HC (1987) The Modified Mini-Mental State (3MS) exami-
nation. Journal of Clinical Psychiatry, 48, 314–8
Tierney MC, Herrmann N, Geslani DM & Szalai JP (2003) Contribution of
informant and patient ratings to the accuracy of the mini-mental state
examination in predicting probable Alzheimer’s disease. Journal of the
American Geriatrics Society, 51, 813–8
Tombaugh TN & McIntyre NJ (1992) The mini-mental state examination:
a comprehensive review. Journal of the American Geriatrics Society, 40,
922–35
Xu G, Meyer JS, Du Huang YF, Chowdhury M & Quach M (2003) Adapting
mini-mental state examination for dementia screening among illiterate
or minimally educated elderly Chinese. International Journal of Geriatric
Psychiatry, 18, 609–16
© 2005 Blackwell Publishing Ltd