J. Wippold M. Duchek,
Franz Janet
II, MD PhD
#{149} Mokhtar #{149} Elizabeth
Senile Dementia A Longitudinal
of atrophy
during
the
study period. However, the overlap of indexes between the patients with SDAT and the control subjects indicates that CT data cannot be used alone to predict the presence or progression
ual
of dementia
in
individ-
cases.
Index terms: Aging #{149} Brain, atrophy, 10.83. Brain,CT, 10.1211 #{149} Brain,volume, 10.1211. Dementia, 10.83
Radiology
1991;
From
179:215-219
the Mallinckrodt
Institute
of Radiolo-
gy (F.J.W., M.H.G.). the Departments of Neurology and Neurosurgery (Neurology) and Pathology (Neuropathology) (J.C.M.) and Occupational Therapy (J.M.D.), the Division of Biostatistics (E.A.G.), and the Alzheimer’s Disease Research Center (F.J.W., M.H.G., J.C.M., J.M.D., E.A.G.), Washington University School of Medicine, 510 5 Kingshighway, St Louis, MO 63110. Received August 6, 1990; revision requested September 18; revision received December 21; accepted December 26. Supported by National Institute on Aging grants AG03991 and AG05681 and National Institute of Mental
Health
grant
quests
to F.J.W.
c
RSNA,
MH31054. 1991
Address
reprint
#{149} John
C. Morris,
and Healthy CT Study’
Volumetric indexes of cerebral atrophy obtained by using computed tomography (CT) were measured longitudinally in patients with senile dementia of the Alzheimer type (SDAT) and in healthy elderly control subjects. Measurements were made three times over a 51-month period. Of the original 44 patients with SDAT, five were available for CT examination at the last time of assessment (51 months); in contrast, 41 of the original 58 control subjects were still available for study at 51 months. As a group, scans of SDAT subjects showed greater atrophy than those of control subjects in all volumetric indexes at each time of testing and demonstrated greater progression
H. Gado, MD A. Grant, PhD
re-
S
MD
Aging:
dementia of the Alzheimem (SDAT) is a disorder chanactemized by insidious onset of progressive deterioration of memory and global cognitive function (1,2). Since the introduction of computed tomography (CT) for clinical use, considerable interest has focused on whether the cerebral atrophy associated with SDAT is greaten than that associated with normal aging. It is generally accepted that cerebral atrophy, as measurcd with CT, occurs during normal aging (3-9). However, the degree and progression of additional atrophy in SDAT remain uncertain. Some investigators have reported that atrophy is strongly associated with SDAT or other dementing processes (10-20), whereas others have found only a ENILE
type
weak
association
(2i-23).
These
con-
flicting results may be due in part to differences in the use of linear versus volumetric CT measurement techniques. In a previous report, Gado et al (24) directly compared linear and volumetric measurements of a group of persons with SDAT with those of healthy control subjects and found that linear measurements showed less pronounced group differences in atrophy than did volumetric measunements. This suggests that volumetric measurement may be a more sensitive indicator of brain volume and atrophy. The purpose of this study was to compare longitudinal differences in the progression of cerebral atrophy between patients with SDAT and control subjects by using volumetric measurements.
SUBJECTS
AND
#{149}
METHODS
cal diagnostic criteria (25). The full characteristics of the research sample have been described previously (26). The diagnosis of SDAT was made on the basis of information obtained in an extensive clinical interview and neurologic examination, which allow determination of a clinical dementia rating (CDR) for each subject (27,28). The CDR is a five-point scale describing the degree of dementia: CDR 0 = no dementia, CDR 0.5 = questionable dementia, CDR 1 = mild dementia, CDR 2 = moderate dementia, and CDR 3 = severe dementia. The reliability of the CDR has been established (29,30). All subjects underwent longitudinal assessments of clinical and psychometric performance, electroencephalographic analysis, laboratory tests, and CT examinations. The CT examinations were conducted at entry and 15, 34, and 66 months after entry into the study. Because CT scans at entry were obtained with a 7070 EM! scanner (Hayes, England) and lent themselves only to linear rather than volumetnc measurements of the cerebrospinal fluid (CSF) space (31), only the CT data from the follow-up scans will be reported in this article. The period of this CT study, therefore, is 51 months, although clinical evaluations spanned the entire 66 months. Furthermore, only scans of subjects who were either CDR 0 or CDR 1 or greater at the first follow-up examination were included in this study. At the 0-, 19-, and 51-month CT volumetric assessments, there were 51, 49, and 41 control subjects, respectively, for whom technically suitable CT scans were available for review. Attrition factors included migration from St Louis, death (three at 51 months), and refusal to participate. All control subjects remained CDR 0 throughout the 51-month period. Forty-four subjects with SDAT were classified as CDR 1 at the initial evaluation. During the study, four of the 16 subjects originally graded as CDR 0.5 pro-
Subjects Forty-four
elderly patients with mild questionable SDAT, and 58 healthy control subjects of comparable age, education, and social position were enrolled in a longitudinal study of SDAT and healthy aging using validated cliniSDAT,
16 with
Abbreviations: Rating, CSF
CDR Clinical Dementia cenebrospinal fluid, SDAT nile dementia of the Alzheimer type, 5% umetnic index of sulcal size, V% volumetric index of ventricular size, V5% volumetric dex of total CSF volume. =
=
Sevolin-
215
gressed
to CDR
scan,
1 by
performed
and
thus
plc.
Technically
the
first
follow-up
15 months
were
included
after
in the
optimal
SDAT
scans
The
composition
summarized
the
of the
in Table
SDAT
group
of dementia study with
CDR
group
In
summary,
the control group (Table
who
had
scans
testing
the for
group and 1). Scans
to CDR
1 at
CDR
The
CT
scans
NJ),
both
at all
with scanner
follow-up
of which
were
the cranial nium were scanning
cavity. obtained
collimation,
Axial scans with the
view.
Data
of
125
of the standard kVp,
stored
on
magnetic
tive volumetric of volumetric ously
described
(24).
on segmenting parenchyma “threshold”
ed by the section.
low area
operator
and
sulci)
section. The after selecting
cluded cerebral
for each
5 seconds
for
quantita-
it depended
the
CT repeated that in-
and excluded the obtaining a value for
volume within between the
the CT values
section; for
values volume
for
and sulci and alone yielded
the the
sections
selected.
the
of sulci. Similarly, the volume of the cranial cavity within the CT section was determined by selecting a threshold value that separated bone (cranium) from soft tissue. Seven
CT
of the
skull
were
excluded
artifacts.
sulci,
tions values,
and The
were the
uppermost
to avoid volumes
base
beam-hardening of the
ventricles,
cranial cavity in the seven secwere summed. From the resultant indexes
of ventricular
size (V%), sulcal size (5%), and volume (VS%) were computed
216
The sections
and
volumetric
#{149} Radiology
SDAT
.
. .
.
.
49
.
12
5
.
51
25
0 no dementia, and are numbers of patients.
CDR
1, 2, and
. .
.
.
Control
Total
.
. .
.
. .
49 3
41
.
1 1 3
.
5
mild,
Small
peripheral
focal
Table
moderate,
.
.
.
.
.
.
.
.
.
141 25 24 11
41 and
severe
201
dementia,
respec-
the
sulcal
a potential
measurements,
(small)
they
source
2
Correlations
Volumetric Size, Sulcal Volume at ar-
of Age with CT Indexes of Ventricular Size, and Total CSF 0 Months
Volumetric Indexes V%
5% VS%
Patients with SDAT* (n = 30)
Control Subjects (n 51)
.05 .20 .09
.18 .43 .37
rep-
of error.
4
Partialled
for
dementia
severity
by using
level.
Analysis
were
nested
within
the
SDAT
group,
be-
the (yen-
particular
same procedure was a region of interest
ventricular the difference
Control
8
thus accounting for any correlations secondary to repeated measures on subjects.
(voxels)
value determined of the CSF spaces within
. .
Data were analyzed with SAS software (SAS Institute, Cary, NC) (32). To assess the changes in volumetric indexes longitudinally, a split-plot analysis of variance was conducted by using all available scans. There were three class variables: group, with two levels (control subjects and patients with SDAT); time, with three levels (0, 19, and 51 months followup); and subjects, with 89 levels (81 subjects with available scans at 0 months and eight additional subjects with available scans at 19 months). In the model, subjects
independent
of pixels
the ventricles sulci, thus
ventricles ventricles
mAs, of
The method has been previ-
Briefly,
number
the threshold (or volume)
tricles
Data
areas of attenuation into and CSF by means of a CT attenuation value select-
The
51 Months
interval
reconThe reconon film
disk
evaluations. assessment
45
was
for a single section. Data were structed on a 256 X 256 matrix. structed images were recorded and
resent
cra-
field
time
of Follow-up
with
25-cm2
acquisition
30
Note.-CDR tively. Values
from
2 or SoIselin,
equipped
and
Total
. .
.
eas of low attenuation, such as lacunar infarctions, were excluded from the ventricular regions of interest. Because they could not, however, reliably be excluded
volumetric assessCSF spaces and
parameters
8-mm
51
.
measurements.
in
visits
a Somatom (Siemens,
software that allowed ment of the intracranial
.
16 11 3
group.
dim-
Measurements
were obtained matom DRH
.
(standard
were
mean age follow-up deviation
5.06), compared with 72.9 years deviation = 4.68) in the control
Volumetric
SDAT
ages of the cranial volume. These volumetric indexes were used as indexes of cerebral atrophy. Subjects with large areas of periventricular white-matter attenuation were cxcluded from this study because of potential contamination of ventricular volume
0.5 at 0
19 months
mated from this study. The the SDAT group at the first scan was 72.6 years (standard
Months
51
from
60 from the SDAT from two subjects
from
and
0, 19,
obtained
periods-141
progressed
months
were
at
l9Months Control
SDAT
0 1 2 3
because
throughout unavailability
201
three
Groups
is
within
largely
Control
and
OMonths
CT scanning (eg, at 51 months, 17 subjects were institutionalized, nine had died, and several had left the St Louis area). for the
of SDAT
avail-
1. Attrition
progression subsequent
1
Composition
with respective-
study
occurred
sam-
were
able for 30, 25, and five patients SDAT at 0, 19, and 51 months, ly.
Table
entry,
total CSF as percent-
RESULTS The correlations three volumetric atrophy (V%, 5%, months are shown cause there was a tion between age
of age and the indexes of cerebral and VS%) at 0 in Table 2. Bemoderate correlaand stage of dementia severity as measured by the CDR (r -.36), the CDR was partialbed out of this portion of the analysis. Although the correlations are moderate, age is significantly cormebated with 5% (P < .002) and VS% (P < .007) in the control group. The oldem the control subjects, the greater the S% and VS% were. Age was not significantly correlated with the measures of cerebral atrophy in the SDAT group, perhaps because the atrophy caused by the disease is greater
than the atrophy that occurs during normal aging. The SDAT group had significantly greater volumetric indexes than the control group on all three measures: V%, F(1, 87) = 22.67; S%, F(1, 87) 32.43; and VS%, F(1, 87) 51.7. All were significant at P < .0001. There was also a significant group-by-time interaction for all three measures: V%, F(2, 108) = 29.14, P < .0001; S%, F(2, 108) = 13.73, P < .0001; VS%, F(2, 108) = 32.76, P < .0001. Only V% changed over time in the control subjects: F(2, 108) = 12.88, P < .0001. A comparison of the means in the control group showed that a slight increase occurred between 19 and 51 months (P < .0007). All three measures increased significantly in the SDAT group over time: V%, F(2, 108) = 58.98, P < .0001; S%, F(2, 108) 13.61, P < .0001; VS%, F(2, 108) 46.92, P < .0001. V% increased significantly between 0 and 19 months (P < .0001) and between 19 and 51 months (P < .01). Both S% and VS% increased significantly from 0 to 19 months but remained stable between 19 and 51 months (Table 3). The box plots of each of the volumetric indexes over time are shown in the Figure. The plots for the controb subjects show a fairly stable course, while the plots for the SDAT group show more variability and an increase over time. Some overlap cxApril
1991
Table Mean
3
Volumetric
Indexes
for Control
and SDAT
Subjects
at
No. of Subjects
V%
51 30
5.7 (2.8) 9.7 (4.3)
Control SDAT Note-Numbers
51 Months
Follow-up
19 Months
5%
in parentheses
and
0, 19,
0 Months
VS%
7.2(2.9) 1 1.2(3.4)
12.9(4.7) 20.9(4.8)
are standard
deviations.
51 Months
No. of Subjects
V%
S%
49 25
5.7 (2.6) 11.8 (5.7)
6.6(2.6) 12.4 (3.2)
See text for discussion
of the significant
interaction
N
__
-r
t
T
I
I
>
20.5), 82% of the subjects had SDAT; only 18% were control subjects. Furthemmore, no systematic association with the severity of SDAT was cvident across the middle and higher mange of VS% values. That is, subjects from each CDR category were in both the middle and higher mange of the distribution.
L
DISCUSSION
z
40
N C’) 5-
U
00
20
U>
LU
-:--.4-
C
IL-
1-
-1-4-
I
-J
11
0
l
-‘-
I
-‘-
‘
I
I
--
______________________
>
This study was conducted to assess the longitudinal differences in cenebra! atrophy between SDAT and healthy aging. The results cleanly indicate large differences in ventricular size, sulcal size, and total CSF volume between patients with SDAT and healthy control subjects over a 51-month period. Scans of the SDAT group showed greater atrophy compared with those of healthy control
subjects use
40
E 0
I[
II
I
I
percentile,
time.
horizontal
(included yond
over
1
1
0
19
the
in data
indexes
Upper
I
1
line
within
interquartile
Vertical
of box box
lines
179
Number
#{149}
outside
‘
“
‘
IY
,)
1
I
OF FOLLOW-UP size
(V%),
75th
percentile, small
sulcal horizontal
of boxes
size
(S%),
lower extend
and
margin lines
and
to nearest
total
CSF
of box dots
point
=
vol-
25th outliers
not be-
range.
ists between the distributions for control and demented subjects. For example, in the distribution of VS% at 0 months, 98% of subjects with VS% values between 4.5 and 14.5
Volume
median,
I
O
51
of ventricular
margin
analysis).
#{149}
I
MONTHS (VS%)
-
T
I
of volumetric
-
L
T
0
plots
#{149} _____
_____
20
1
at the
of volumetric
first
evaluation
were control subjects. Only one subject with SDAT had a VS% within this mange. Greater overlap between SDAT and control subjects occurred in the middle of the distribution
with
measurements.
Cerebra! atrophy in patients SDAT was more pronounced examination. Gado et al (31)
I
>z.
ume
by SDAT
14.5-20.5);
LU
Box
S%
6.3 (2.7) 12.4(3.9)
for time
(VS%,
V%
were control patients with
40
D
41 5
12.3(4.2) 24.2(5.6)
SDAT
Control
No. of Subjects
VS%
with at each also
found greater atrophy in subjects with mild SDAT at entry into their study by using linear measurements. Therefore, some degree of cerebral atrophy already has occurred at an early stage of SDAT. Our results demonstrate that the amount of atmophy in SDAT increased over time. Except for the slight increase in V%, the volumetric indexes of cerebral atrophy were fairly stable for the cognitively normal control subjects during the 51-month study. These findings are in agreement with those from the recently reported longitudinal study of deLcon et a! (17), in which 50 patients with SDAT and 45 control subjects were followed up for 3 years. Similar results were found in earlier longitudinal studies performed with much smaller samples (18,19). the
Our data show that enlargement CSF spaces in SDAT occurs
Radiology
of in
#{149} 217
both
the
ventricles
and
cerebral
indicating
a process of mixed and cortical atrophy. Previous tigators have maintained that
sulci, central invesyen-
tricular enlargement was a better indicator of the presence of SDAT (16). However, their studies relied primarily
on
linear
measurements.
Because
of the geometric complexity of the cerebral sulci, linear measures of a few sulci underrcpnesent differences in their overall volume. Gado et a! (24) compared linear and volumetric measurements of the same subjects and found that volumetric measurements provided better distinction between patients with SDAT and control subjects. DeLeon et a! (17) maintamed that the linear method is as accurate as the volumetric method and easier to use in CT analysis, but they
did
not
measure
the
sulci.
A po-
tential source of error in our study was the inclusion of small peripheral white-matter
areas
of low
attenua-
tion, such as lacunar infarctions, into the sulcal measurement. This was minimized by excluding all individuals with large areas of low attenuation in the white matter from the study
population.
Beam-hardening
artifacts were avoided by eliminating from consideration the uppermost CT sections. Given the limitations sulcal measurements, results were evaluated
as indexes
en than nations. Although
the phy
of atrophy
as absolute
mean
volume
there
values
between
were
the
of math-
differences
SDAT
in
of atro-
and
control
groups, considerable overlap existed between individual patients with SDAT and control subjects. Thus, it was not possible to predict the presence on absence of SDAT for a given case solely on the basis of CT findings.
Although
the
extreme
values
of
the VS% distribution corresponded well with the presence on absence of SDAT, the middle range of VS% distribution was problematic. On the basis of magnetic resonance (MR) imaging volume measurements of the hippocampus,
Scab
et a! (33)
suggested
that detectable hippocampal atrophy occurs early in SDAT and may be more indicative of the presence of the disease than measurements of ventricular size, sulcal size, or brain atrophy.
Dementia severity by the CDR did not highly with CT volumetric (r ranged from .02 to .27).
suned
with mean ing
ence 218
the overlap between indexes of atrophy, suggests
that
neither
non the severity #{149} Radiology
as meacorrelate measures Combined
groups
subjects months
can
differ
still and
problems longitudinal
between
those
of conducting studies
extended in SDAT popula-
in SDAT.
The
control
MR the fact CSF
bone
is not
visualized
Investigators
have
already
demented
phy
progressively
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
1 1.
during
Except
U
Adams RD. Victor M. Principles of neurology. 4th ed. New York: McGraw-Hill, 1989; 923-932. Berg L, Morris JC. Aging and dementia. In: Pearlman AL, Collins RC, eds. Neurobiology of disease. New York: Oxford University Press, 1990; 299-322. Roberts MA, Caird Fl, Grossart KW, Steyen JL. Computerized tomography in the diagnosis of cerebral atrophy. J Neurol Neurosung Psychiatry 1976; 39:909-915. Hahn FJY. Rim K. Frontal ventricular dimensions on normal computed tomography. AJR 1976; 126:593-596. Barron SA, Jacobs L, Kinkel WR. Changes in size of normal lateral ventricles during aging determined by computerized tomography. Neurology 1976; 26:1011-1013. Gyldensted C. Measurements of the nonma! ventricular system and hemispheric sulci of 100 adults with computed tomography. Neuroradiology 1977; 14:183-192. Haug G. Age and sex dependence of the size of normal ventricles on computed tomognaphy. Neunoradiology 1977; 14:201204. Jacoby RJ, Levy R, Dawson JM. Computed tomography in the elderly. I. The normal population. Br J Psychiatry 1980; 136:249-255. Yamaura H, Ito M, Kubota K, Matsuzawa T. Brain atrophy during aging: a quantitative study with computed tomography. Gerontol 1980; 35:492-498. Huckman MS. Fox J, Topel J. The validity of criteria for the evaluation of cerebral atrophy by computed tomography. Radiology 1975; 116:85-92. Roberts MA, Caird Fl. Computenised tomography and intellectual impairment in
the elderly. J Neurol try 1976; 39:986-989. 12.
13.
14.
15.
subjects.
increased
study.
References
demon-
In conclusion, our results demonstrate greater cerebra! atrophy in SDAT than in healthy aging. This increased atrophy was apparent in patients with mild SDAT, and the atro-
of the
Acknowledgments: We thank Leonard Berg, MD. for critical review of this material and Beverly McDonald for manuscript preparation.
with
strated the utility of MR imaging in SDAT (33,38). Validation studies of MR imaging-based volume measurements have also been published (39). One of the major disadvantages of MR imaging is the lengthy time needed to complete the image. This is particularly critical when studying and
period.
group,
imaging. The identification of high cerebral sulci without antifrom the skull should improve volume determinations.
51 months
for minimal increase in V%, the volumetric indexes of atrophy observed in healthy, aging control subjects were fairly stable during the same
available for CT at 51 those who were not. The
however, only diminished from 58 at entry to 41 at 51 months, permitting reasonable estimation of CT features of atrophy in healthy aging oven this period. In the future, MR imaging undoubtedly will be the modality of choice in volumetric studies of the brain. Compared with CT, MR imaging offers superior tissue contrast and multiplanar display. These featunes facilitate analysis of specific structures such as the hippocampus of the temporal lobe, which is a site of selective atrophy in SDAT (36,37). Cortical
the
SDAT
tions are well recognized (34,35). Attnition due to severe debility or death reduces the study population between the times of follow-up. Themefore, the data from the five SDAT patients who were scanned at 51 months may actually underrepresent the deterioration in the original SDAT group. For this study, the small sample size at 51 months limits conclusions about the long-term CT
elderly
in
this findthe pres-
of SDAT
nificantly
changes
determi-
of indexes
be inferred from CT measures. Attrition was notable in the SDAT subjects during the study. The numben of SDAT subjects with technically optimal CT scans diminished from 30 at the first follow-up examination (0 months) to only five at the last foblow-up (51 months). At the time of the first follow-up scan, these five subjects did not differ significantly from the 25 SDAT subjects who were not scanned at 51 months. Subject age, duration of dementia, quantitative measures of cognitive impairment, and CT measures did not sig-
16.
17.
Neurosung
Psychia-
Gonzalez CF. Lantieni RL, Nathan RJ. The CT scan appearance of the brain in the normal elderly population: a correlative study. Neuronadiology 1978; 16:120122. deLeon MJ, Ferris SH, Blau I, et al. Comelations between computenised tomographic changes and behavioural deficits in senile dementia (letter). Lancet 1979; 2:859860. Jacoby RJ, Levy R. Computed tomography in the elderly. II. Senile dementia: diagnosis and functional impairment. Br Psychiatry 1980; 136:256-269. deLeon MJ, Ferris SH, George AE, Reisberg B, Kricheff II, Genshon S. Computed tomography evaluations of brain-behavior relationships in senile dementia of the Alzheimer’s type. Neurobiol Aging 1980; 1:69-79. Bninkman SD, Samwar M, Levin HS, Monmis HH III. Quantitative indexes of computed tomography in dementia and nonmal aging. Radiology 1981; 138:89-92. deLeon MJ, George AE, Reisberg B, et al. Alzheimer’s disease: longitudinal CT studies of ventricular change. AJNR 1989;
April
1991
18.
19.
20.
21.
22.
23.
24.
10:371-376. Luxenbeng JS, Haxby JV, Creasey H, Sundaram M, Rapoport SI. Rate of ventriculan enlargement in dementia of the Alzheimer type correlates with the rate of neuropsychological deterioration. Neurology 1987; 37:1135-1 140. Brinkman SD, Largen JW. Changes in brain ventricular size with repeated CAT scans in suspected Alzheimer’s disease. Am J Psychiatry 1984; 141:81-83. George AE, deLeon MJ, Rosenbloom 5, et al. Ventricular volume and cognitive deficit: a computed tomographic study. Radiology 1983; 149:493-498. Earnest MP, Heaton RK, Wilkinson WE, Manke WF. Cortical atrophy, ventricular enlargement, and intellectual impairment in the aged. Neurology 1979; 29:11381143. Kaszniak AW, Gamron DC, Fox JH, Bergen D, Huckman M. Cerebral atrophy, EEG slowing, age, education, and cognitive functioning in suspected dementia. Neurology 1979; 29:1273-1279. Hughes CP, Gado M. Computed tomognaphy and aging of the brain. Radiology 1981; 139:391-396. Gado M, Hughes CP, Danziger W, Chi D, Jost G, Berg L. Volumetric measurements of cerebrospinal fluid spaces in demented subjects and controls. Radiology 1982;
25.
26.
27.
28.
29.
1988; 30.
McCulla
MM,
31.
32.
Coats
179
#{149} Number
1
M, VanFleet
E, Morris
JC.
Reliability
TF.
NMR measurements of hippocampal atnophy in Alzheimer’s disease. Magn Reson Med 1988; 8:200-208. Zimmen AW, Calkins E, Hadley E, Ostfeld AM, Kaye JM, Kaye D. Conduction of nesearch in geriatric populations. Ann In-
tern Med 1985; 103:276-283. 35.
36.
Botwinick J, Storandt M, Berg L, Boland S. Senile dementia of the Alzheimer type: subject attrition and testability in research. Arch Neunol 1988; 45:493-496. Hyman BT, Van Hoesen GW, Damasio AR, Barnes CL. Alzheimer’s disease: cell-specific pathology isolates the hippocampal
formation. 37.
Ball
Science
MJ, Fisman
new definition
N, Du-
of
clinical nurse specialists in the staging of dementia. Arch Neurol 1989; 46:12101211. Gado M, Hughes CP, Danzigen W, Chi D. Aging, dementia, and brain atrophy: a longitudinal computed tomognaphic study. AJNR 1983; 4:699-702. SAS Institute, Inc. SAS/STAT guide for personal computers. 6th ed. Cary, NC: SAS Institute, 1987. Scab JP, Jagust WJ, Wong STS, Roos MS.
Reed BR, Budinger
Volume
34.
38.
45:31-32.
chek J, Grant
33.
144:535-538.
Morris JC, McKeel DW Jr. Fulling K, Tonack RM, Berg L. Validation of clinical diagnostic criteria for Alzheimer’s disease. Ann Neurol 1988; 24:17-22. Berg L, Hughes CP, Coben LA, Danziger WL, Martin RL, Knesevich J. Mild senile dementia of the Alzheimer type: research diagnostic criteria, recruitment and description of a study population. J Neurol Neurosurg Psychiatry 1982; 45:962-968. Hughes CP, Berg L, Danziger W, Cohen LA, Martin RL. A new clinical scale for the staging of dementia. Br J Psychiatry 1982; 40:566-572. Berg L. Clinical Dementia Rating (CDR). Psychopharmacol Bull 1988; 24:637-639. Burke WJ, Miller JP, Rubin EH, et al. The reliability of the Washington University Clinical Dementia Rating. Arch Neurol
V. et al.
of Alzheimer’s
disease:
A
a
hippocampal dementia. Lancet 1985; 1:1416. Bowen BC, Barker WW, Loewenstein DA, Sheldon J, Duara R. MR signal abnormalities in memory disorder and dementia.
AJNR 39.
1984; 225:1168-1170.
M, Hachinski
1990; 11:283-290.
Jack CR Jn, Bentley MD, Twomey CK, Zinmeister AR. MR imaging-based volume measurements of the hippocampal formation and anterior temporal lobe: validation studies. Radiology 1990; 176:205-209.
Quantitative
Radiology
#{149} 219