The Addenbrooke’s Cognitive Examination

(ACE) in the Differential Diagnosis of Early

Dementias Versus Affective Disorder
Robert B. Dudas, M.D., M.Phil.
German E. Berrios, M.D., D.M., FRCPsych
John R. Hodges, M.D., FRCP

Objective: The authors describe the profile of performance of patients whose cognitive
complaint is due to dementia, affective disorder, or combinations thereof on the Ad-
denbrooke’s Cognitive Examination (ACE) test battery. Methods: Authors tested 90
subjects with dementia (63 Alzheimer disease [AD]; 27 fronto-temporal dementia
[FTD]), 60 subjects with “pure” affective disorder (23 major depression [MDD], 37
whose affective symptoms did not meet criteria for major depression [Affective]); 22
patients with symptoms of affective disorder and organic dementia (Mixed); and 127
healthy volunteers (NC). Results: The total ACE scores for the AD, FTD, and Mixed
groups were significantly lower than for the NC group. Likewise, on total score, the
AD and FTD groups scored significantly lower than either of the “pure” affective-dis-
order groups. Within the dementia group, the AD group scored significantly lower
than the fronto-temporal group. Conclusions: The profile of performance on the ACE
of patients with dementia is different from that of patients suffering from affective
illness. Mild impairment in the total ACE score, along with a low score on the memory
domain tasks and letter fluency (in contrast to normal category fluency), are strongly
indicative of an affective, as opposed to organic, pathology. A total score of ⬍88 in
suspected dementia patients with affective symptoms appears strongly predictive of
an underlying organic disorder. (Am J Geriatr Psychiatry 2005; 13:218–226)

T he Addenbrooke’s Cognitive Examination is a

brief, 15–20-minute test battery originally de-
signed to detect and classify different kinds of de-
mentia, particularly Alzheimer disease (AD) and
fronto-temporal dementia (FTD), without the use of
specialized test equipment.1 It incorporates the Mini-
Mental State Exam (MMSE),2 with additional mem-
ory, language, and visuospatial components. The
maximum score is 100, weighed as follows: orienta-
tion (10), attention (8), memory (35), verbal fluency
(14), language (28), and visuospatial ability (5). The
ACE was validated on 115 dementia (AD, FTD, and
vascular dementia), and 24 non-dementia patients,
and 127 NC subjects. Of note is the fact that patients

Received March 14, 2003; revised September 15, 2003; July 26, 2004; accepted July 31, 2004. From the Department of Psychiatry, University of
Cambridge (RBD,GEB), and the Department of Neurology, University of Cambridge and the MRC Cognition and Brain Sciences Unit (JRH). Send
correspondence and reprint requests to Robert B. Dudas, M.D., Department of Psychiatry, University of Cambridge, Addenbrooke’s Hospital (Box
189), Hills Road, Cambridge, UK. e-mail: rbd21@cam.ac.uk
䉷 2005 American Association for Geriatric Psychiatry

218 Am J Geriatr Psychiatry 13:3, March 2005


with affective pathology were excluded from this
study. When the lower cut-off score of 83 was used,
the sensitivity of the ACE to detect dementia was
79%, one-third more than the MMSE using the con-
ventional cut-off score of 24. The ACE appeared to be
especially sensitive in FTD, where it nearly doubled
the rate of detection, compared with the MMSE. At a
cut-off score of 88, in a memory clinic with a dementia
prevalence rate of 40%, a high proportion (68%) of
those who screen positive will have dementia, and
94% of those above the cut-off score will not. Age,
number of years of education, or gender did not in-
fluence predictive outcome during the initial valida-
tion.
Patients presenting with combinations of cognitive,
affective, and behavioral problems pose a clinical co-
nundrum. In some cases, it is difficult to establish
whether the cognitive impairment is secondary to an
affective disorder, or a more sinister, organic de-
menting process—most often AD. Depressive symp-
toms are common in the elderly population without
dementia.3,4 A small proportion of depressed individ-
uals present with very significant cognitive under-
functioning, formerly known as depressive pseudo-
dementia, also termed “functional (as opposed to
organic) dementia,” “memory disorder in the context
of depressive illness,” or “the dementia syndrome of
depression.” In clinical practice, this group remains
difficult to identify. Approximately 10% of the pa-
tients with depression are misdiagnosed as having
dementia,5,6 leading to a misdiagnosis of depression
rather than early dementia. One of the major con-
founding factors is that many patients with organic
dementia also have affective symptoms. Patients with
AD often contact their general practitioner or the
memory clinic with complaints of depressive symp-
toms.
Several screening and diagnostic tests for dementia
(such as the CAMDEX7 or the Mattis Dementia Rat-
ing Scale8) are available, but their complexity and
need for specialized test equipment put them beyond
routine bedside use.9
The aim of this study was to investigate the ability
of the ACE to discriminate between cognitive defi-
cits resulting from dementia versus those from af-
fective disorder. Correlations were sought between
the cognitive deficits (reflected by the cognitive do-
main scores of ACE) and affective symptoms (as
captured by the Hamilton Rating Scale for Depres-
Am J Geriatr Psychiatry 13:3, March 2005
Dudas et al.
sion [Ham-D] and other standard instruments for
affective change). We were particularly interested in
the ability of the ACE to detect progressive degen-
erative dementias in patients presenting with diffi-
cult, mixed symptoms.
METHODS
Subjects
A total of 299 patients participated: major depres-
sive disorder (MDD; N⳱23), Affective (N⳱37), AD
(N⳱63), FTD (N⳱27), mixed: possible progressive
degenerative dementing disorder with an affective
component (N⳱22), and healthy volunteers (NC;
N⳱127). The research protocol was approved by the
Local Research Ethics Committee. Data for the pa-
tients with dementia were collected from the Cam-
bridge Memory Clinic, and for those with MDD from
the Medical Research Council Link Study between
June 1996 and April 2001. Patients were excluded if
they had 1) any concurrent psychiatric illness other
than those on the affective spectrum (excluding bi-
polar disorder according to ICD-10), such as schizo-
phrenia, or obsessive-compulsive disorder, or 2)
causes of cognitive impairment other than primary
degenerative or affective pathology (e.g., vascular de-
mentia, closed head injury, or alcoholism). The
healthy volunteers for the NC group were ascer-
tained from a previous study.1 Exclusion criteria were
history of head injury, drug abuse, alcoholism, and
cognitive complaints of neurological or psychiatric
origin.
The diagnosis of AD was in accordance with the
National Institute of Neurological and Communica-
tive Disorders and Stroke–AD and Related Disorders
Association (NINCDS-ADA) criteria.10 The diagnosis
of FTD was used in all cases with focal lobar degen-
eration, including patients with the core feature of
personality and behavioral changes (frontal-variant
FTD), non-fluent spontaneous speech (progressive
non-fluent aphasia), and fluent empty spontaneous
speech with semantic breakdown (semantic demen-
tia).11,12 The severity of dementia was assessed with
the Clinical Dementia Rating scale.13,14 Patients with
MDD met the DSM-IV criteria for major depression.
Patients in the affective group (Affective) had signifi-
cant memory complaints and depressive symptom-
219
Dementia Versus Depression
atology (e.g., low mood, anhedonia, lack of motiva-
tion, lack of energy, etc., alone or in combination) of
various degrees, but did not meet the DSM-IV criteria
for MDD. Importantly, they were also not felt to suffer
from a progressive degenerative dementing condi-
tion. In contrast, the Mixed group consisted of pa-
tients with a mixture of cognitive and affective symp-
toms; and, in this group, it was difficult to arrive at a
formal diagnosis with high certainty and to establish
whether or not the cognitive impairment was due to
a degenerative disease, and, if so, to what extent. A
large proportion of these patients developed frank
progressive degenerative dementia over an average
follow-up period of 2 years. The NC consisted of 127
age- and education-matched neuropsychiatrically
healthy volunteers who were attendees at an ortho-
pedic (N⳱36) or gynecologic (N⳱28) clinic, spouses
of the before-mentioned (N⳱26), or members of the
Medical Research Council subject panel (N⳱37).
All patients were assessed by a senior neurologist
(JRH) and psychiatrist (GEB), using standardized as-
sessments described elsewhere.15 In brief, a care-
giver/relative is also always interviewed. Patients
undergo a neuropsychological evaluation, including
the Wechsler Memory Scale–Revised,16 the Rey Com-
plex Figure Test,17 the Rey Auditory Verbal Learning,
Wechsler Adult Intelligence Scale (WAIS),18 the Trail-
Making Test,19 Stroop Test, Verbal Fluency,20 Wiscon-
sin Card-Sorting Test, Graded Naming Test,21 Na-
tional Adult Reading Test,22 Visual Object and Space
Perception battery,23 and the Warrington Recognition
Memory Test.24 The team met at the end of the day,
and reviewed evidence from all clinical and test re-
sults. All patients have structural brain scans (CT or
MRI), and are followed up in the clinic after 6 months.
Those with progressive degenerative dementias or
ambiguous diagnoses are followed thereafter at 6–12-
month intervals.
Procedures
The ACE was completed in the memory clinic as
part of the routine assessment. The orientation and
attention components of the ACE are identical to the
ones in the MMSE. The memory component evalu-
ates episodic and semantic memory. In addition to the
recall of three items from the MMSE, there is a “name
and address learning and delayed recall test,” which
appears to be a remarkably sensitive measure in the
220
detection of early AD. The language component com-
prises naming 12 line-drawings of medium and low
familiarity, comprehension of sentences, repeating
words and phrases, reading regular and irregular
words, and writing a sentence. Naming has been
found to be dependent upon concept frequency and
age of acquisition. Alterations made to the non-
semantic aspects of language evaluation in the ACE
are intended for the early detection of FTD, especially
the progressive aphasic variants. Frontal executive
function is tested by verbal fluency in two tasks: letter
fluency (generating words beginning with the letter
P in 1 minute), and category fluency (generating
names of animals in 1 minute). Letter fluency relies
upon phonologic processing, and category fluency on
semantic memory in addition to other executive pro-
cesses. The former is a particularly sensitive indicator
of FTD, whereas the latter is a marker of semantic
memory impairment. Visuospatial testing includes
copying overlapping pentagons (from the MMSE)
and a wire cube, and drawing a clock face. Adding
the three-dimensional wire cube copying and the
clock face-drawing test provide a greater scope for
detecting impaired constructional abilities. Raw
scores are used for each item except the two fluency
tasks, where scaled scoring systems derived from the
Gaussian distributions of the raw scores from NC
subjects are used instead. The calculated six domain
scores add up to the composite score on the ACE.
We chose the 17-question version of the Hamilton
Rating Scale for Depression (Ham-D) as an observer-
rated instrument to assess the actual affective status
of our patients.25
The statistical analysis was carried out with the
SPSS for Windows package. One-way ANOVAs and
suitable post-hoc multiple comparisons (Scheffé test
when equal variances were assumed and Games-
Howell test when the Levene test was significant)
were carried out to compare the relevant group
means. As in our original sample, ascertained
through the natural occurrence and presentation of
patients in the memory clinic, the groups presenting
with diagnoses other than a progressive degenerative
dementia tended to be younger than those with a pro-
gressive degenerative process, we carried out analy-
ses of covariance for age with confidence-interval ad-
justment, using the Bonferroni method for each
variable. As a check, but not reported, the appropri-
ate nonparametric tests were also carried out, and
Am J Geriatr Psychiatry 13:3, March 2005

they were consistent in each case. Logistic-regression
analysis was used to predict group membership, and
we applied a receiver operating characteristic curve
(ROC) analysis to examine the sensitivity and speci-
ficity of our measures.
RESULTS
Clinical and demographic characteristics of the pa-
tient and NC groups are summarized in Table 1.
Comparison of the age of the subject groups revealed
an overall significant difference (one-way ANOVA:
F[5, 291]⳱14.31; p⳱0.0000). The AD patients, although
not significantly different from the subjects in the NC
and patients in the Mixed group, were older than
those in the other groups, and, at the other extreme,
the Affective group was significantly younger than
all the other groups except the MDD and the FTD.
The NC group matched each patient group for age,
except the Affective group who were younger. There
were no significant differences in education level be-
tween the groups (one-way ANOVA: F[5, 242]⳱1.79;
p⳱0.115).
The MMSE scores reflected the fact that the AD
group (mean score: 20.4; standard deviation [SD]: 5.6)
consisted of mild cases, and patients in the FTD
group had very mild overall cognitive impairment
(mean score: 26.0; SD: 3.7). The Clinical Dementia
Rating Scale (CDR)26 scores, likewise, indicated rela-
tively mild disease.
As expected from the entry criteria, the MDD
group scored high on the Ham-D (mean: 24.7; [SD:
4.7]), and had significantly more signs and symptoms
of depression than any other group (one-way AN-
OVA: F[4, 133]⳱134.28; p⳱0.0000). The Affective and
the Mixed groups (Ham-D scores: 10; [SD: 4.4] versus
TABLE 1. Clinical and Demographic Characteristics of the Patient and Control Groups
Diagnostic Group Age, years Sex (M/F)
Major depression (MDD) (N⳱23) 59.1 (7.9) 11/12
Affective (N⳱37) 54.4 (10.2)* 21/16
Alzheimer disease (AD) (N⳱63) 68.9 (8.0) 27/36
Frontotemporal dementia (FTD) (N⳱27) 61.3 (7.3) 18/9
Mixed (N⳱22) 64.1 (8.1) 12/10
Normal-control (NC) (N⳱127) 64.4 (9.4) 63/64
Note: Values are mean (standard deviation). MMSE: Mini-Mental State Exam; CDR: Clinical Dementia Rating Scale; Ham-D: Hamilton Rating
Scale for Depression.
*Scheffé: p⳱0.0000 relative to NC group.
Am J Geriatr Psychiatry 13:3, March 2005
Dudas et al.
9.4; [SD: 5.0]) had equivalent levels of depression,
and, although significantly less depressed than the
MDD group (Games-Howell p⳱0.0000), they were
still significantly more depressed than either of the
progressive-degenerative dementia groups (p⳱0.0000
and p⳱0.0003, respectively, for AD and FTD). The
Ham-D mean scores indicated minimal affective
symptoms in the AD (1.8; SD: 2.0) and FTD (2.8; [3.7])
groups. In an ANCOVA, age was not a significant
covariate of the Ham-D score.
ACE Scores
The mean scores on the ACE of the various groups
are shown in Table 2. An ANCOVA showed a signifi-
cant between-group difference (F[5, 290]⳱73.73; p⳱
0.0000). It is of note that there was a lack of correlation
between age and the ACE total score. To explore this
further, we carried out post-hoc pairwise compari-
sons. The AD, FTD, and Mixed groups showed sig-
nificant impairment relative to the NC group (Bon-
ferroni; p⳱0.0000), and, importantly, they were also
more impaired than any of the affective groups
(p⳱0.0000). Also, the AD group scored significantly
lower than the FTD group (p⳱0.0004). The Mixed
group was indistinguishable from either of the pro-
gressive-degenerative dementia groups. Compared
with the NC group, the small decrements in the over-
all performance of the MDD and Affective groups
failed to reach significance. In view of the relatively
low number of cases in the MDD and Affective
groups, we repeated the analysis with a single affec-
tive (MDD Ⳮ Affective) group. Again, an ANCOVA
demonstrated a significant difference between the
progressive-degenerative dementia groups and the
Mixed group versus the collapsed affective group,
but it could not distinguish the collapsed affective
group from the NC group.
Education, years MMSE CDR Ham-D
11.2 (1.7) 27.6 (2.4) 24.7 (4.7)
12.4 (3.1) 28.3 (2.5) 10.0 (4.4)
10.6 (2.3) 20.4 (5.6) 1.1 (0.6) 1.8 (2.0)
11.3 (2.7) 26.0 (3.7) 0.6 (0.3) 2.8 (3.7)
11.1 (2.9) 23.9 (4.7) 0.8 (0.3) 9.4 (5.0)
11.3 (2.6) 29.1 (0.9)
221
Dementia Versus Depression

To examine further the usefulness of the ACE total
score as a quick guide to assist the clinician in the
detection of progressive-degenerative dementias in a
patient population with cognitive deficits of various
etiologies, we carried out a logistic-regression anal-
ysis with only two target variables: progressive-de-
generative dementia patients (AD and FTD) versus
subjects with no progressive-degenerative dementia
(MDD, Affective, and NC). The patients originally
classified into the Mixed group for whom we had fol-
low-up diagnoses were allocated into the appropriate
group according to their follow-up diagnoses; that is,
those who developed a progressive-degenerative de-
mentia were put into the “progressive degenerative”
group; those who still suffered from an affective ill-
ness without significant cognitive deterioration, into
the “no organic dementia” group; and those with still
mixed affective Ⳮ cognitive symptomatology or no
follow-up information were excluded from this anal-
ysis. The total ACE score correctly classified 88.5% of
the cases. The trade-off between sensitivity (true posi-
tive rate) and 1–specificity (false positive rate) of the
ACE in diagnosing progressive dementia in a patient
population with and without a later confirmed pro-
Sensitivity
and NC subjects scoring less than 83 and 88 (the
recommended cut-off scores on the ACE for research
and for screening, respectively) are shown in Table
3. These proportions indicated a need for further
indices in the domain scores to enable us to better
separate the affective patients from those with
progressive-degenerative dementia, especially MDD/
Affective patients from FTD patients, on an individual
basis.
A series of ANCOVAs revealed significant mean
differences (p⳱0.0000) between the groups for all do-
main scores (Table 2). Age only seemed to correlate
FIGURE 1. Receiver Operating Characteristics (ROC) of the
Addenbrooke’s Cognitive Examination (ACE) as a
Test for Progressive Degenerative Dementia
ROC Curve
1.00
0.75

gressive-degenerative pathology is shown in the ROC 0.50

curve in Figure 1. The area under the ROC curve is
0.95, which suggests that the ACE has a high specific-
ity for a large range of sensitivities. At 88, the previ-
ously recommended cut-off score for clinical use in the 0.25
detection of dementia,1 the ACE showed a sensitivity
of 93%, and a specificity of 82% for progressive-
degenerative pathology in our study.
A scatterplot showing the individual scores of our
0.00 0.25 0.50 0.75 1.00
subjects on the ACE is displayed in Figure 2, and the 1-Specificity
proportions of MDD, Affective, AD, FTD, Mixed,

TABLE 2. Addenbrooke’s Cognitive Examination (ACE) Cognitive Domain and Total Scores
MDD Affective AD FTD Mixed NC ANCOVA: F[5]
Orientation (max: 10) 9.7 (0.6) 9.7 (0.8) 6.7 (2.6) 9.4 (0.9) 8.6 (1.7) 9.9 (0.4) 46.62
Attention (max: 8) 6.9 (1.4) 7.5 (1.4) 5.6 (2.2) 7.3 (1.2) 6.2 (1.9) 7.9 (0.4) 46.27
Memory (max: 35) 30.4 (5.0) 29.1 (5.9) 15.7 (8.3) 24.4 (7.9) 20.0 (8.7) 32.7 (2.0) 79.99
Verbal fluency (max: 14) 9.8 (2.7) 9.9 (3.5) 6.6 (3.1) 6.0 (3.4) 7.3 (3.1) 11.0 (2.2) 30.15
Naming (max: 12) 12.0 (0.0) 11.9 (0.3) 10.3 (2.6) 9.2 (3.8) 11.0 (1.8) 12.0 (0.3) 16.72
Language (max: 16) 15.4 (0.7) 15.1 (2.2) 14.1 (2.5) 13.8 (2.1) 14.7 (1.9) 15.9 (0.4) 13.64
Visuospatial (max: 5) 4.5 (0.5) 4.6 (0.9) 2.9 (1.8) 4.0 (1.3) 3.8 (1.5) 4.6 (0.6) 20.58
ACE Total (max: 100) 89.2 (9.2) 89.0 (9.2) 61.9 (18.3) 74.2 (15.4) 71.0 (16.6) 93.9 (3.5) 73.73*
Note: Values are mean (standard deviation). MDD: major depression; AD: Alzheimer disease; FTD: frontotemporal dementia; NC: normal-
control subjects.
*p⳱0.0000; AD, FTD, and Mixed groups showed significant impairment relative to the NC group (Bonferroni; p⳱0.0000), and were also
more impaired than any of the affective groups (p⳱0.0000). Also, the AD group scored significantly lower than the FTD group (p⳱0.0004).

222 Am J Geriatr Psychiatry 13:3, March 2005

 Dudas et al.

FIGURE 2. Scatterplot of the Addenbrooke’s Cognitive Examination (ACE) Total Scores in the Diagnostic Groups

100

75
ACE Score

50

25

AD FTD MD Affective Mixed NC

Diagnostic Group

Note: Reference line is a score of 88, the recommended cut-off score on the ACE for screening purposes.

with verbal fluency but none of the other domain
scores. The Bonferroni test was used for pairwise
comparisons. As would be predicted, the AD group
showed significant deficits in all domains, relative to
the NC group. In contrast, the FTD group was not
impaired in the orientation, attention, and visuospa-
tial domains, but showed deficits in all the other do-
mains (p⳱0.0000). The MDD and Affective groups
were again found to be statistically indistinguishable
on the basis of the domain scores. Compared with the
MDD group, the AD group was impaired on all mea-
sures except language, and the FTD group on mem-
ory, verbal fluency, naming, and language. Compared
TABLE 3. Proportion (in percent) of Subjects Scoring Below
the Recommended Cut-off Scores on the
Addenbrooke’s Cognitive Examination (ACE)
Cut-Off Score
⬍83 ⬍88
Major depression (MDD) 17
Affective 19
Alzheimer disease (AD) 87
Frontotemporal dementia (FTD) 63
Mixed 73
Normal-control (NC) 0 4
Note: 83 and 88 are the recommended cut-off scores for
research and screening, respectively.
with the Affective group, the AD group was, again,
impaired in each domain except language, and the
FTD group was impaired on verbal fluency, and nam-
ing. In summary, the AD group performed more
poorly in all domains (except language) than either
of the affective disorder groups, and, not surprisingly,
it was the lower verbal fluency and naming scores
that could clearly differentiate the FTD group from
both affective disorders groups. From the non–
progressive-degenerative groups, only the Affective
group was impaired in memory relative to the NCs
(p⳱0.0033). The combined AffectiveⳭMDD group,
again, showed significant impairment in memory
and verbal fluency (p⳱0.0062 and p⳱0.0377, respec-
tively) relative to the NC group.
The Mixed group was not as impaired as the AD
group in the orientation (p⳱0.0000) and visuospatial
domains (p⳱0.0421), and, relative to the FTD group,
39 performed better on the naming (p⳱0.0097), and
35 slightly worse on the attention (p⳱0.0429) tasks.
97
81
In view of our findings with the verbal fluency
77 scores and previous reports of selective impairment
of category fluency in AD, we analyzed the perfor-
mance of the groups on the two measures of fluency:
initial letter (P) and category (animals; see Figure 3).

Am J Geriatr Psychiatry 13:3, March 2005 223

 Dementia Versus Depression
ANCOVAs indicated significant mean differences be-
tween the groups on both scores, and age covaried
with category (p⳱0.0017), but not with letter fluency;
for letter fluency, F[5, 290]⳱14.68; p⳱0.0000; for cate-
gory fluency, F[5, 290]⳱37.28; p⳱0.0000. Pairwise com-
parisons revealed that both the AD group and the
FTD group were impaired on both category and letter
fluency, compared with the NC group (p⳱0.0000).
Relative to the MDD and Affective groups, they were
also similarly impaired on category fluency (p⳱
0.0000), whereas on letter fluency only the FTD group
showed a statistically demonstrable difference. Of
note was the finding that neither the MDD nor the
Affective group was statistically different from the
AD group on the letter fluency task.
The performance of the Mixed group was also sig-
nificantly poorer on both the letter and category flu-
ency tasks compared with the NC group (p⳱0.0001
and p⳱0.0000, respectively), and was not distin-
guishable from that of the AD and FTD groups on
either measure.
A logistic-regression analysis using the domain
scores of the ACE again to predict membership in the
FIGURE 3. Dot-line Chart of Letter Fluency and Category
Fluency (Derived Scores) Means in the Subject
Groups
Letter fluency
Category fluency
5 patients scoring lower than 88 had a confirmed pro-
Fluency Mean Scores
4
3
MD Affective AD FTD Mixed
Diagnostic Group
Note: Maximum attainable score is 7 on both tests of fluency.
This scoring system was previously derived by means of a Gaussian
distribution of the raw scores from normal-control subjects.
224
target groups with and without a progressive-degen-
erative dementia indicated a satisfactorily high pro-
portion, 89.5%, of the observed cases correctly pre-
dicted, and the naming, orientation, attention,
category fluency, memory, and attention scores being
the discriminating variables in this order of size of
effect; and age had a negligible effect.
Follow-Up of the Mixed Group
The most interesting group, from the point of dif-
ferential diagnosis, was the Mixed group. At presen-
tation, 8 out of a total of 22 patients in this group were
suspected to have a progressive-degenerative demen-
tia with concomitant affective disorder, and the other
14 were diagnosed as suffering primarily from an af-
fective disorder with a probable progressive-degen-
erative component. After an average 2 years of fol-
low-up, there was a marked shift to organic
diagnoses: of the 18 subjects for whom data were
available, 15 had clinically confirmed progressive-de-
generative dementias (8 still with concomitant affec-
tive symptoms). The organic etiologies were the fol-
lowing: AD: 10; FTD: 2; Lewy-body dementia: 1;
vascular dementia: 1; and dementia of other origin:
1. Three patients suffered from affective disorder
only. Of particular note is the fact that all the subjects
who eventually developed a progressive-degenera-
tive dementia scored below the recommended cut-off
point (88) for screening on the ACE (except one who
scored just at the cut-off point), and 15 out of the 16
gressive-degenerative dementia diagnosis within 2
years.
DISCUSSION
The earlier study by Mathuranath et al. established
that the Addenbrooke’s Cognitive Examination is
sensitive to both AD and FTD, and is able to distin-
guish between these two disorders.1 We have ex-
tended this work by showing that patients with af-
fective disorders are clearly separable from those
NC
with a progressive degenerative dementia whether
considered as a combined Affective-MDD group or
as separate subgroups. There was a small difference
between the affective groups and the normal-com-
parison group that failed to reach statistical signifi-
Am J Geriatr Psychiatry 13:3, March 2005

cance. Overall, therefore, the affective groups showed
very little impairment in the total ACE score, and this
could be attributed almost entirely to mild deficits in
memory and verbal fluency.
Considering first memory, it was surprising that
the memory impairment (as reflected by the ACE
Memory Domain score) reached statistical signifi-
cance in the Affective (but not the MDD) group. Nev-
ertheless, it is of note that this impairment was also
demonstrable in the combined Affective-MDD group.
This finding was unexpected for two reasons: First,
the memory deficit in depression has long been de-
scribed in the literature,27 but we found a significant
deficit in the generally less depressed Affective group
only. Second, our sample consisted of depressed pa-
tients with significant complaints of poor memory.
Other researchers reported a lack of relationship be-
tween the severity of memory complaints and mem-
ory performance28 and alternatively that the linkage
between depression and memory impairment was
present only in a subset of the depressed subjects,
rather than in all depressed individuals.29 It should
also be borne in mind that the complaint of “poor
memory” is a general catch-phrase used by subjects
in a folk-psychology sense to describe a broad range
of cognitive deficits, including poor attention, work-
ing, episodic, or semantic memory (see Berrios and
Hodges15). The memory complaints in patients with
affective disorders are likely to reflect, therefore, de-
fective attentional processing.
Moving to the verbal fluency findings, although
the total Verbal Fluency Domain scores were not sig-
nificantly impaired in either of the groups without a
progressive degenerative dementia diagnosis relative
to the normal-comparison group, a finer-grained
analysis of the two components uncovered some in-
teresting findings. Whereas the MDD and Affective
group showed unimpaired category fluency, their
performance on letter fluency was the same as that of
the AD group. In comparison to the huge literature
References
1. Mathuranath PS, Nestor PJ, Berrios GE, et al: A brief cognitive
test battery to differentiate Alzheimer’s disease and frontotem-
poral dementia. Neurology 2000; 55:1613–1620
2. 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
3. Copeland JR, Gurland BJ, Dewey ME, et al: Is there more demen-
tia, depression, and neurosis in New York? a comparative study
Am J Geriatr Psychiatry 13:3, March 2005
Dudas et al.
of letter- and category-based fluency in the demen-
tias,30–33 relatively little attention has been paid to
verbal fluency in affective disorders. Some research-
ers have found no significant difference between de-
pressed and normal-comparison subjects,31 whereas
others have reported impairment.32,33 A more recent
study, in accordance with our findings, reported im-
paired letter but unimpaired semantic fluency in
mildly depressed elderly subjects.34 Impaired verbal
fluency in severely depressed patients has also been
found to normalize after successful treatment of de-
pression.35 A previous study also found impairment
on letter fluency in patients with AD and depression,
and a trend toward poorer performance than in pa-
tients with AD with no depression.36 More recently,
however, Berger et al.37 found that AD and “AD-
with-depression” patients were indistinguishable on
category and letter fluency tasks. Our finding of a
correlation between age and category but not letter
fluency is consistent with the findings of others.37
The Mixed group, in whom we suspected under-
lying dementia but were unable to reach a clear di-
agnosis at presentation because of the prominence of
mood-related symptoms, represents a clinically im-
portant but neglected group. Of the 18 patients for
whom follow-up data were available, 15 developed a
clear-cut dementia within 2 years (seven pure pro-
gressive-degenerative [four AD, one FTD, one Lewy-
body dementia, and one vascular dementia], and
eight still mixed, but clearly diagnosed progressive
degenerative dementia). On the ACE, the Mixed
group was not separable from the AD and FTD
groups in terms of total scores. Our finding was con-
sistent with that of Lopez et al.38 in patients with AD
and concomitant depression. The ACE appears to be
particularly valuable in this challenging group. Pa-
tients scoring above 88 on the ACE are unlikely to
develop a progressive degenerative dementia,
whereas, for those scoring below 88, the likelihood is
very high.
of the elderly in New York and London using the computer di-
agnosis AGECAT. Br J Psychiatry 1987; 151:466–473
4. Blazer D, Williams CD: Epidemiology of dysphoria and depres-
sion in an elderly population. Am J Psychiatry 1980; 137:439–
444
5. DesRosiers G: Primary or depressive dementia: hedonic tone
and memory. Ph.D. dissertation, University of Cambridge, UK,
1994
225
Dementia Versus Depression
6. Garcia CA, Reding MJ, Blass JP: Overdiagnosis of dementia. J Am
Geriatr Soc 1981; 29:407–410
7. Roth M, Huppert FA, Tym E, et al: CAMDEX: The Cambridge
Examination for Mental Disorders of the Elderly. Cambridge, UK,
Cambridge University Press, 1988
8. Mattis S: Dementia Rating Scale. Odessa, FL, Psychological As-
sessment Resources Inc., 1988
9. Gifford DR, Cummings JL: Evaluating dementia screening tests:
methodologic standards to rate their performance. Neurology
1999; 52:224–227
10. McKhann G, Drachman D, Folstein M, et al: Clinical diagnosis of
Alzheimer’s disease: report of the NINCDS-ADRDA Work Group
Under the Auspices of the Department of Health and Human
Services Task Force on Alzheimer’s disease. Neurology 1984;
34:939–944
11. Neary D, Snowden JS, Mann DM: Classification and description
of frontotemporal dementias. Ann NY Acad Sci 2000; 920:46–51
12. Hodges JR, Miller B: The classification, genetics, and neuropa-
thology of frontotemporal dementia: introduction to the special
topic papers, part I: Neurocase 2001; 7:31–35
13. Hughes CP, Berg L, Danziger WL, et al: A new clinical scale for
the staging of dementia. Br J Psychiatry 1982; 140:566–572
14. Morris JC: The Clinical Dementia Rating (CDR): current version
and scoring rules. Neurology 1993; 43:2412–2414
15. Hodges JR, Berrios GE, Breen K: The multidisciplinary approach,
in Memory Disorders in Psychiatric Practice. Edited by Berrios
GE, Hodges JR. Cambridge, UK, Cambridge University Press,
2000, pp 101–121
16. Wechsler DA: Wechsler Memory Scale–Revised Manual. San An-
tonio, TX, The Psychological Corporation, 1987
17. Meyers J, Meyers K: Rey Complex Figure Test and Recognition
Trial, Professional Manual. Odessa, FL, Psychological Assessment
Resources, Inc., 1995
18. Wechsler D: WAIS–R Manual. San Antonio, TX, The Psychological
Corporation, 1981
19. Reitan RM, Wolfson D: The Halstead-Reitan Neuropsychological
Test Battery, 2nd Ed., Tucson, AZ, Neuropsychology Press, 1993,
pp 279–288
20. Benton AL, Hamsher K: Multilingual Aphasia Examination. Iowa
City, IA, University of Iowa Press, 1976
21. McKenna P, Warrington EK: The Graded Naming Test. Windsor,
Berks, UK, Nelson, 1983
22. Nelson HE: The National Adult Reading Test (NART). Windsor,
Berks, UK, NFER-Nelson, 1982
23. Warrington EK, James M: The Visual Object and Space Perception
226
Battery. Bury St. Edmunds, UK, Thames Valley Test Company,
1991
24. Warrington EK: Recognition Memory Test. Windsor, Berks, UK,
Nelson, 1984
25. Berrios GE, Bulbena-Villarasa A: The Hamilton Depression Rating
Scale and the Numerical Description of the Symptoms of De-
pression, in The Hamilton Scales: Psychopharmacology Series 9.
Edited by Bech P, Coppen A. Berlin, Heidelberg, Springer-Verlag,
1990, pp 80–92
26. Hughes CP, Berg L, Danziger W, et al: A new clinical scale for the
staging of dementia. Br J Psychiatry 1982; 140:566–572
27. DesRosiers G: Depressive pseudodementia, in Memory Disorders
in Psychiatric Practice. Edited by Berrios GE, Hodges JR. Cam-
bridge, UK, Cambridge University Press, 2000, pp 268–290
28. Derouesne C, Labomblez L, Thibault S, et al: Memory complaints
in young and elderly subjects. Int J Geriatr Psychiatry 1999;
14:291–301
29. Burt DB, Zembar MJ, Niederehe G: Depression and memory im-
pairment: a meta-analysis of the association, its pattern, and spec-
ificity. Psychol Bull 1995; 117:285–305
30. Monsch AU, Bondi MW, Butters N, et al: Comparisons of verbal
fluency tasks in the detection of dementia of the Alzheimer type.
Arch Neurol 1992; 49:1253–1258
31. Fossati P, Deweer B, Raoux N, et al: Deficits in memory retrieval:
an argument in favour of frontal subcortical dysfunction in de-
pression. Encephale 1995; 21:295–305
32. Geffen G, Bate A, Wright M, et al: A comparison of cognitive
impairments in dementia of the Alzheimer type and depression
in the elderly. Dementia 1993; 4:294–300
33. Kuzis G, Sabe L, Tiberti C, et al: Cognitive functions in major
depression and Parkinson disease. Arch Neurol 1997; 54:982–
986
34. Ravdin LD, Katzen HL, Agrawal P, et al: Letter and semantic flu-
ency in older adults: effects of mild depressive symptoms and
age-stratified normative data. Clin Neuropsychol 2003; 17:195–
202
35. Trichard C, Martinot JL, Alagille M, et al: Time course of prefron-
tal lobe dysfunction in severely depressed inpatients: a longitu-
dinal neuropsychological study. Psychol Med 1995; 25:79–85
36. Wefel JS, Hoyt BD, Massma PJ: Neuropsychological functioning
in depressed versus nondepressed participants with Alzheimer’s
disease. Clin Neuropsychol 1999; 13:249–257
37. Berger AK, Fahlander K, Wahlin A, et al: Negligible effects of
depression on verbal and spatial performance in Alzheimer’s dis-
ease. Dement Geriatr Cogn Disord 2002; 13:1–7
38. Lopez OL, Boller F, Becker J, et al: Alzheimer’s disease and de-
pression: neuropsychological impairment and progression of the
illness. Am J Psychiatry 1990; 147:855–860
Am J Geriatr Psychiatry 13:3, March 2005