The Alzheimer's Disease Assessment Scale-Cognitive subscale: normative data for older adult controls

Abstract

The Alzheimer's Disease Assessment Scale-Cognitive subscale (ADAS-cog) is commonly used to assess cognitive dysfunction in individuals with Alzheimer disease and other dementias. The purpose of this study was to provide normative scores for the ADAS-cog 11 individual items and total score, as well for delayed recall errors, using normal, elderly volunteers. The ADAS-cog was administered to 124, non-cognitively impaired volunteers ages 55 to 89, with 10 to 21 years of education. The mean total ADAS-cog score was five. The ADAS-cog error score was not associated with education in this highly educated group, and was positively correlated (P < 0.001) with the age of the participant. Age stratified ADAS-cog normative data are reported for the ADAS-cog total and the delayed recall error score.

Full text

ORIGINAL ARTICLE
The Alzheimer’s Disease Assessment
Scale - Cognitive Subscale
Normative Data for Older Adult Controls
David P. Graham, MD,*†
k
Jeffrey A. Cully, PhD,*†‡
k
A. Lynn Snow, PhD,*†‡
k
Paul Massman, PhD,{# and Rachelle Doody, MD, PhD§#
Abstract: The Alzheimer’s Disease Assessment Scale - Cognitive
subscale (ADAS-cog) is commonly used to assess cognitive dysfunc-
tion in individuals with Alzheimer disease and other dementias. The
purpose of this study was to provide normative scores for the ADAS-
cog 11 individual items and total score, as well for delayed recall
errors, using normal, elderly volunteers. The ADAS-cog was admin-
istered to 124, non-cognitively impaired volunteers ages 55 to 89,
with 10 to 21 years of education. The mean total ADAS-cog score
was five. The ADAS-cog error score was not associated with educa-
tion in this highly educated group, and was positively correlated (P ,
0.001) with the age of the participant. Age stratified ADAS-cog
normative data are reported for the ADAS-cog total and the delayed
recall error score.
Key Words: Alzheimer disease, assessment, dementia
(Alzheimer Dis Assoc Disord 2004;18:236–240)
T
he Alzheimer’s Disease Assessment Scale-cognitive sub-
scale (ADAS-cog) is commonly used to assess cognitive
dysfunction in individuals with Alzheimer disease.
1
The
ADAS-cog is a reliable and valid instrument
1,2
that has become
a standard outcome measure in multi-national Alzheimer dis-
ease treatment trials, and is widely used in studies as diverse as
caregiver burden,
3
functional status,
4
and staging of Alzheimer
disease.
5
As a testament to the widespread use of the ADAS-
cog, the reliability and validity of the ADAS-cog have been
confirmed by a number of studies in a variety of languages.
6–10
While not a substitute for extensive neuropsychological
assessment, the ADAS-cog is considered more complete than
most other cognitive screening measures.
11
The ADAS-cog scale (total score range from 0 to 70)
assesses performance on eleven cognitive tasks: orientation,
three trials of a 10-word list learning task (average score), three
trials of a 12 word recognition task (average score), recall of
instructions, comprehension of commands, object and finger
naming, word finding difficulty, expressive language, lan-
guage comprehension, ideational praxis, and constructional
praxis. A frequently used modification of the scale includes
a five minute delayed recall trial for the 10-word list.
Studies have shown that several sociodemographic
variables may influence ADAS-cog performance. Lower levels
of education may affect ADAS-cog scores.
7,12
Results for
highly educated individuals with dementia may overlap with
minimally educated, non-demented individuals,
13
particularly
those with less than seven years of schooling,
10
resulting in
a higher risk of false positives.
14
The ADAS-cog may be influ-
enced by age (younger age associated with modestly superior
performance),
15–17
and gender (male gender associated with
modestly superior performance).
18
Despite the frequency of research and clinical trials
using the ADAS-cog, only four studies reference ADAS-cog
normative data for Alzheimer disease and mild cognitive
impairment (MCI) patient populations,
19,20
and elderly
controls.
15,21
Additionally, only two studies describe individual
item performance in normal subjects.
20,21
Schmeidler’s study
reported values for various severity strata of Alzheimer disease
patients, but not for normal controls. Grundman’s study
reported numerical values for only four of the ADAS-cog
items. Although Zec et al presented ADAS-cog total error data
for stratified ages 7 to 80+,
15
the study did not provide specific
normative data (e.g. means and standard deviations for the
eleven individual tasks).
21
A recent study of MCI patients
indicated that the ADAS-cog may be a useful measure in
treatment models for patients with this condition,
22
however,
adequate normative data is needed to separate normal control
subjects from those with MCI.
The purpose of the present study was to evaluate the
ADAS-cog, including delayed recall, controlling for the effects
of age, gender, and educational level in normal controls. We
offer an age-stratified normative reference base for score
interpretation and comparison of the ADAS-cog for normal
controls.
METHODS
Subjects: This study used data from 124 non-demented,
non-MCI, non-depressed, older adult volunteers recruited to
a prospective longitudinal study at the Baylor College of
Medicine (BCM) Alzheimer’s Disease Center (ADC). The
Received for publication April 21, 2004; accepted September 30, 2004.
From the *Veterans Affairs South Central Mental Illness Research, Education,
and Clinical Center (MIRECC), †Psychiatry and Behavioral Sciences,
‡Medicine, and §Neurology Departments, Baylor College of Medicine,
k
Houston Center for Quality of Care and Utilization Studies, Health
Services Research & Development, Michael E. Debakey Veterans Affairs
Medical Center, {Department of Psychology, University of Houston,
#Baylor College of Medicine Alzheimer’s Disease Center.
Reprints: David P. Graham, MD, Michael E. DeBakey Veterans Affairs
Medical Center, 2002 Holcombe Blvd. (152), Houston, Texas 77030
(e-mail: david.graham@med.va.gov).
Copyright Ó 2004 by Lippincott Williams & Wilkins
236 Alzheimer Dis Assoc Disord

Volume 18, Number 4, October–December 2004
JOBNAME: alz 18#4 2004 PAGE: 1 OUTPUT: Mon November 29 12:09:45 2004
lww/alz/88441/04-032

normal controls were recruited, after BCM IRB approval, from
a registry of normal senior volunteers or from community
volunteers.
Initial evaluations included a clinical interview and
assessment of cognitive abilities, including the ADAS-cog and
the Mini-Mental State Examination.
23
The authors also
assessed delayed recall of the ADAS-cog 10-item word list
after a delay of five minutes as part of the memory testing.
Additionally, a neurologic examination, a health-screening
questionnaire, and an extensive neuropsychological assess-
ment battery were included in the initial evaluation to confirm
‘‘normal control’’status. All subjects were required to perform
in the normal range on neuropsychological testing which was
objectified as not scoring below the 5th percentile on more
than two measures. All subjects scored 26 or better on the
MMSE.
Procedures
An independent sample t test was used to evaluate
whether ADAS-cog scores differed between men and women.
Pearson correlations were used to evaluate whether age, years
of education, and MMSE scores were related to the ADAS-cog
total score and each of 11 individual ADAS-cog item scores
and the delayed recall error score. An ANOVA with Games-
Howell procedure for multiple comparisons with unequal
variance and/or unequal group sample sizes was used to
examine the effects of age cohorts on the ADAS-cog total
score. The age cohorts constructed included: 55 to 64, 65 to
69, 70 to 74, 75 to 79, and 80 to 89 years of age. The Kruskal-
Wallis test was used to examine the relationship between the
ADAS-cog and education categorized into three groups: high
school, college, and post-Baccalaureate. Bonferonni adjust-
ments were used to prevent capitalization on chance (ADAS-
cog evaluation 48 tests, a = 0.001; covariate analysis 6 tests,
a = 0.008, age-stratified analysis 5 tests, a = 0.01). Statistical
analyses were performed using SPSS for Windows version
11.5.0.
24
To evaluate the limitations of our single-ethnicity
sample, post-hoc comparisons of the mean ADAS-cog total
score and years of education were made between our results,
and the results of Liu et al
10
and Grundman et al
20
using a two
sample independent t test with Satterthwaite’s approximation
for unequal variances using Stata version 8.2.
25
Liu et al’s
study was comprised of one ethnicity (Chinese), while
Grundman et al’s control group was 71% Caucasian, 14%
African American, 10% Hispanic, 3% Asian American, 1%
American Indian, and 1% ‘‘Other.’’
RESULTS
The mean age of the124 participants in the study was
71.2 years (SD = 5.89, range 55 to 89). There were 65 men
(52.4%) and 59 women (47.6%), and all were Caucasian. They
had a mean of 15.6 total years of education (SD = 2.21).
Fourteen (11.3%) had some high school education (10 to 12
years), 72 (58.1%) had some college education (13 to 16
years), and 38 (30.6%) had some post-Baccalaureate education
(16 to 21years). ADAS-cog and MMSE scores for the sample
of 124 participants are summarized in Table 1. The majority of
errors on the ADAS-cog were made on word recall and word
recognition tasks and the median error score for the nine other
items was zero, with a range of zero to two. Men and women
did not differ on the total ADAS-cog score (t = 1.32, two-tailed
P $ 0.19, 95% CI = -0.27 to 1.33). ADAS-cog scores were not
significantly correlated with years of education (r = 0.04, P .
0.65). Results were the same when education was examined
categorically; neither having some college education or some
post-baccalaureate education affected the ADAS-cog score
compared with some high-school education (x
2
(2) = 0.089,
TABLE 1. Normative Scores for the ADAS-cog Individual Items and Total Score,
Delayed Recall Item Score, and MMSE Score
Median Mode Mean S.D. 95% C.I. Range
ADAS-cog total error score 5 5 4.98 2.25 4.58 to 5.38 1–11
Memory Components
Word Recall 3 3 3.19 1.33 3.19 to 3.42 1–6
Word Recognition 1 1 1.17 1.17 .97 to 1.38 0–7
Remembering Instructions 0000 0to0 0
Orientation 0 0 .10 .31 .05 to .16 0–1
Language Components
Naming (fingers & objects) 0 0 .05 .22 .01 to .09 0–1
Word Finding Difficulty 0 0 .01 .09 2.01 to .02 0–1
Follow oral commands 0 0 .11 .34 .05 to .17 0–2
Expressive Language 0 0 .02 .15 .00 to .05 0–1
Comprehension 0 0 .03 .22 2.01 to .07 0–2
Praxis Components
Constructional Praxis 0 0 .30 .49 .21 to .39 0–2
Ideational Praxis 0 0 .00 .00 0 to 0 0
Delayed Recall error score 4 5 3.82 2.12 3.45 to 4.20 0–10
MMSE total score 29 30 29.29 .85 29.14 to 29.44 26–30
ADAS-cog, Alzheimer’s Disease Assessment Scale - Cognitive subscale; MMSE, Mini-Mental State Examination.
q 2004 Lippincott Williams & Wilkins 237
Alzheimer Dis Assoc Disord

Volume 18, Number 4, October–December 2004 The Alzheimer’s Disease Assessment Scale
JOBNAME: alz 18#4 2004 PAGE: 2 OUTPUT: Mon November 29 12:09:46 2004
lww/alz/88441/04-032

P . 0.95). Years of education were not significantly correlated
with either age (all age strata had between a mean of 15.27 to
16.00 years of education) or the MMSE total score.
ADAS-cog scores were correlated significantly with age
(r = 0.34, P , 0.001). When divided into age strata, the effect
remained (one-way ANOVA (d.f. 4, 123), F = 3.767, P =
0.006). The results summary for the ADAS-cog total score by
age strata is presented in Table 2. Age was also positively
correlated with the individual ADAS-cog item of word recall
(r = 0.30, P = 0.001), but not with any other ADAS-cog items.
A trend toward statistical significance (P = 0.002, just above
Bonferroni correction of P = 0.001) was exhibited by two
ADAS-cog items: word finding (r = 0.27) and expressive lan-
guage (r = 0.28); as well as the additional task of delayed recall
errors (r = 0.28, P = 0.002).
The ADAS-cog error scores were inversely, but non-
significantly, related to the MMSE scores (Pearson r = 20.13,
P = 0.157), and were positively correlated to the individual
items: word recall (r = 0.75, P , 0.001), word recognition (r =
0.68, P , 0.001), constructional praxis (r = 0.31, P , 0.001),
follow commands (r = 0.31, P , 0.001), and the naming score
(r = 0.31, P , 0.001). The total ADAS-cog error score was
positively correlated with the delayed recall error score (r =
0.56, P , 0.001). Age, years of education, and the total MMSE
score were not significantly correlated (r values all , 0.12, all
P values . 0.20).
DISCUSSION
The main purpose of this study is to report preliminary
age-stratified normative control data for the Alzheimer Disease
Assessment Scale - Cognitive subscale total score and scores
for each of eleven individual components as well as for
delayed recall. Although controls, on average, score about five
errors on the ADAS-cog, those under age 70 are more likely to
score four errors, while those 70 to 79 will score a median of
five errors, and those 80 and over normally score a median of
eight errors. With respect to delayed recall errors, the data
show those under age 65 scored a median of two errors, while
those over age 65 typically scored a median of four errors. Our
average delayed recall error score was probably higher than
that of Grundman et al
20
because his normal control subjects
were selected, in part, on the basis of paragraph recall
performance, and we had no such a priori restriction. These
data are important because the ADAS-cog has become
a standard measure in clinical trials involving Alzheimer
disease patients,
26
and may become a standard in MCI trials,
where separation of normal from MCI subjects is critical.
Our study found an influence of age on the ADAS-
cog total error scores, supporting the views of Zec et al,
15
Doraiswamy et al,
16
and Pena-Casanova et al,
17
, particularly
over age 80. This positive correlation was seen whether age
was viewed as a continuous variable or grouped into cohorts.
The age of the patients in our study was similar to that of Liu
et al and Grudman et al (mean of 72.7 years versus 71.2 versus
70.0 years) and gender composition was also similar (52%
male versus 52% male versus 40% male). Our study supports
reports that gender has no apparent effect on the total ADAS-
cog score, contrasting with Doraiswamy et al.
18
Although the mean ADAS-cog total scores were not
statistically different, the education in our populations did
differ significantly; our study population had a mean of 15.6
years of education, Liu et al had a mean of 9.9 years, and
Grundman et al had a mean education of 14.8 years. Liu et al
10
found that there may be an influence of very low education
levels (zero to six years). Since we found no influence of
ten years or more of education, it could be that a threshold
level of education between seven and ten years is required to
TABLE 2. Descriptive Statistics for Total ADAS-cog Error Score, Delayed Recall Error Score, and Years of Education
by Age Cohort
Median Mode Mean S.D. Z-score 95% C.I. Range
n = 124 ADAS-cog total score 5 5 4.98 2.25 4.58 to 5.38 1–11
n=15 Ages 55 to 64 4 4 3.60 1.64 2.61 2.69 to 4.51 1–6
n=30 Ages 65 to 69 4 4 4.51 2.18 2.21 3.70 to 5.32 1–10
n=48 Ages 70 to 74 5 5 5.13 2.06 .07 4.53 to 5.72 1–9
n=20 Ages 75 to 79 5 5 5.51 2.11 .24 4.53 to 6.50 2–11
n=11 Ages 80 to 89 8 8 6.57 3.02 .71 4.55 to 8.60 2–11
n = 124 Delayed recall error score 4 5 3.82 2.12 3.45 to 4.20 0–10
n=15 Ages 55 to 64 2 1 2.73 2.19 2.51 1.52 to 3.94 0–7
n=30 Ages 65 to 69 3.5 5 3.33 1.94 2.23 2.61 to 4.06 0–6
n=48 Ages 70 to 74 4 5 3.92 1.97 .05 3.35 to 4.49 0–9
n=20 Ages 75 to 79 4.5 4 & 5 4.60 2.19 .37 3.58 to 5.62 1–10
n=11 Ages 80 to 89 4 4 4.82 2.40 .47 3.21 to 6.43 1–9
n = 124 Years of Education 16 16 15.56 2.21 15.17 to 15.96 10–21
n=15 Ages 55 to 64 15 14 15.27 2.55 13.86 to 16.68 10–20
n=30 Ages 65 to 69 16 16 15.60 1.94 14.88 to 16.32 12–20
n=48 Ages 70 to 74 16 16 15.65 2.05 15.05 to 16.24 12–21
n=20 Ages 75 to 79 15.5 14 & 16 15.30 2.76 14.01 to 16.59 11–20
n=11 Ages 80 to 89 16 16 16.00 2.32 14.44 to 17.56 12–20
ADAS-cog, Alzheimer’s Disease Assessment Scale - Cognitive subscale.
238 q 2004 Lippincott Williams & Wilkins
Graham et al Alzheimer Dis Assoc Disord

Volume 18, Number 4, October–December 2004
JOBNAME: alz 18#4 2004 PAGE: 3 OUTPUT: Mon November 29 12:09:48 2004
lww/alz/88441/04-032

successfully evaluate ADAS-cog performance. This would be
consistent with the conclusions from Doraiswamy et al,
12
who
stated that worse performance was seen in subjects with less
than a high school education; as well as with Zec et al,
15
who
stated that higher ADAS-cog error scores were seen in subjects
ages seven to thirteen (who generally should have less than
eight years of education). Our data cannot be used to deter-
mine whether low education has a negative effect, or whether
meeting a particular threshold of education lends a positive
effect.
We found a mean total ADAS-cog value of 5.0, identical
to the mean of 5.0 found by Liu et al,
10
and very similar to the
mean of 5.5 found by Zec et al,
15
and 5.6 by Grundman et al.
20
All of these scores are lower than the mean scores reported for
demented individuals.
10,16,19,20
Further, it is important to note
the mean ADAS-cog total scores are not statistically different
despite the different ethnicities comprising the studies. This
study was comprised of no minorities, Liu et al
10
evaluated
a 100% Chinese population, and Grundman et al
20
used a popu-
lation in which 20% of the control sample was of minority
status. These similarities in mean scores across samples and
ethnicities suggests ethnicity may not be an important deter-
minant of ADAS-cog scores. Caution should be applied to
comparisons between studies regarding different ethnicities, as
subtle differences in testing procedures may have an un-
recognized affect on ADAS-cog total or component scores. An
example is the study by Liu et al,
10
who attempted to accom-
modate illiteracy in subjects by including pictorial versus
verbal stimuli in their testing methodology.
In summary, our study reports new preliminary data on
expected performance of normal controls on the ADAS-cog,
and gives actual expected age-adjusted norms. Several limita-
tions affect generalizability: First, our population consisted of
one ethnicity, Caucasian, but our mean scores do not differ
from studies performed with other ethnicities. Second, our
population consisted of a highly educated group, restricting us
from examining the possible influences of very low education
on the ADAS-cog. Third, our sample size was not matched
across age stratification. Despite these limitations, and given
the widespread use of the ADAS-cog, the reported preliminary
normative data generated by this study provide a useful refer-
ence for clinicians and researchers, especially those attempting
to differentiate between normal and atypical cognitive func-
tioning in older adults.
More normative studies are clearly needed to enhance
our understanding of the interactions among the various points
including age, education and ethnicity. Furthermore, we recom-
mend future studies seriously consider the inclusion of larger
sample sizes, both age and education stratification, better repre-
sentation of minorities, and better representation of low edu-
cational levels in their design.
ACKNOWLEDGMENTS
This work was supported by NIH AGO P50 8664 (P.M.
and R.D.) as well as by a Zenith Award from the Alzheimer’s
Association (R.D.). Additionally, this material is based in part
upon work supported by the Office of Academic Affiliations, VA
Special MIRECC Fellowship Program in Advanced Psychiatry
and Psychology, Department of Veteran Affairs (D.G.), and the
Health Services Research and Development Service, Office of
Research and Development, Department of Veterans Affairs
(RCD 01-008-1).
REFERENCES
1. Rosen WG, Mohs RC, Davis KL. A new rating scale for Alzheimer’s
disease. Am J Psychiatry. 1984;141:1356–1364.
2. Stern RG, Mohs RC, Davidson M, et al. A longitudinal study of
Alzheimer’s disease: measurement, rate, and predictors of cognitive dete-
rioration. Am J Psychiatry. 1994;151:390–396.
3. Caro J, Ward A, Ishak K, et al. To what degree does cognitive impairment
in Alzheimer’s disease predict dependence of patients on caregivers? BMC
Neurol. 2002;2:6.
4. Farias ST, Harrell E, Neumann C, et al. The relationship between
neuropsychological performance and daily functioning in individuals with
Alzheimer’s disease: ecological validity of neuropsychological tests. Arch
Clin Neuropsychol. 2003;18:655–672.
5. Hannesdottir K, Snaedal J. A study of the Alzheimer’s Disease Assess-
ment Scale-Cognitive (ADAS-Cog) in an Icelandic elderly population.
Nord J Psychiatry. 2002;56:201–206.
6. Tsolaki M, Fountoulakis K, Nakopoulou E, et al. Alzheimer’s disease
assessment scale: the validation of the scale in Greece in elderly demented
patients and normal subjects. Dement Geriatr Cogn Disord. 1997;8:273–
280.
7. Weyer G, Erzigkeit H, Kanowski S, et al. Alzheimer’s disease assess-
ment scale: Reliability and validity in a multicenter clinical trial. Int
Psychogeriatr. 1997;9:123–138.
8. Inzitari D, Rossi R, Lamassa M, et al. Validity of different linguistic
versions of the Alzheimer’s disease assessment scale in an international
multicentre Alzheimer’s disease trial. Dement Geriatr Cogn Disord. 1999;
10:269–277.
9. Chu LW, Chiu KC, Hui SL, et al. The reliability and validity of the
Alzheimer’s disease assessment scale cognitive subscale (ADAS-cog)
among the elderly Chinese in Hong Kong. Ann Acad Med Singapore.
2000;29:474–485.
10. Liu HC, Teng EL, Chuang YY, et al. The Alzheimer’s disease assessment
scale: findings from a low-education population. Dement Geriatr Cogn
Disord. 2002;13:21–26.
11. Pena-Casanova J, Aguilar M, Santacruz P, et al. Adaptation and
normalization of the Alzheimer’s disease Assessment Scale for Spain
(NORMACODEM) (II). Neurologia. 1997;12:69–77.
12. Doraiswamy PM, Krishen A, Stallone F, et al. Cognitive performance of
the Alzheimer’s disease assessment scale: effect of education. Neurology.
1995;25:1980–1984.
13. Schultz RR, Siviero MO, Bertolucci PH. The cognitive subscale of the
‘‘Alzheimer’s disease assessment scale’’ in a Brazilian sample. Braz J Med
Biol Res. 2001;34:1295–1302.
14. Kolibas E, Korinkova V, Novotny V, et al. ADAS-cog (Alzheimer’s
Disease Assessment Scale - cognitive subscale) - validation of the Slovak
version. Bratisl Lek Listy. 2000;101:598–602.
15. Zec RF, Landreth ES, Vicari SK, et al. Alzheimer disease assessment
scale: useful for both early detection and staging of dementia of the
Alzheimer type. Alzheimer Dis Assoc Disord. 1992a;6:89–102.
16. Doraiswamy PM, Bieber F, Kaiser L, et al. The Alzheimer’s Disease
Assessment Scale: patterns and predictors of baseline cognitive perfor-
mance in multicenter Alzheimer’s disease trials. Neurology. 1997a;48:
1511–1517.
17. Pena-Casanova J. Alzheimer’s Disease Assessment Scale–cognitive in
clinical practice. Int Psychogeriatr. 1997;9:105–114.
18. Doraiswamy PM, Bieber F, Kaiser L, et al. Memory, language, and praxis
in Alzheimer’s disease: norms for outpatient clinical trial populations.
Psychopharmacol Bull. 1997b;33:123–128.
19. Schmeidler R, Mohs RC, Aryan M. Relationship of disease severity to
decline on specific cognitive and functional measures in Alzheimer
disease. Alzheimer Dis Assoc Disord. 1998;12:146–151.
20. Grundman M, Petersen RC, Ferris SH, et al. Mild cognitive impairment
can be distinguished from Alzheimer disease and normal aging for clinical
trials. Arch Neurol. 2004;61:59–66.
q 2004 Lippincott Williams & Wilkins 239
Alzheimer Dis Assoc Disord

Volume 18, Number 4, October–December 2004 The Alzheimer’s Disease Assessment Scale
JOBNAME: alz 18#4 2004 PAGE: 4 OUTPUT: Mon November 29 12:09:48 2004
lww/alz/88441/04-032

21. Zec RF, Landreth ES, Vicari SK, et al. Alzheimer Disease Assessment
Scale: a subtest analysis. Alzheimer Dis Assoc Disord. 1992b;6:164–181.
22. Salloway SP, Kumar D, Ieni J, et al. Benefits of Donepezil treatment in pa-
tients with Mild Cognitive Impairment. Neurology. 2003;60(Suppl 1):A411.
23. Folstein MF, Folstein SE, McHugh PR. Mini-Mental State Examination.
J Psychiatr Res. 1975;2:189–198.
24. SPSS for Windows version 11.5.0. SPSS Inc. Headquarters, 233 S.
Wacker Drive, 11th floor Chicago, Illinois 60606. Copyright 1989 - 2002.
http://www.spss.com.
25. Stata version 8.2. StataCorp, 4905 Lakeway Drive, College Station, Texas
77845 USA. Copyright 1984 – 2003: http://www.stata.com.
26. Gillen TE, Gregg KM, Yuan H, et al. Clinical trials in Alzheimer’s disease.
Calculating Alzheimer’s Disease Assessment Scale-cognitive subsection
with the data from the consortium to establish a registry for Alzheimer’s
disease. Psychopharmacol Bull. 2001;35:83–96.
27. Weyer G, Ihl R, Schambach M. Alzheimer’s disease assessment scale
(ADAS). Deutschsprachige Bearbeitung. Dokumentationsbogen und
Testmaterial. Weinheim, Beltz, 1992.
240 q 2004 Lippincott Williams & Wilkins
Graham et al Alzheimer Dis Assoc Disord

Volume 18, Number 4, October–December 2004
JOBNAME: alz 18#4 2004 PAGE: 5 OUTPUT: Mon November 29 12:09:49 2004
lww/alz/88441/04-032