/clinical/,/clinical/cckm-tools/,/clinical/cckm-tools/content/,/clinical/cckm-tools/content/questionnaires/,/clinical/cckm-tools/content/questionnaires/related/,

/clinical/cckm-tools/content/questionnaires/related/name-97063-en.cckm

201606168

page

100

UWHC,UWMF,

Tools,

Clinical Hub,UW Health Clinical Tool Search,UW Health Clinical Tool Search,Questionnaires,Related

Comparison of SLUMS and MMSE for Detecting Dementia and Mild Neurocognitive Disorder

Comparison of SLUMS and MMSE for Detecting Dementia and Mild Neurocognitive Disorder - Clinical Hub, UW Health Clinical Tool Search, UW Health Clinical Tool Search, Questionnaires, Related


Comparison of the Saint Louis University
Mental Status Examination and the Mini-
Mental State Examination for Detecting
Dementia and Mild Neurocognitive
Disorder—A Pilot Study
Syed H. Tariq, M.D., F.A.C.P., Nina Tumosa, Ph.D.,
John T. Chibnall, Ph.D., Mitchell H. Perry III, M.D.,
John E. Morley, M.B., B.Ch.
Context: The Mini-Mental State Examination (MMSE) is commonly used as a screen-
ing tool to detect dementia. However, it performs poorly in identifying persons with
mild neurocognitive disorder. The Saint Louis University Mental Status (SLUMS)
examination is a 30-point screening questionnaire that tests for orientation, mem-
ory, attention, and executive functions. Objective: The objective of this study was to
compare SLUMS and the MMSE for detecting dementia and mild neurocognitive
disorder (MNCD) using Diagnostic and Statistical Manual of Mental Disorders,
Fourth Edition (DSM–IV) criteria. Methods: Patients at the Veterans’ Affairs Geriat-
ric Research, Education and Clinical Center, St. Louis, MO (N�702) were clinically
classified as having normal cognitive functioning, MNCD, or dementia based on
DSM–IV criteria. The SLUMS and MMSE were administered for comparison. Results:
Mean age was 75.3 years (standard deviation: 5.5). Regarding education, 62.4% of the
sample had at least completed high school and 30.6% had not. Sensitivity and specificity
were calculated and receiver operator curves (ROCs) generated for SLUMS and MMSE as
a function of diagnosis (MCND versus dementia) and education. Both the SLUMS and
MMSE produced acceptable ROCs for the diagnosis of dementia, but the ROCs for SLUMS
were better than the MMSE for the diagnosis of MNCD in both education groups.
Conclusion: These results suggest that the SLUMS and MMSE have comparable sensitiv-
ities, specificities, and area under the curve in detecting dementia. Although the defini-
tion of MNCD is controversial, the authors believe that the SLUMS is possibly better at
detecting mild neurocognitive disorder, which the MMSE failed to detect, but this needs to
be further investigated. (Am J Geriatr Psychiatry 2006; 14:900–910)
Key Words: SLUMS-a new screening tool, dementia, mild neurocognitive disorder
Received January 18, 2006; revised February 22, 2006; accepted March 17, 2006. From the Division of Geriatric Medicine (SHT, NT, MHP, JEM)
and the Department of Psychiatry (JTC), St. Louis University School of Medicine, and The VAMC, GRECC Jefferson Barrack’s Medical Center, St.
Louis, MO (SHT, NT, MHP, JEM). Send correspondence and reprint requests to Dr. Syed H. Tariq, Assistant Professor, Division of Geriatric
Medicine, 1402 S Grand M-238, St. Louis University School of Medicine, St. Louis, MO 63104. e-mail: Tariqsh@slu.edu
? 2006 American Association for Geriatric Psychiatry
Am J Geriatr Psychiatry 14:11, November 2006900
REGULAR RESEARCH ARTICLES

T
he aging of the U.S. population has been accom-
panied by a dramatic increase in the prevalence
of dementia. Dementia is seen in 3%–11% of persons
older than 65 years of age and 25%–47% of those
older than 85 years of age.
1–5
In 1997, there were 2.3
million persons with Alzheimer disease (AD) in the
United States and 90% were 60 years of age or older.
6
Dementia causes a high burden of suffering for pa-
tients, their families, and society.
7–11
The annual cost
of dementia is approximately $100 billion.
2,7
Never-
theless, dementia is underdiagnosed as evidenced by
the fact that 50% of persons with dementia, including
mild to moderate forms of dementia, have never
received the diagnosis of dementia from their physi-
cian.
12–15
A transitional period exists between normal aging
and the diagnosis of clinical probable early AD. This
transitional zone is described using a variety of terms
such as mild neurocognitive impairment (MCI), de-
mentia prodrome, incipient dementia, isolated mem-
ory impairment, and the Diagnostic and Statistical
Manual of Mental Disorders, Fourth Edition (DSM–IV)
construct is called mild neurocognitive disorder.
Currently, there is no agreed-on definition of this
transitional state. The criterion for mild neurocogni-
tive disorder as defined by DSM–IV is shown in
Table 1. Similarly, Petersen has defined mild cogni-
tive impairment criteria, which consist of memory
complaints usually corroborated by an informant,
objective memory impairment for age, essentially
preserved general cognitive function, largely intact
functional activities, and being not demented.
16,17
In
this article, we use mild neurocognitive disorder
(MNCD) instead of MCI because we used DSM–IV
criteria for diagnostic purposes. The diagnostic crite-
rion for dementia of Alzheimer type as defined by
DSM–IV is shown in Table 2. Similar criteria for
other dementias were used as defined by DSM–IV.
A number of studies have addressed the conver-
sion of this transitional state of mild neurocognitive
disorder or mild cognitive impairment to dementia
using different criteria. In the Mayo Clinic’s longitu-
dinal studies of aging and dementia in community-
dwelling older adults, those who met the criteria for
mild cognitive impairment were diagnosed using the
following: Mini-Mental State Examination, WAIS–
Revised, Wechsler Memory Scale–Revised, Dementia
Rating Scale, Free and Cued Selective Reminding
Test, and Auditory Verbal Learning Test. Clinical
classifications of dementia and AD were determined
according to the Diagnostic and Statistical Manual of
Mental Disorders, Revised Third Edition and the Na-
tional Institute of Neurological and Communicative
Disorders and Stroke–Alzheimer’s Disease and Re-
lated Disorders Association criteria, respectively,
were converted to AD at a rate of 12% per year over
the course of 4 years.
18
Similarly, in a Canadian
study (using a research battery of neuropsychologic
tests), 28% of 107 subjects with memory impairment
without dementia converted to AD over a 2-year
period, for an annual conversion rate of 14%.
19,20
Other studies using similar methodology in older
patients with “questionable dementia,” “memory
impairment,” and “mild cognitive impairment” have
reported annual conversion rates to dementia of be-
tween 6% and 25%, with a mean conversion rate of
approximately 14% across studies.
21–24
Thus, early
detection of mild cognitive impairment with proac-
tive treatment has the potential to delay the onset of
AD in vulnerable populations in the future.
Given the importance of the detection of mild neu-
rocognitive disorder, a simple, efficient, and sensi-
tive/specific screening tool for this diagnosis is
TABLE 1. Diagnostic and Statistical Manual of Mental
Disorders, Fourth Edition Research Criteria for
Mild Neurocognitive Disorder
A. The presence of two (or more) of the following impairments in
cognitive functioning lasting most of the time for a period of 2
weeks (as reported by the individual or a reliable informant):
1) Memory impairment as identified by a reduced ability to
learn or recall information;
2) Disturbances in executive functioning (i.e., planning,
organizing, sequencing, abstracting);
3) Disturbances in attention or speed of information
processing;
4) Impairment in perceptual–motor abilities; and
5) Impairment in language (e.g., comprehension, word
finding).
B. There is objective evidence from physical examination or
laboratory findings (including neuroimaging) or general medical
condition that is judged to be etiologically related to the
cognitive disturbance.
C. There is evidence from neuropsychologic testing or quantified
cognitive assessment of an abnormality or decline in
performance.
D. The cognitive deficits causes marked distress or impairment in
social, occupational, or other important areas of functioning
and represent a decline from a previous level of functioning.
E. The cognitive disturbance does not meet criteria for delirium, a
dementia, or an amnestic disorder and is not better accounted
for by another mental disorder (e.g., a substance-related
disorder, major depressive disorder).
Tariq et al.
Am J Geriatr Psychiatry 14:11, November 2006 901

clearly needed, particularly for use in primary care.
The Mini-Mental State Examination (MMSE)
25
is
widely used in practice as a screening tool for de-
mentia. The MMSE has limitations, however, espe-
cially with regard to its use in more educated pa-
tients and as a screen for mild neurocognitive
disorder.
26–28
The Saint Louis University Mental Sta-
tus Examination (SLUMS) was developed to address
this limitation. The SLUMS is a 30-point, 11-item,
clinician-administered scale that is similar in format
to the MMSE.
29
The SLUMS, however, supplements
the MMSE with enhanced tasks corresponding to
attention, numeric calculation, immediate and de-
layed recall, animal naming, digit span, clock draw-
ing, figure recognition/size differentiation, and im-
mediate recall of facts from a paragraph. In
particular, the clock drawing test is designed to as-
sess impairment in executive function,
30
one of the
earliest forms of cognition affected in mild neurocog-
nitive disorder and dementia. The SLUMS also elim-
inates some tasks that appear on the MMSE, includ-
ing repetition and construction. By enhancing the
SLUMS relative to the MMSE, it was predicted that
the SLUMS would be more sensitive and specific
than the MMSE, particularly for diagnoses of mild
neurocognitive disorder. In the present study, the
SLUMS examination is compared with the MMSE
regarding sensitivity/specificity for mild neurocog-
nitive disorder and dementia in a large sample of
Veterans’ Administration patients. The Saint Louis
University Mental Status Examination is shown in
Figure 1.
METHODS
Seven hundred five participants in this study were
recruited prospectively from January 2003 to Sep-
tember 2003 at the Geriatric Research Education and
Clinical Center (GRECC), Veterans’ Affairs Medical
Center (VAMC) hospitals in Saint Louis, MO. The
recruitment process was conducted during a routine
clinic visit in which participants were informed
about the screening tests and consented as required
by the Institutional Review Board of the Saint Louis
VAMC. The investigators evaluated each participant
during a routine clinic visit and, in addition, a his-
tory was obtained from corroborating sources. In
addition, a complete physical and mental status ex-
amination was performed and laboratory findings
were reviewed.
31,32
The investigators used DSM–IV
criteria to make the diagnosis of mild neurocognitive
disorder or dementia. The investigators then com-
pleted the MMSE
24
and SLUMS.
28
Those who did not
qualify by any of DSM–IV criteria were considered
having normal cognition or no cognitive problems
without considering the scores of MMSE or SLUMS.
The inclusion criteria for the study was that the
participant should be 60 years of age or older and
must consent for the screening tests. The exclusion
criteria was refusal or to be unable to complete the
screening tests for any reason.
Data Analysis
Statistical analyses were conducted using a statis-
tical software package (SPSS 11.0; SPSS Inc., Chicago,
IL). Demographic data were summarized using de-
TABLE 2. Diagnostic Statistical Manual of Mental
Disorders, Fourth Edition Diagnostic Criteria for
Dementia of Alzheimer Type
A. The development of multiple cognitive deficits manifested by
both:
1. Memory impairment(impaired ability to learn new
information and to recall previously learned information);
and
2. One (or more) of the following cognitive disturbances:
a. Aphasia(language disturbances);
b. Apraxia (impaired ability to carry out motor activities
despite intact motor function);
c. Agnosia (failure to recognize or identify objects despite
intact sensory function); and
d. Disturbance in executive functioning (i.e., planning,
organizing, sequencing, abstracting).
B. The cognitive disturbance in criteria A1 and A2 each causes
significant impairment in social or occupational functioning and
represents a significant decline from previous level of
functioning.
C. The course is characterized by gradual onset and continuing
cognitive decline.
D. The cognitive deficits in criteria A1 and A2 are not the result of
any of the following:
1. Other central nervous system conditions that may cause
progressive deficit in memory and cognition (e.g.,
cerebrovascular disease, Parkinson disease, Huntington
disease, subdural hematoma, normal-pressure hydrocephalus,
brain tumor);
2. Systemic conditions that may cause dementia (e.g.,
hypothyroidism, vitamin B12 deficiency, niacin deficiency,
hypercalcemia, neurosyphilis, HIV infection); or
3. Substance-induced conditions.
E. The deficit does not occur exclusively during the course of
delirium.
F. The disturbance is not better accounted for by another axis I
disorder (e.g., major depression disorder, schizophrenia).
Comparison of the SLUMS and the MMSE
Am J Geriatr Psychiatry 14:11, November 2006902

FIGURE 1. Saint Louis University Mental Status Examination
Tariq et al.
Am J Geriatr Psychiatry 14:11, November 2006 903

scriptive statistics. Using the standard approach, sen-
sitivity, specificity, and positive and negative predic-
tive values were calculated for various cutoff scores
of the SLUMS and the MMSE as a function of diag-
noses of mild neurocognitive disorder and dementia
(against normal) for patients with less than high
school education and those with high school educa-
tion or more. Standard receiver operator curves
(ROCs) were plotted using sensitivity and 1-specific-
ity values. Area under the curve (AUC) was esti-
mated from the ROCs and evaluated for statistical
significance using the nonparametric method. The
latter method uses the trapezoidal rule for calculat-
ing area.
36
The obtained AUC is evaluated for statis-
tical significance against the null hypothesis that the
true AUC is 0.50. The significance test used in SPSS is
the equivalent of the nonparametric Wilcoxon statis-
tic.
36
RESULTS
The mean age of the sample was 75.3 years (standard
deviation [SD]: 5.5). Patients were classified by edu-
cation into those with less than a high school educa-
tion (N�216 [30.6%]) and those with a high school
education or more (N�489 [69.4%]). With respect to
cognitive status, 440 of the patients (62.4%) had nor-
mal cognitive functioning by DSM–IV criteria, 180
(25.5%) were diagnosed with mild neurocognitive
disorder, and 82 (11.6%) were diagnosed with de-
mentia (N�73 AD; N�7 vascular dementia; N�2
other dementia). Three patients (0.4%) were blind
and, therefore, a valid cognitive status score could
not be obtained using the study measures and were
excluded from the final analysis. The SLUMS exam-
ination required a mean of 7 minutes (SD: 3) to
complete. Table 3 displays the mean age, SLUMS
score, and MMSE score for each patient group as a
function of education and cognitive status. Using
analysis of variance and the Student-Newman-Keuls
post hoc test for multiple mean comparisons, de-
mented patients across both education groups were
found to be significantly older compared with those
with normal and mild neurocognitive disorder
(F2,696�22.9, p �0.001).
Sensitivity and Specificity
Tables 4 through 7 display the cutoff scores for the
SLUMS and the MMSE for various levels of sensitiv-
ity and specificity as a function of education and
cognitive status. The optimal cutoff scores for
SLUMS for MNCD with less than high school edu-
cation and high school or higher education were 23.5
and 25.5, respectively. The cutoff scores for SLUMS
for dementia were 19.5 and 21.5 for less than high
school education and high school or higher educa-
tion, respectively. The optimal cutoff score for MMSE
for MNCD with less than high school education and
high school education or higher were 28.5 and 29.5,
respectively. The cutoff scores for MMSE for demen-
tia were 26.5 and 27.5 for less than high school edu-
cation and high school or higher education, respec-
tively.
Receiver operator curves were calculated from
sensitivity and 1-specificity values for the SLUMS
and the MMSE for diagnoses of mild neurocognitive
disorder and dementia (against normal) for patients
with less than high school education and those with
high school education or more. Area under the curve
was estimated using the Wilcoxon method. Figures 2
and 3 contrast the AUCs for the SLUMS and MMSE
for patients with MNCD and those with dementia
with less than high school education. As shown in
Figure 2, for patients with less than high school
TABLE 3. Descriptive Data by Education and Cognitive Status (DSM–IV)
Less Than High School High School or More
Normal MNCD Dementia Normal MNCD Dementia
(N�137) (N�50) (N�27) (N�303) (N�130) (N�55)
Age, mean (SD) 74.6 (5.2) 77.8 (6.0) 79.1 (6.7) 74.2 (5.0) 75.6 (5.6) 77.8 (5.1)
SLUMS, mean (SD) 25.7 (2.8) 20.2 (2.4) 11.3 (5.1) 26.9 (2.0) 22.3 (2.1) 14.9 (5.2)
MMSE, mean (SD) 28.6 (1.7) 27.3 (2.4) 21.0 (7.4) 29.0 (1.3) 28.2 (1.7) 23.0 (5.0)
MNCD: mild neurocognitive disorder; SD: standard deviation; SLUMS: St. Louis University Mental Status examination; MMSE: Mini-Mental Status
Examination.
Comparison of the SLUMS and the MMSE
Am J Geriatr Psychiatry 14:11, November 2006904

education, the AUC for the SLUMS and MNCD was
0.927 (p �0.001) with a corresponding 95% confi-
dence interval (CI) of 0.891–0.963 based on 137 nor-
mal cases and 50 MNCD cases. The AUC for the
MMSE and MNCD was 0.671 (p �0.001; 95% CI:
0.582–0.759). As shown in Figure 3, for patients with
less than high school education, the AUC for the
TABLE 4. Cutoff Scores for Mild Neurocognitive
Impairment (MNCD) in Patients With Less Than
High School Education
Test/Category
Cutoff
Score
(<) Sensitivity Specificity PPV NPV
SLUMS/MNCD 19.5 0.40 0.98 0.91 0.82
20.5 0.56 0.96 0.82 0.86
21.5 0.70 0.91 0.74 0.89
22.5 0.78 0.86 0.67 0.91
23.5 0.92 0.81 0.64 0.97
24.5 0.96 0.74 0.57 0.98
25.5 1.00 0.55 0.45 1.00
26.5 1.00 0.40 0.38 1.00
27.5 1.00 0.29 0.34 1.00
MMSE/MNCD 21.0 0.02 1.00 1.0 0.74
22.5 0.04 1.00 1.0 0.74
23.5 0.04 0.99 0.75 0.74
24.5 0.20 0.96 0.62 0.77
25.5 0.24 0.92 0.52 0.77
26.5 0.30 0.87 0.45 0.77
27.5 0.46 0.79 0.44 0.80
28.5 0.60 0.65 0.38 0.82
29.5 0.82 0.39 0.33 0.86
PPV: positive predictive value; NPV: negative predictive value;
SLUMS: Saint Louis University Mental Status examination; MMSE:
Mini-Mental State Examination.
TABLE 5. Cutoff Scores for Dementia in Patients With Less
Than High School Education
Test/Category
Cutoff
Score
(<) Sensitivity Specificity PPV NPV
SLUMS/dementia 15.0 0.74 1.00 1.00 0.95
16.5 0.85 1.00 1.00 0.97
17.5 0.89 0.98 0.92 0.98
18.5 0.93 0.98 0.93 0.98
19.5 1.00 0.98 0.93 1.00
20.5 1.00 0.96 0.82 1.00
21.5 1.00 0.91 0.69 1.00
22.5 1.00 0.86 0.59 1.00
23.5 1.00 0.81 0.51 1.00
MMSE/dementia 21.5 0.37 1.00 1.00 0.89
22.5 0.41 1.00 1.00 0.90
23.5 0.56 0.99 0.94 0.92
24.5 0.63 0.96 0.74 0.93
25.5 0.78 0.92 0.66 0.95
26.5 0.81 0.87 0.55 0.96
27.5 0.85 0.79 0.44 0.96
28.5 0.89 0.65 0.33 0.97
29.5 1.00 0.39 0.25 1.00
PPV: positive predictive value; NPV: negative predictive value;
SLUMS: Saint Louis University Mental Status examination; MMSE:
Mini-Mental State Examination.
TABLE 6. Cutoff Scores for Mild Neurocognitive
Disorder (MNCD) in Patients With High School
Education or More
Test/Category
Cutoff
Score
(<) Sensitivity Specificity PPV NPV
SLUMS/MNCD 21.5 0.38 1.00 1.00 0.75
22.5 0.55 0.99 1.00 0.79
23.5 0.68 0.95 0.96 0.84
24.5 0.85 0.87 0.85 0.87
25.5 0.95 0.76 0.74 0.93
26.5 0.98 0.61 0.63 0.99
27.5 0.99 0.42 0.52 0.99
28.5 1.00 0.22 0.42 0.99
29.5 1.00 0.12 0.35 1.00
MMSE/MNCD 23.5 0.03 1.00 1.00 0.71
24.5 0.05 0.99 0.64 0.71
25.5 0.08 0.98 0.65 0.71
26.5 0.14 0.93 0.47 0.72
27.5 0.27 0.86 0.45 0.73
28.5 0.45 0.76 0.43 0.76
29.5 0.75 0.48 0.38 0.82
PPV: positive predictive value; NPV: negative predictive value;
SLUMS: Saint Louis University Mental Status examination; MMSE:
Mini-Mental State Examination.
TABLE 7. Cutoff Scores for Dementia in Patients With High
School Education or More
Test/Category
Cutoff
Score
(<) Sensitivity Specificity PPV NPV
SLUMS/dementia 17.5 0.58 1.00 1.00 0.93
18.5 0.78 1.00 1.00 0.96
19.5 0.93 1.00 1.00 0.99
20.5 0.96 1.00 1.00 0.99
21.5 0.98 1.00 1.00 1.00
22.5 0.98 0.99 0.95 1.00
23.5 0.98 0.95 0.77 1.00
24.5 0.98 0.87 0.58 1.00
25.5 0.98 0.76 0.42 1.00
MMSE/dementia 21.5 0.27 1.00 1.00 0.88
22.5 0.38 1.00 1.00 0.90
23.5 0.45 1.00 1.00 0.91
24.5 0.53 0.99 0.88 0.92
25.5 0.67 0.98 0.86 0.94
26.5 0.76 0.93 0.68 0.96
27.5 0.89 0.86 0.53 0.98
28.5 0.94 0.75 0.40 0.99
29.5 0.96 0.48 0.25 0.99
PPV: positive predictive value; NPV: negative predictive value;
SLUMS: Saint Louis University Mental Status examination; MMSE:
Mini-Mental State Examination.
Tariq et al.
Am J Geriatr Psychiatry 14:11, November 2006 905

SLUMS and dementia was 0.998 (p �0.001; 95% CI:
0.995–1.00) based on 137 normal cases and 27 demen-
tia cases. The AUC for the MMSE and dementia was
0.915 (p �0.001; 95% CI: 0.853–0.976). Figures 4 and
5 contrast the curves for patients with MNCD and
those with dementia with a high school education or
more. As shown in Figure 4, for patients with a high
school education or more, the AUC for the SLUMS
and MNCD was 0.941 (p �0.001; 95% CI: 0.919–
0.963) based on 303 normal cases and 130 MNCD
FIGURE 2. Receiver Operator Curve for the Saint Louis University Mental Status Examination and the Mini-Mental State
Examination for Mild Neurocognitive Disorder in Patients With Less Than High School Education
SLUMS: Saint Louis University Mental Status Examination; MMSE: Mini-Mental State Examination.
FIGURE 3. Receiver Operator Curve for the Saint Louis University Mental Status Examination and the Mini-Mental State
Examination for Dementia in Patients With Less Than High School Education
SLUMS: Saint Louis University Mental Status Examination; MMSE: Mini-Mental State Examination.
Comparison of the SLUMS and the MMSE
Am J Geriatr Psychiatry 14:11, November 2006906

cases. The AUC for the MMSE and MNCD was 0.643
(p �0.001; 95% CI: 0.587–0.700). As shown in Figure
5, for patients with a high school education of more,
the AUC for the SLUMS and dementia was 0.983 (p
�0.001; 95% CI: 0.950–1.00) based on 303 normal
cases and 55 dementia cases. The AUC for the MMSE
and dementia was 0.934 (p �0.001; 95% CI: 0.892–
0.977).
ROCs were calculated from sensitivity and 1-spec-
ificity values for the range of scores obtained on each
item of the SLUMS as a function of diagnosis and
education. AUC was estimated for each item using
the nonparametric approach. Table 8 summarizes the
results. For patients with less than a high school
education (137 normal, 50 MNCD, 27 dementia),
SLUMS items corresponding to animal naming, de-
FIGURE 4. Receiver Operator Curve for the Saint Louis University Mental Status Examination and the Mini-Mental State
Examination for Mild Cognitive Disorder in Patients With High School Education or More
SLUMS: Saint Louis University Mental Status Examination; MMSE: Mini-Mental State Examination.
FIGURE 5. Receiver Operator Curve for the Saint Louis University Mental Status Examination and the Mini-Mental State
Examination for Dementia in Patients With High School Education
SLUMS: Saint Louis University Mental Status Examination; MMSE: Mini-Mental State Examination.
Tariq et al.
Am J Geriatr Psychiatry 14:11, November 2006 907

layed recall, digit span, correct time on the clock
drawing, and the four items representing immediate
paragraph recall significantly discriminated normal
patients from patients with MNCD. Also for this
group, items corresponding to orientation to day of
the week, calculation, animal naming, delayed recall,
digit span, clock drawing, and immediate paragraph
recall significantly discriminated normal patients
from patients with dementia. For the group with a
high school education or more (303 normal, 130
MNCD, 55 dementia), SLUMS items corresponding
to the subtraction portion of the calculation problem,
animal naming, delayed recall, digit span, clock
drawing and time, and the immediate paragraph
recall significantly discriminated normal patients
from patients with MNCD. Also in this group,
SLUMS items corresponding to orientation to day of
the week, calculation, animal naming, delayed recall,
digit span, clock drawing, and immediate paragraph
recall significantly discriminated normal patients
from patients with dementia. Orientation to year and
state of residence, figure identification, and differen-
tiation of figure size did not discriminate normal
patients from either MNCD or dementia in both
education groups. Across both groups and diag-
noses, animal naming, delayed recall, digit span, and
immediate paragraph recall were the strongest and
most consistent discriminators.
DISCUSSION
The present analysis showed that the sensitivity and
specificity appear similar for both SLUMS and
MMSE in detecting dementia, but the SLUMS ap-
peared to be possibly better than the MMSE for
differentiating MNCD from normal cognitive func-
tioning From these data, we suggest different cutoff
scores for the SLUMS examination than initially
published before validation.
33
The scores for mild
neurocognitive disorder and dementia for patients
with less than high school education are 23.5 and
19.5, respectively. These cutoffs yield sensitivity/
TABLE 8. SLUMS Item Discrimination as a Function of Education and Cognitive Status (DSM–IV)
Less Than High School High School or More
MNCD Dementia MNCD Dementia
SLUMS Items
Area Under Curve
(95% CI)
a
p
Area Under Curve
(95% CI) p
Area Under Curve
(95% CI) p
Area Under Curve
(95% CI) p
Day of the week 0.51 (0.42–0.61) 0.83 0.65 (0.52–0.78) �0.05 0.51 (0.45–0.57) 0.74 0.61 (0.52–0.70) �0.05
Year 0.52 (0.42–0.62) 0.68 0.59 (0.46–0.72) 0.13 0.51 (0.45–0.57) 0.84 0.57 (0.48–0.66) 0.09
State 0.50 (0.41–0.59) 1.00 0.56 (0.43–0.68) 0.36 0.50 (0.44–0.56) 1.00 0.54 (0.46–0.63) 0.28
Calculation story problem
Money spent 0.52 (0.42–0.61) 0.69 0.67 (0.55–0.80) �0.01 0.53 (0.47–0.59) 0.36 0.62 (0.52–0.71) �0.01
Money left 0.57 (0.47–0.66) 0.16 0.78 (0.67–0.89) �0.001 0.57 (0.51–0.63) �0.05 0.71 (0.62–0.79) �0.001
Animal naming 0.62 (0.53–0.71) �0.01 0.85 (0.77–0.94) �0.001 0.66 (0.60–0.72) �0.001 0.84 (0.77–0.91) �0.001
Delayed recall—five objects 0.74 (0.66–0.82) �0.001 0.91 (0.86–0.96) �0.001 0.75 (0.70–0.80) �0.001 0.92 (0.87–0.96) �0.001
Digit span 0.65 (0.56–0.74) �0.01 0.72 (0.61–0.84) �0.001 0.63 (0.57–0.69) �0.001 0.80 (0.73–0.87) �0.001
Clock drawing—numbers
Numbers 0.54 (0.45–0.64) 0.35 0.81 (0.72–0.91) �0.001 0.56 (0.50–0.62) �0.05 0.70 (0.62–0.78) �0.001
Time 0.61 (0.52–0.71) �0.05 0.85 (0.76–0.94) �0.001 0.59 (0.53–0.65) �0.01 0.82 (0.75–0.89) �0.001
Identification—triangle 0.53 (0.44–0.63) 0.49 0.55 (0.42–0.67) 0.43 0.51 (0.45–0.57) 0.77 0.55 (0.46–0.64) 0.26
Differentiation—figure size 0.50 (0.41–0.59) 0.98 0.56 (0.44–0.69) 0.30 0.50 (0.44–0.56) 0.93 0.52 (0.44–0.61) 0.63
Immediate paragraph recall
Who (name) 0.61 (0.52–0.71) �0.05 0.78 (0.66–0.89) �0.001 0.57 (0.51–0.63) �0.05 0.68 (0.59–0.77) �0.001
What (work/job) 0.62 (0.52–0.72) �0.05 0.84 (0.75–0.93) �0.001 0.58 (0.52–0.64) �0.01 0.72 (0.64–0.81) �0.001
When (returned to work) 0.72 (0.63–0.81) �0.001 0.81 (0.71–0.91) �0.001 0.61 (0.55–0.67) �0.001 0.78 (0.70–0.86) �0.001
Where (state lived in) 0.68 (0.58–0.77) �0.001 0.83 (0.73–0.92) �0.001 0.69 (0.63–0.75) �0.001 0.76 (0.68–0.84) �0.001
a
Based on receiver operator curve (ROC) plots of sensitivity and 1-specificity. Area under the curve (AUC) and 95% confidence interval (CI)
were estimated using the nonparametric approach.
DSM-IV: Diagnostic and Statistical Manual of Mental Disorders-IV; SLUMS: Saint Louis University Mental Status Examination; MNCD: mild
neurocognitive disorder; CI: confidence interval.
Comparison of the SLUMS and the MMSE
Am J Geriatr Psychiatry 14:11, November 2006908

specificity values of 0.92/0.81 and 1.0 /0.98, respec-
tively. The cutoff scores for mild neurocognitive
disorder and dementia for patients with high school
education or higher are 25.5 and 21.5, respectively.
Sensitivity/specificity values for these cutoffs are
0.95/0.76 and 0.98/1.0, respectively. On the item dis-
crimination analysis, four items did not significantly
affect the outcomes for the SLUMS, namely orienta-
tion, year and state of residence, figure identification,
and differentiation of figure size. The present results
suggest that these items do not contribute to the
sensitivity/specificity of the SLUMS.
Both SLUMS and MMSE can be used by clinicians
to identify dementia, but SLUMS has the advantage
that it can help the clinician identify patients with
MNCD on the initial visit compared with MMSE,
which requires a follow-up screening and following
a decline in scores.
34
This early detection of MNCD
offers the opportunity for the clinicians to begin early
treatment for MNCD as they become available. Also,
MMSE has the limitation of a ceiling effect in patients
with higher education.
35
A limitation of this study is that the data were
obtained from primarily white, male patients at a VA
medical center. The extent to which the results re-
ported here generalize to other patient populations is
an open question for future research. An important
limitation is the current controversies over the defi-
nition of MNCD. The study methodology is another
limitation as the same clinicians who administered
the SLUMS and MMSE made the classifications of
normal cognitive functioning, MNCD, or dementia.
The lack of independence of the two activities intro-
duces a potential investigator expectation confound
into the data. Finally, the number of patients with
dementia and the number of patients with MNCD
with less than a high school education were rela-
tively low, prompting the need for further research
on the SLUMS in larger samples and different ethnic
backgrounds.
This study indicates that the SLUMS and MMSE
both can be used as a screening tool to detect demen-
tia. SLUMS also recognizes a group of patients with
mild neurocognitive disorder, which the MMSE
failed to recognize as defined by DSM–IV criteria.
This needs to be further investigated. Neither the
SLUMS nor the MMSE can substitute for clinical
assessment and neuropsychologic testing in the di-
agnosis of MNCD or dementia. The authors further
acknowledge that there is no single agreed-on defi-
nition of MNCD, and data from this study should be
used with caution when diagnosing mild neurocog-
nitive disorder that needs further investigation.
Presented at the 56th Annual Scientific meeting of
Gerontological Society of America, San Diego, 2003; and
the annual meeting of the American Geriatric Society, Las
Vegas, May 2004.
References
1. Evans DA, Funkenstein HH, Albert MS, et al: Prevalence of Alz-
heimer’s disease in a community population of older persons.
Higher than previously reported. JAMA 1989; 262:2551–2556
2. Evans DA, Smith LA, Scherr PA, et al: Risk of death from Alzhei-
mer’s disease in a community population of older persons. Am J
Epidemiol 1991; 134:403–412
3. US General Accounting Office: Alzheimer’s Disease: Estimates of
Prevalence in the United States. Washington, DC, US General
Accounting Office, 1998. Publication HEHS 98-16
4. Patterson CJ, Gauthier S, Bergman H, et al: The recognition,
assessment and management of dementing disorders: conclusions
from the Canadian Consensus Conference on Dementia. CMAJ
1999; 160:1–15
5. Canadian study of health and aging: study methods and preva-
lence of dementia. CMAJ 1994; 150:899–913
6. Brookmeyer R, Gray S, Kawas C: Projections of Alzheimer’s dis-
ease in the United States and the public health impact of delaying
disease onset. Am J Public Health 1998; 88:1337–1342
7. Ernst RL, Hay JW: The US economic and social costs of Alzhei-
mer’s disease revisited. Am J Public Health 1994; 84:1261–1264
8. Dunkin JJ, Anderson-Hanley C: Dementia caregiver burden: a
review of the literature and guidelines for assessment and inter-
vention. Neurology 1998; 51:53–60
9. Grafstro¨m M, Fratiglioni L, Sandman PO, et al: Health and social
consequences for relatives of demented and non-demented el-
derly. A population-based study. J Clin Epidemiol 1992; 45:861–
870
10. Schulz R, O’Brien AT, Bookwala J, et al: Psychiatric and physical
morbidity effects of dementia caregiving: prevalence, correlates,
and causes. Gerontologist 1995; 35:771–791
11. Gold DP, Reis MF, Markiewicz D, et al: When home caregiving
ends: a longitudinal study of outcomes for caregivers of relatives
with dementia. J Am Geriatr Soc 1995; 43:10–16
12. O’Connor DW, Pollitt PA, Hyde JB, et al: Do general practitioners
miss dementia in elderly patients? BMJ 1988; 297:1107–1110
13. Lagaay AM, van der Meij JC, Hijmans W: Validation of medical
history taking as part of a population based survey in subjects
aged 85 and over. BMJ 1992; 304:1091–1092
14. Cooper B, Bickel H, Schaufele M, et al: Early development and
progression of dementing illness in the elderly: a general-practice
based study. Psychol Med 1996; 26:411–419
15. Valcour VG, Masaki KH, Curb JD, et al: The detection of dementia
Tariq et al.
Am J Geriatr Psychiatry 14:11, November 2006 909

in the primary care setting. Arch Intern Med 2000; 160:2964–
2968
16. Petersen RC, Stevens JC, Ganguli M, et al: Practice parameter:
early detection of dementia: mild cognitive impairment (an evi-
dence based review): report of quality of the standards Subcom-
mittee of the American Academy of Neurology. Neurology 2001;
56:1133–1142
17. Petersen RC: Mild cognitive impairment as a diagnostic entity.
J Intern Med 2004; 256:183–194
18. Petersen RC, Smith GE, Waring SC, et al: Mild cognitive impair-
ment: clinical characterization and outcome. Arch Neurol 1999;
56:303–308
19. Tierney MC, Szalai JP, Dunn E, et al: Prediction of probable
Alzheimer’s disease in memory-impaired patients: a prospective
longitudinal study. Neurology 1996; 46:661–665
20. Tierney MC, Szalai JP, Snow WG, et al: A prospective study of the
clinical utility of ApoE genotype in the prediction of outcome in
patients with memory impairment. Neurology 1996; 46:149–154
21. Devanand DP, Folz M, Gorlyn M, et al: Questionable dementia:
clinical course and predictors of outcome. J Am Geriatr Soc 1997;
45:321–328
22. Daly E, Zaitchik D, Copeland M, et al: Predicting conversion to
Alzheimer’s disease using standardized clinical information. Arch
Neurol 2000; 57:675–680
23. Bowen J, Teri L, Kukull W, et al: Progression to dementia in
patients with isolated memory loss. Lancet 1997; 349:763–765
24. Flicker C, Ferris SH, Reisberg B: Mild cognitive impairment in the
elderly: predictors of dementia. Neurology 1991; 41:1006–1009
25. Folstein MF, Folstein SE, McHugh PR : Mini-Mental State. A prac-
tical method for grading the cognitive state of patients for the
clinician. J Psychiatr Res 1975; 12:189–198
26. Anthony JC, LeResche L, Niaz U, et al: Limits of the ‘Mini-Mental
State’ as a screening test for dementia and delirium among hos-
pital patients. Psychol Med 1982; 12:397–408
27. Dick JP, Guiloff EJ, Stewart A, et al: Mini-Mental State Examination
in neurological patients. J Neurol Neurosurg Psychiatry 1984;
47:496–499
28. Tombaugh TN, McIntyre NJ: The Mini-Mental State Examination:
a comprehensive review. J Am Geriatr Soc 1992; 40:922–935
29. Morley JE, Tumosa N: Saint Louis University Mental Status Exam-
ination (SLUMS). Aging Successfully 2002; XII:4
30. Royall DR, Cordes JA, Polk M: CLOX: an executive clock drawing
test. J Neurol Neurosurg Psychiatry 1998; 64:588–594
31. Sadock BJ, Sadock VA: Kaplan and Sadock’s Comprehensive Text-
book of Psychiatry, 7th Ed. Philadelphia, Lippincott Williams &
Wilkins, 2000, pp. 899–901
32. Sadock BJ, Sadock VA: Kaplan and Sadock’s Comprehensive Text-
book of Psychiatry, 7th Ed. Philadelphia, Lippincott Williams &
Wilkins, 2000, p. 912
33. Banks WA, Morley JE: Memories are made of this: recent advances
in understanding cognitive impairments and dementia. J Gerontol
A Biol Sci Med Sci 2003; 58A:314–321
34. Kukull WA, Larson EB, Teri L, et al: The Mini-Mental State Exam-
ination score and the clinical diagnosis of dementia. J Clin Epi-
demiol 1994; 47:1061–1067
35. Fountoulakis KN, Tsoloki M, Mohs RC, et al: Epidemiological
Dementia Index: a multiculture screening instrument for Alzhei-
mer’s disease and other type of dementia suitable for use in
populations with low education level. Dement Geriatr Cogn Dis-
ord 1998; 9:329–338
36. Hanley JA, McNeil BJ: The meaning and use of the area under a
receiver operating characteristic (ROC) curve. Radiology 1982;
143:29–36
Comparison of the SLUMS and the MMSE
Am J Geriatr Psychiatry 14:11, November 2006910