Hemispheric Differences in Hippocampal Volume
Predict Verbal and Spatial Memory Performance
in Patients With Alzheimer’s Disease
Leyla de Toledo-Morrell,1–3*Brad Dickerson,1
M.P. Sullivan,1C. Spanovic,1Robert Wilson,1–3
and D.A. Bennett1,3
1Department of Neurological Sciences, Rush University,
2Department of Psychology, Rush University,
3Rush Alzheimer’s Disease Center, Rush University,
for the acquisition of new declarative knowledge, has been well-documented
in Alzheimer’s disease (AD), although the relation of such atrophy to the
extent of memory dysfunction in these patients has been less clear. In the
present study, 18 patients with a clinical diagnosis of probable AD were
studied with a high-resolution, quantitative magnetic resonance imaging
(MRI) protocol, as well as the verbal and spatial versions of the Buschke
controlled learning task. The volumes of the hippocampal formation and, as
a control for generalized atrophy, parahippocampal gyrus and temporal
neocortex were computed from gapless coronal slices taken perpendicular to
the long axis of the hippocampus. To correct for individual differences in
brain size, volumes of regions of interest were divided by total intracranial
volume. Separate stepwise regression analyses (with age, right and left hip-
pocampal, parahippocampal gyrus, and temporal lobe volumes as the inde-
pendent variables) showed that left hippocampal volume was the best pre-
dictor of free recall and delayed free recall of verbal information (P ? 0.0042
and P < 0.0001, respectively). Recall and delayed recall of the spatial
location of verbal items were best predicted by right hippocampal volume (P
? 0.0054 and P ? 0.0118, respectively). Memory scores did not correlate
either with parahippocampal gyrus or temporal lobe volume. Furthermore,
the relation between hippocampal volume and memory function observed in
cases with AD did not hold for healthy aged control subjects. Hippocampus
Atrophy of the hippocampal formation, a region important
© 2000 Wiley-Liss, Inc.
poral lobe; parahippocampal gyrus
hippocampal formation; MRI; imaging; volumetry; tem-
aging (MRI) techniques provide a unique tool for studying brain-behavior
relations in vivo. Atrophy of the hippocampal forma-
tion,1a region critically important for the acquisition of
edge), has been well-documented in Alzheimer’s disease
(AD), using both quantitative (e.g., Seab et al., 1988;
Kesslak et al., 1991; Jack et al., 1992, 1997; Killiany et
al., 1993; de Toledo-Morrell et al., 1997; Köhler, et al.,
1998) and semiquantitative or qualitative MRI analyses
(e.g., Jobst et al., 1992, 1994; de Leon et al., 1997).
However, the relation between measures of memory
function and hippocampal atrophy in patients with a
clinical diagnosis of probable AD has not been clear-cut.
While some investigators reported a significant correla-
tion between hippocampal volume and memory perfor-
al., 1998; Köhler et al., 1998), others found that hip-
pocampal atrophy was not a good predictor of the extent
of memory impairment in AD (de Leon et al., 1997;
Mori et al., 1997), possibly due to “floor” effects.
The robustness of a relationship between the extent of
a great extent, on the behavioral measures used, as well as
on whether hippocampal atrophy is assessed quantita-
tively or qualitatively. First, the nature of the behavioral
task is critical, since some may be more sensitive to hip-
pocampal dysfunction than others. For example, in Wil-
son et al. (1996), the extent of recall of a standardized
pal volume, while the recall of short passages did. Sec-
ondly, a qualitative scale of hippocampal atrophy may be
with quantitative measures.
Grant sponsor: National Institute on Aging; Grant number: P01 AG09466.
M.P. Sullivan is now at the Biomedical Department, University of Alaska,
*Correspondence to: Leyla de Toledo-Morrell, Department of Neurological
Sciences, Rush Medical College, 1653 W. Congress Parkway, Chicago, IL
60612. E-mail: email@example.com
Accepted for publication 2 November 1999
1As used here, the term “hippocampal formation” refers
to the entire hippocampal region including the dentate
gyrus, hippocampus proper, and subiculum.
HIPPOCAMPUS 10:136–142 (2000)
© 2000 WILEY-LISS, INC.
Buschke and Grober (1986) and Grober and Buschke (1987)
argued that it may be difficult to distinguish between “apparent”
have impaired attention or may use inefficient strategies in acquir-
ing information. By controlling the type of processing carried out,
they were able to identify “genuine” memory deficits in aged sub-
jects and in patients in the early stages of AD. Others have also
reported on the utility of the verbal version of the Buschke task in
In the present study, patients with a clinical diagnosis of AD
were examined with high-resolution, quantitative MRI protocols
as well as the verbal and spatial versions of the Buschke controlled
learning task. The major aim of the study was to determine the
hemisphere- and material-specific nature of a possible relationship
between hippocampal volume and memory performance.
MATERIALS AND METHODS
Subjects consisted of 18 mildly demented patients (mean age,
able AD according to NINCDS/ADRDA criteria (McKhann et
al., 1984). All patients were evaluated at the Rush Alzheimer’s
Disease Center (RADC, Chicago, IL), as previously described in
Wilson et al. (1996) and in deToledo-Morrell et al. (1997). The
evaluation included a medical history, neurological examination,
neuropsychological testing that conformed to NINCDS/ADRDA
guidelines (McKhann et al., 1984), as implemented by the Con-
sortium to Establish a Registry for Alzheimer’s Disease (CERAD,
Morris et al., 1989), and blood tests. The clinical diagnosis of
probable AD required a history of cognitive decline and neuropsy-
chological test evidence of impairment in at least two cognitive
domains, one of which had to be memory. Exclusion criteria were
evidence of other neurologic, psychiatric, or systemic conditions
that could cause cognitive impairment (e.g., stroke, alcoholism,
major depression). The Mini Mental State Examination (MMSE,
range, 20–28); their mean delayed CERAD word list recall was
3.17 (range, 0–6, with only one patient scoring 6).
For comparison purposes, normative data were obtained from
30 healthy aged control participants (mean age, 72.4; range, 64–
bers of patients and the RADC staff, as well as from hospital
volunteers. They had a standard evaluation including a medical
history, neurological examination, and abbreviated neuropsycho-
logical testing. Inclusion criteria for the control subjects were
MMSE ?28, CERAD delayed word list recall ?6, and a normal
used for the AD diagnosis.
Informed consent was obtained from all participants according
to the rules of the Human Investigation Committee of Rush Med-
Acquisition and Quantitation of MRI Data
All MR images were acquired on a 1.5 Tesla General Electric
ular to the long axis of the hippocampal formation with the fol-
lowing parameters: matrix ? 256 ? 256; field of view ? 16 cm;
eight acquisitions, TR ? 400, TE ? 13. In addition, gapless,
5-mm sagittal slices were taken spanning the entire brain with the
acquisition, TR ? 200, TE ? 12.2
An interactive three-dimensional (3-D) image analysis program
(Amersham Image Analysis System with software by Loates Asso-
ciates) was used to compute the volume of regions of interest. To
correct for normal individual differences in brain size, the volumes
of given regions were normalized by dividing with intracranial
volume derived from sagittal slices. To compute intracranial vol-
ume, the inner table of the cranium was traced in consecutive
sagittal sections spanning the entire brain. At the level of the fora-
men magnum, a straight line was drawn from the inner surface of
the clivus to the most anterior extension of the occipital bone.
Volumes for the hippocampal formation, parahippocampal gy-
rus, and temporal lobe were computed separately for the right and
left hemispheres from coronal slices. The latter two brain regions
were used as a control for the effects of generalized atrophy. Nor-
Morrell et al., 1997; Wilson et al., 1996).
Tracings of the hippocampal formation started with the first
section caudal to the amygdala, where the dentate gyrus could be
clearly identified, and included the fimbria, the dentate gyrus, the
hippocampus could be clearly seen without partial volume averag-
ing were included (usually 6–7 slices). The volume of the parahip-
pocampal gyrus was derived from seven consecutive coronal im-
ages beginning at the level of the amygdala, one slice anterior to
that used for hippocampal measurements. The amygdala and hip-
pocampal formation were excluded from this measure. Temporal
lobe volume was also computed from seven consecutive coronal
images, starting at the level of the amygdala. To separate the tem-
poral lobe from the temporal stem, a straight line was drawn from
the most lateral extension of the inferior horn of the lateral ventri-
cle to the most medial extension of the gray matter of the superior
temporal gyrus. The hippocampal formation, the amygdala, and
Volumetric determinations were carried out by two investiga-
tors (M.P.S. and B.D.); their interrater agreement was better than
96% on volumes of all three regions of interest.
2Our MRI protocol has changed since the initiation of this study.
We now acquire 1.6-mm coronal images of the entire head with
an SPGR pulse sequence. For purposes of consistency, this inves-
tigation was restricted to those patients scanned with our old
MEMORY PERFORMANCE IN AD
presented four at a time. Items were shown as line drawings, one
the experimenter gave a category cue verbally (e.g., bird), the sub-
ject had to search the display, and point to and name the object
from that category (e.g., owl). After this procedure was completed
by presenting each category cue to the subject. If he/she failed to
recall an item in response to its cue, the item was shown again, the
search was performed, and so on, until immediate cued recall was
as described above, until all 16 items were identified and retrieved
correctly during immediate cued recall. The search and naming
in processing information, while immediate cued recall of each
item ensured that it had been correctly encoded.
The search phase was followed by three trials of free recall, each
trial being preceded by 20 s of interference. On each such trial,
previously memorized items as they could. Next, a category cue
was provided by the examiner for each item missed on that trial. If
the subject still failed to recall the item with the cue, the examiner
reminded the participant of the missed item, which he/she then
later to examine delayed recall. Recall following a delay has been
demonstrated to be sensitive to hippocampal dysfunction (Squire
and Zola-Morgan, 1991; Miller et al., 1993) and to be one of the
best discriminators of very mild AD cases (Welsh et al., 1991;
Memory for the spatial location of verbally identified items in
this task was evaluated immediately after the third trial of free
of the 16 pictures was presented one at a time and the subject was
asked to identify in which of the four quadrants it originally ap-
peared (see Fig. 1B). Scoring for both the verbal and nonverbal
verbal material, data reported here are based on scores for the third
delayed free recall. It should be noted that scores for the third trial
of free recall are not equivalent to those of immediate recall, since
there was some delay between the acquisition of information and
Mean normalized hippocampal formation, parahippocampal
sphere. Volumes of these three regions of interest in patients and
analyses of variance (ANOVA), with groups (AD vs. control) and
hemispheres (right vs. left) as the two factors. The two groups
differed significantly in hippocampal volume only (F(1,46) ?
11.48, P ? 0.0015). The effect of hemisphere or the interaction
between hemisphere and groups did not reach significance. Even
when data for the 18 patients were examined separately, the vol-
umes of the right and left hippocampal formation did not signifi-
cantly differ from each other (t ? 0.624, df ? 17, P ? 0.541).
spatial location of verbal items for controls and AD cases are pre-
sented in Table 2. Separate two-way repeated measures ANOVAs
trolled learning task that subjects had to learn and remember. B: The
card with labeled quadrants that was used to test recall of the spatial
location of verbally learned items.
A: An example of four items from the Buschke con-
DE TOLEDO-MORRELL ET AL.
in the recall of verbal as well as spatial information (F(1,46) ?
ly). The effect of delay, however, was significant only for verbal
recall (F(1,46) ? 6.57, P ? 0.014), while the interaction between
groups and delay did not reach significance in either case. The lack
of an effect of delay for spatial recall found in our study is in line
with previous reports (Smith and Milner, 1981).
mance was examined using separate stepwise forward regression
analyses with age and normalized right hippocampal, left hip-
pocampal, parahippocampal gyrus, and temporal lobe volumes as
the independent variables. These analyses demonstrated that in
patients with AD, left hippocampal volume was the best predictor
of free verbal recall and delayed free verbal recall (R?0.637, P ?
0.0045, and R ? 0.799, P ? 0.0001, respectively; see Fig. 2).
Recall and delayed recall of the spatial location of verbal items, on
the other hand, were best predicted by right hippocampal volume
(R ? 0.627, P ? 0.0054, and R ? 0.579, P ? 0.0118, respec-
tively; see Fig. 3). Temporal lobe or parahippocampal gyrus vol-
ume did not enter the model. In addition, Pearson’s product mo-
ment correlations were performed to further assess the relation
between material-specific memory scores and right or left parahip-
pocampal gyrus or temporal lobe volumes; none were found to be
Results for the relationship between each region of interest and
material-specific memory performance in patients with AD are
presented in Table 3. It should be noted that the relation between
memory performance and hippocampal volume was significant in
patients, but not in aged controls (data not shown). In fact, there
was no significant correlation between any region of interest and
memory performance in the controls.
The data presented here confirm and extend previous reports
showing that hippocampal atrophy, as measured in vivo by quan-
titative MRI, is related to memory dysfunction in AD. Our results
pal atrophy and the degree of memory impairment in AD cases.
Most important is the fact that the strongest relation observed was
material-specific; left hippocampal volume was found to be the
best predictor of verbal recall, and right hippocampal volume of
spatial recall. Lesion and imaging studies have demonstrated that
verbal information processing, while the right processes nonverbal
or spatial information (Smith and Milner, 1981; Jones-Gotman,
1986; Maguire et al., 1997; Abrahams et al., 1997). Our results
agree well with these previous findings and provide a clear-cut
demonstration of hemispheric specialization of memory function
in patients with Alzheimer’s disease.
relations in AD, but these were not specific to the hippocampal
formation. Using the Warrington Recognition Memory Test
right temporal gray matter volume. Thus, in their hands, the hip-
Normalized Hippocampal Formation, Parahippocampal
Gyrus, and Temporal Lobe Volumes for Elderly Control
Subjects and Patients With Alzheimer’s Disease*
A. Normalized hippocampal volume
B. Normalized parahippocampal gyrus volume
C. Normalized temporal lobe volume
*AD, Alzheimer’s disease.
Percent Correct Recall on Verbal and Spatial Versions of the
Buschke “Controlled Learning” Task for Healthy Aged
Control Subjects and Patients With Alzheimer’s Disease
A. Healthy aged control subjects
B. Patients with Alzheimer’s disease
MEMORY PERFORMANCE IN AD
however, was not related to word recognition when the entire
group of patients was considered.
In our study, temporal lobe or parahippocampal gyrus vol-
ume did not significantly relate to any measure of verbal or
spatial recall, while hippocampal volume did. These dissocia-
tions strongly suggest that our findings cannot be attributed to
nonspecific generalized atrophy in AD. Although evidence in-
dicating the involvement of entorhinal and perirhinal cortices
in memory function is accumulating, the lack of a relation
between parahippocampal gyrus volume and recall of verbal or
spatial material in our hands is not surprising. Our measure of
delayed free recall (right) of verbal information.
Relationship between left hippocampal volume and free recall (left) as well as
(right) of the spatial location of verbal items.
Relationship between right hippocampal volume and recall (left) or delayed recall
DE TOLEDO-MORRELL ET AL.
parahippocampal gyrus volume included both white and gray
matter and went beyond the anatomical boundaries of the en-
torhinal and perirhinal cortices.
In summary, these results demonstrate a clear-cut and highly
significant relation between hippocampal volume and the recall of
material specific information in AD patients. In previous research,
hippocampal atrophy and performance on the Buschke controlled
learning task were each shown separately to be sensitive discrimi-
al., 1997; Jack et al., 1997). Our report extends these findings by
in the Buschke task are highly correlated in the same group of
We thank Drs. Carol Barnes and Irina Goncharova for helpful
comments on earlier versions of this paper.
Abrahams S, Pickering A, Polkey CE, Morris RG. 1997. Spatial memory
deficits in patients with unilateral damage to the right hippocampal
formation. Neuropsychologia 35:11–24.
disease. Proc Natl Acad Sci USA 93:13547–13551.
Buschke H, Grober E. 1986. Genuine memory deficits in age-associated
memory impairment. Dev Neuropsychol 2:287–307.
Cahn AD, Sullivan EV, Shear PK, Marsh L, Fama R, Lim KO, Yesavage
JA, Tinklenberg JR, Pfefferbaum A. 1998. Structural MRI correlates
de Leon MJ, George AE, Golomb A, Tarshish C, Convit A, Cluger A, De
M, Quinn B, Milled DC, Wishniewski H. 1997. Frequency of hip-
pocampal formation atrophy in normal aging and Alzheimer’s disease.
Neurobiol Aging 18:1–11.
de Toledo-Morrell L, Sullivan MP, Morrell F, Wilson RS, Bennett DA,
Spencer S. 1997. Alzheimer’s disease: in vivo detection of differential
vulnerability of brain regions. Neurobiol Aging 18:463–468.
Deweer B, Lehericy S, Pillon B, Baulac M, Chiras J, Marsault C, Agid Y,
Dubois B. 1995. Memory disorders in probable Alzheimer’s disease:
the role of hippocampal atrophy as shown with MRI. J Neurol Neu-
rosurg Psychiatry 58:590–597.
Folstein MF, Folstein SE, McHugh PR. 1975. “Mini-Mental State”: a
practical method for grading the mental status of patients for the
clinician. J Psychiatr Res 12:189–198.
Grober E, Buschke H. 1987. Genuine memory deficits in dementia. Dev
Jack CR, Petersen RC, O’Brian PC, Tangalos EG. 1992. MR-based hip-
pocampal volumetry in the diagnosis of Alzheimer’s disease. Neurol-
Jack CR, Petersen RC, Xu YC, Waring SC, O’Brien PC, Tangalos EG,
Smith GE, Ivnik RJ, Kokmen E. 1997. Medial temporal atrophy on
MRI in normal aging and very mild Alzheimer’s disease. Neurology
Jobst KA, Smith AD, Szatmari M, Molyneaux AJ, Esiri ME, King E,
Smith A, Jaskowski A, McDonald B, Wald N. 1992. Detection in life
of confirmed Alzheimer’s disease using a simple measurement of me-
dial temporal lobe atrophy by computed tomography. Lancet 340:
Jobst KA, Smith AD, Szatmari M, Esiri MM, Jaskowski A, Hindley H,
McDonald B, Molyneaux AJ. 1994. Rapidly progressing atrophy
of medial temporal lobe in Alzheimer’s disease. Lancet 343:
Jones-Gotman M. 1986. Right hippocampal excision impairs learning
and recall of a list of abstract designs. Neuropsychologia 24:659–670.
Kesslak JP, Nalcioglu O, Cotman CW. 1991. Quantification of magnetic
resonance scans for hippocampal and parahippocampal atrophy in
Alzheimer’s disease. Neurology 41:51–54.
Killiany RJ, Moss MB, Albert MS, Sandor T, Tieman J, Jolesz F. 1993.
Temporal lobe regions on magnetic resonance imaging identify pa-
tients with early Alzheimer’s disease. Arch Neurol 50:949–954.
Moscovitch M, Winocur G, Szalai JP, Bronskill MJ. 1998. Memory
impairments associated with hippocampal versus parahippocampal-
gyrus atrophy: an MR volumetry study in Alzheimer’s disease. Neuro-
Maguire EA, Frackowiak RSJ, Frith CD. 1997. Recalling routes around
McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan
EM. 1984. Clinical diagnosis of Alzheimer’s disease: report of the
NINCDS/ADRDA Work Group under the auspices of Department
of Health and Human Services Task Force on Alzheimer’s Disease.
Miller LA, Munoz DG, Finmore M. 1993. Hippocampal sclerosis and
human memory. Arch Neurol 50:391–394.
Medial temporal structures relate to memory impairment in Alzhei-
mer’s disease: an MRI volumetric study. J Neurol Neurosurg Psychi-
Relation Between Normalized Hippocampal Formation,
Parahippocampal Gyrus, and Temporal Lobe Volumes, and
Material-Specific Memory Performance in Patients With
recall Spatial recall
Right HFR ? 0.490
P ? 0.0389
R ? 0.637
P ? 0.0045
R ? 0.171
R ? 0.254
R ? ?0.145
R ? ?0.237
R ? 0.670
P ? 0.0023
R ? 0.799
P ? 0.00007
R ? 0.364
R ? 0.318
R ? 0.035
R ? ?0.157
R ? 0.612
P ? 0.007
R ? 0.510
P ? 0.031
R ? 0.427
R ? 0.325
R ? ?0.018
R ? ?0.0327
R ? 0.555
P ? 0.0167
R ? 0.510
P ? 0.0304
R ? 0.278
R ? 0.282
R ? ?0.233
R ? ?0.157
*HF, hippocampal formation; PHG, parahippocampal gyrus; TL, tem-
poral lobe; n.s., no significance.
MEMORY PERFORMANCE IN AD
MorrisJC,HeymanA,MohsRC,HughesJP,vanBelleG,FillenbaumG, Download full-text
Alzheimer’s Disease (CERAD). Part I. Clinical and neuropsychologi-
cal assessment of Alzheimer’s disease. Neurology 39:1159–1165.
ory function in very early Alzheimer’s disease. Neurology 44:867–872.
Seab JP, Jagust WJ, Wong ST, Roos MS, Reed BR, Budinger TF. 1988.
Quantitative NMR measurements of hippocampal atrophy in Alzhei-
mer’s disease. Magn Reson Med 8:200–208.
Smith ML, Milner B. 1981. The role of the right hippocampus in the
recall of spatial location. Neuropsychologia 19:781–793.
Squire LR, Zola-Morgan S. 1991. The medial temporal lobe memory
system. Science 253:1380–1386.
Warrington E. Recognition memory test manual. Windsor, UK: Nelson
Publishing Co. 1984.
Welsh K, Butters N, Hughes JP, Mohs RC, Heyman A. 1991. Detection
CERAD neuropsychological measures. Arch Neurol 48:278–281.
Morrell F. 1996. Association of memory and cognition in Alzheimer’s
disease wirth volumetric estimates of temporal lobe structures. Neuro-
DE TOLEDO-MORRELL ET AL.