Neuroanatomical and neuropsychological features of elderly euthymic depressed patients with early- and late-onset.
ABSTRACT Whether or not cognitive impairment and brain structure changes are trait characteristics of late-life depression is still disputed. Previous studies led to conflicting data possibly because of the difference in the age of depression onset. In fact, several lines of evidence suggest that late-onset depression (LOD) is more frequently associated with neuropsychological deficits and brain pathology than early-onset depression (EOD). To date, no study explored concomitantly the cognitive profile and brain magnetic resonance imaging (MRI) patterns in euthymic EOD and LOD patients.
Using a cross-sectional design, 41 remitted outpatients (30 with EOD and 11 with LOD) were compared to 30 healthy controls. Neuropsychological evaluation concerned working memory, episodic memory, processing speed, naming capacity and executive functions. Volumetric estimates of the amygdala, hippocampus, entorhinal and anterior cingulate cortex were obtained using both voxel-based and region of interest morphometric methods. White matter hyperintensities were assessed semiquantitatively.
Both cognitive performance and brain volumes were preserved in euthymic EOD patients whereas LOD patients showed a significant reduction of episodic memory capacity and a higher rate of periventricular hyperintensities compared to both controls and EOD patients.
Our results support the dissociation between EOD thought to be mainly related to psychosocial factors and LOD that is characterized by increasing vascular burden and episodic memory decline.
- SourceAvailable from: arxiv.org[show abstract] [hide abstract]
ABSTRACT: A hydrophobic constriction site can act as an efficient barrier to ion and water permeation if its diameter is less than the diameter of an ion's first hydration shell. This hydrophobic gating mechanism is thought to operate in a number of ion channels, e.g. the nicotinic receptor, bacterial mechanosensitive channels (MscL and MscS) and perhaps in some potassium channels (e.g. KcsA, MthK and KvAP). Simplified pore models allow one to investigate the primary characteristics of a conduction pathway, namely its geometry (shape, pore length, and radius), the chemical character of the pore wall surface, and its local flexibility and surface roughness. Our extended (about 0.1 micros) molecular dynamic simulations show that a short hydrophobic pore is closed to water for radii smaller than 0.45 nm. By increasing the polarity of the pore wall (and thus reducing its hydrophobicity) the transition radius can be decreased until for hydrophilic pores liquid water is stable down to a radius comparable to a water molecule's radius. Ions behave similarly but the transition from conducting to non-conducting pores is even steeper and occurs at a radius of 0.65 nm for hydrophobic pores. The presence of water vapour in a constriction zone indicates a barrier for ion permeation. A thermodynamic model can explain the behaviour of water in nanopores in terms of the surface tensions, which leads to a simple measure of 'hydrophobicity' in this context. Furthermore, increased local flexibility decreases the permeability of polar species. An increase in temperature has the same effect, and we hypothesize that both effects can be explained by a decrease in the effective solvent-surface attraction which in turn leads to an increase in the solvent-wall surface free energy.Physical Biology 07/2004; 1(1-2):42-52. · 2.62 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: We have used the patch-clamp electrical recording technique on giant spheroplasts of Escherichia coli and have discovered pressure-activated ion channels. The channels have the following properties: activation by slight positive or negative pressure; voltage dependence; large conductance; selectivity for anions over cations; dependence of activity on the species of permeant ions. We believe that these channels may be involved in bacterial osmoregulation and osmotaxis.Proceedings of the National Academy of Sciences 05/1987; 84(8):2297-301. · 9.74 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Human red cell AQP1 is the first functionally defined member of the aquaporin family of membrane water channels. Here we describe an atomic model of AQP1 at 3.8A resolution from electron crystallographic data. Multiple highly conserved amino-acid residues stabilize the novel fold of AQP1. The aqueous pathway is lined with conserved hydrophobic residues that permit rapid water transport, whereas the water selectivity is due to a constriction of the pore diameter to about 3 A over a span of one residue. The atomic model provides a possible molecular explanation to a longstanding puzzle in physiology-how membranes can be freely permeable to water but impermeable to protons.Nature 11/2000; 407(6804):599-605. · 38.60 Impact Factor
Neuroanatomical and neuropsychological features of elderly euthymic depressed
patients with early- and late-onset
Christophe Delaloyea,b,⁎, Guenaël Moya, Fabienne de Bilbaoa, Sandra Baudoisa, Kerstin Webera,
Françoise Hofera, Claire Ragno Paquiera, Alessia Donatic, Alessandra Canutoa, Umberto Giardinia,
Armin von Guntenc, Raluca Iona Stancuc, François Lazeyrasd, Philippe Millete, Philip Scheltensf,
Panteleimon Giannakopoulosa,c, Gabriel Goldg
aDivision of Geriatric Psychiatry, University Hospitals of Geneva and Faculty of Medicine, Switzerland
bFaculty of Psychology and Science of Education, University of Geneva, Switzerland
cDivision of Old Age Psychiatry, University Hospitals of Lausanne, Hospices-CHUV, Switzerland
dDepartment of Radiology, University Hospitals of Geneva, Switzerland
eClinical Neurophysiology and Neuroimaging Unit, University Hospitals of Geneva, Switzerland
fAlzheimer Center and Department of Neurology, VU Medical Center, Amsterdam, The Netherlands
gDepartment of Rehabilitation and Geriatrics, University Hospitals of Geneva, Switzerland
a b s t r a c ta r t i c l ei n f o
Received 8 March 2010
Accepted 25 August 2010
Available online 17 September 2010
Background: Whether or not cognitive impairment and brain structure changes are trait characteristics of
late-life depression is still disputed. Previous studies led to conflicting data possibly because of the difference
in the age of depression onset. In fact, several lines of evidence suggest that late-onset depression (LOD) is
more frequently associated with neuropsychological deficits and brain pathology than early-onset
depression (EOD). To date, no study explored concomitantly the cognitive profile and brain magnetic
resonance imaging (MRI) patterns in euthymic EOD and LOD patients.
Method: Using a cross-sectional design, 41 remitted outpatients (30 with EOD and 11 with LOD) were
compared to 30 healthy controls. Neuropsychological evaluation concerned working memory, episodic
memory, processing speed, naming capacity and executive functions. Volumetric estimates of the amygdala,
hippocampus, entorhinal and anterior cingulate cortex were obtained using both voxel-based and region of
interest morphometric methods. White matter hyperintensities were assessed semiquantitatively.
Results: Both cognitive performance and brain volumes were preserved in euthymic EOD patients whereas
LOD patients showed a significant reduction of episodic memory capacity and a higher rate of periventricular
hyperintensities compared to both controls and EOD patients.
Conclusion: Our results support the dissociation between EOD thought to be mainly related to psychosocial
factors and LOD that is characterized by increasing vascular burden and episodic memory decline.
© 2010 Elsevier B.V. All rights reserved.
Amongst patients who suffer from late-life depression, 30% to 50%
than in younger cohorts , some cognitive deficits in old age may
represent trait characteristics of depression that persist despite the
amendment of symptoms . However, this viewpoint has been
challenged by prospective studies showing that depression at baseline
is not necessarily associated with an increased risk of subsequent
cognitivedecline. These discrepancies might be partlyexplained by
age differences relative to the onset of depression among elderly
patients. Recent studies suggest that the patterns of both neuropsy-
chological deficits and structural imaging changes vary substantially
between late-onset depression (LOD) and early-onset depression
(EOD) [1,5,6]. LOD has been traditionally associated with more
frequent and rapid cognitive decline as well as more severe structural
brain abnormalities compared to age-matched controls and EOD
patients [1,6]. In particular, literature reported executive dysfunction
to be a characteristic of LOD whereas episodic memory dysfunction is
present both in LOD and EOD . Consistent with the presence of
dysexecutive impairment in LOD, neuroimaging studies  revealed a
frontostriatal disruption caused by subcortical, white matter and
periventricular hyperintensities. In contrast, EOD cases display
Journal of the Neurological Sciences 299 (2010) 19–23
⁎ Corresponding author. Service de Psychiatrie Gériatrique, Hôpitaux Universitaires
de Genève, 2 Chemin du Petit-Bel-Air, 1225 Chêne-Bourg, Switzerland. Tel.: +41 22 305
51 30; fax: +41 22 305 50 44.
E-mail address: Christophe.Delaloye@hcuge.ch (C. Delaloye).
0022-510X/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
Contents lists available at ScienceDirect
Journal of the Neurological Sciences
journal homepage: www.elsevier.com/locate/jns
There are two main methodological limitations of these studies.
Most of them explored separately structural changes and neuropsy-
chological performances. Most importantly, they included acutely
depressed patients and did not provide further insight into their
persistent abnormalities. To address these issues, we performed a
prospective group study (EOD, LOD and controls) including a detailed
neuropsychological assessment and MRI investigation.
Thirty EOD patients (depression onset before 60), eleven LOD
patients (depression onset after 60), and thirty healthy controls were
included. The diagnosis of depression or the absence of a psychiatric
condition was established using the Mini International Neuropsychi-
atric Inventory Interview. Euthymia was defined according to DSM-IV
criteria, namely the absence of depressive symptoms for at least two
months. In addition, all participants had to obtain a score below 5 on
the 15 item Geriatric Depression Scale (GDS) at inclusion. Exclusion
criteria for all groups were: 1) history of major neurological disorders,
2) the presence of a current or a past DSM-IV psychiatric diagnosis
(other than depression), 3) the presence of dementia, and 4) current
systemic medical disease requiring inpatient treatment. Physical
health status was examined with the Charlson Comorbidity Index
(CCI). Following the formal acceptance of the research protocol by the
local ethics committee, written informed consent was obtained from
2.2. Cognitive measures
2.2.1. Processing speed
(I) In a simple reaction time test, participants had to press a key as
quickly as possible when the visual signal stimulus appeared. The
score of interest was the mean latency of the 120 test trials. (II) In the
Color Trail Making Test part A, participants had to connect a series of
25 circled numbers in numerical order as quickly as possible. Time to
complete the task was taken into consideration.
2.2.2. Working memory
(I) In a computerized reading span test, participants had to
complete a single and a dual condition. In the single condition,
participants had to decide whether each one of the 16 sentences,
presented sequentially, was semantically correct as fast and accu-
rately as possible. In the dual condition, the participants additionally
had to memorize the final word of each sentence. Sixteen series, from
two to five sentences, were presented. Recall took place at the end of
each series. The retained score was the mean number of correctly
recalled words; (II) In the Letter-number sequence task, participants
had to listen to letters and numbers and then recall all the numbers in
ascending order, followed by all the letters in alphabetical order. The
score of interest was the number of correct sequences produced.
2.2.3. Executive function
(I) In the Stroop color, participants had to complete two subtests.
The first subtask displayed solid color patches in one of four colors.
The second subtask contained color words printed in an incongruous
ink color. In each subtask, participants had to name the ink color of
the stimuli as quickly and accurately as possible. Completion time
for each subtask was the dependent measure. The score of interest
was the interference effect which was computed by a relative ratio
[(subtask 2−subtask 1)/subtask 1)]; (II) In the Color-trail making
test part B (for part A see Processing speed section), participants had
to shift sets by connecting numbers in an increasing order while
alternating color-patterns. The time needed to complete each part
was the dependent measure. The score of interest was the flexibility
cost [(part B−part A)/part A].
2.2.4. Episodic memory
(I) In the Cued recall 48 items test (CR 48), participants had to
learn 48 different words, belonging to 12 different semantic
categories, with the help of semantic cues. A cued recall task, using
semantic categories as a cue, was performed. The score of interest was
the number of words correctly recalled; (II) In the CERAD Word List
Memory test, participants had to memorize a list of ten words that
was presented over three trials. The total number of words recalled
correctly across the three trials was the dependant measure.
2.3. MRI procedures
MRI imaging was performed at 3 T. Coronal slices were obtained
from 3-dimensional MPRAGE sequence with the following para-
meters: TR 2500 ms, TE 2.94 ms, TI 1100 ms, flip angle 9°, isotropic
resolution of 0.9 mm3, acquisition time 520 s. In addition, 3-
dimensional T2 weighted imaging was obtained with the following
parameters: TE=383 ms, TR=3200 ms, FOV=230 mm, acceleration
factor (parallel imaging) 2, matrix size 256×256×240.
2.4. Assessment of WMH
Assessment of whitematterlesions wasperformedin T2-weighted
sequences with the Scheltens semiquantitative scale .
2.5. ROI analysis
Anterior cingular and entorhinal cortices, as well as hippocampal
and amygdala perimeters were traced manually on each contiguous
coronal slice using a ROI procedure of ANALYZE software (version 8,
Mayo Foundation). Neuroanatomic boundaries of the hippocampus
and amygdala were based on those of Watson et al. . Anatomic
guidelines for outlining the entorhinal and anterior cingular cortices
were those described by Bernasconi et al.  and Sassi et al. 
respectively. Normalized volumes for brain regions of interest were
determined by using the following formula: [absolute volume in
mm3/intracranial volume (ICV) in mm3]×1.000.
2.6. Voxel-based morphometry
Standard processing, using SPM5 software, was employed to
analyze MRI for voxel-based morphometry (VBM) . Images were
segmented using the standard T1 template and a priori gray matter,
white matter and CSF atlases provided by SPM. Spatially normalized
(1×1×1 mm3) data were modulated to account for local volume
changes due to non linear co-registration. Gray matter images were
smoothed with a 8-mm Gaussian kernel.
2.7. Statistical analysis
Comparisons between groups for continuous variables were
performed using a one-way independent analysis of variance. To
respect assumptions of normality and homogeneity of variance data
were modified prior to analysis by means of logarithmic or power 2
were used. Comparisons of categorical variables were performed with
a Pearson's Chi-Square. Linear regression models were also built with
cognitive parameters or vascular lesions as the dependent variables
and diagnostic group and age as the independent variables.
C. Delaloye et al. / Journal of the Neurological Sciences 299 (2010) 19–23
3.1. Demographics and clinical characteristics
The demographic and clinical characteristics of the series are
summarized in Table 1. Among the three groups there were no
significant differences in gender, education and somatic comorbidity
as assessed by the CCI. EOD patients were significantly younger than
controls (U=205.50, pb0.001) and LOD patients (U=22.00,
pb0.001). Patients were already under treatment at inclusion. We
did not interfere with their medication as the treatment was
prescribed naturalistically. 73% of LOD patients and 50% of EOD
patients received regular antidepressant medication (selective sero-
EOD cases and 27% of the LOD cases with benzodiazepines andin 3% of
the EOD cases and 18% of the LOD cases with atypical antipsychotics.
Only one LOD case received concomitantly antidepressants, anti-
psychotics, and benzodiazepines.
3.2. MRI data
Table 2 summarizes the mean rating score of WMH according to
the Scheltens semiquantitative scale. Periventricular hyperintensities
score differed among our three groups (H(2)=13.00, pb0.01) and
were significantly higher in LOD patients compared to controls
(U=58.50, pb0.001, r=0.51) and EOD patients (U=57.00, pb0.001,
r=0.52). The severity of deep white matter and basal ganglia
hyperintensities was comparable between the three groups. Except
2 LOD cases, there were no hyperintensities in infratentorial area in
the present series.
The mean normalized volumes of each brain ROI by group of
participants are reported in Table 2. There were no significant
differences in entorhinal cortex and amygdala volume between our
three groups. There was a trend for reduced hippocampal (F (2, 68)=
2.97, p=0.06) and anterior cingulate (F (2, 68)=3.03, p=0.06)
volume in LOD patients. However, this trend did not persist after
adjustment for age in linear regression models. The VBM analysis
confirmed the absence of statically significant difference in gray
matter volumes between our three groups.
3.3. Cognitive test performance
Performances on cognitive tests are provided in Table 3. With
respect to processing speed, the simple reaction time test and the part
A of the color trail making test revealed no significant differences
among our three groups. This was also the case for working memory
assessed by the mean number of words correctly recalled in the
reading span task and the score of the sequence-letter sequence test.
Executive functions were also preserved in EOD and LOD patients.
The relative ratios measuring interference cost induced by the Stroop
effect and switching cost evaluated by the color trail making test were
not significantly different compared to controls.
For episodic memory, a significant group difference was observed
for the cued recall score of the CR 48 items test (F (2, 68)=4.16,
pb0.05) as well as for the total recall score of the CERAD word list
memory test (F (2, 68)=8.26, pb0.01). Post-hoc analyses using the
Bonferroni post hoc criterion for significance indicated that LOD
patients (M=17.73, 95% CI [14.72, 20.73]) obtained a significantly
lowertotalrecall score ontheCERAD wordlist memory testcompared
to controls (M=22.17, 95% CI [20.77, 23.56], pb0.01, r=0.46) and
EOD patients (M=22.70, 95% CI [21.59, 23.81], pb0.01, r=0.55).
Similarly, LOD patients (M=23.55, 95% CI [18.44, 28.65]) obtained a
significantly lower cued recall score on the CR 48 items test compared
to controls (M=29.07, 95% CI [26.96, 31.18], pb0.05, r=0.37) and
EOD patients (M=29.10, 95% CI [26.78, 29.67], pb0.05, r=0.39).
Importantly, linear regression analysis revealed that these differences
were still present after adjustment for age. Among EOD patients,
duration of illness had no impact on ROI volumes, WMH scores and
cognition after control for age.
The preservation of both cognition and brain structures in
euthymic EOD patients, in particular episodic memory capacity and
hippocampal volume, as well as the absence of deleterious effect of
the duration of illness does not support the hypothesis of a
progressive neurotoxic effect of depression . Our results parallel
several lines of evidence supporting the preservation of cognitive
abilities in elderly patients with EOD after remission [4,6]. In
particular, Brodaty et al.  found no evidence for long-term cognitive
deficits following depressive episodes even after 25 years of follow-
up. It is worth mentioning that in the present study we carefully
excluded all lifetime psychiatric comorbidities such as substance
abuse, known to influence neuropsychological performances and
hippocampal volumes . This might explain the difference between
our findings and those of certain previous studies in this field [3,8].
Unlike EOD, LOD patients showed decreased performances limited
to episodic memory tasks. Although one could argue that this finding
may be primarily due to confounding factors such as medication, this
is an unlikely scenario since we indexed episodic memory capacity by
Demographic and clinical characteristics.
CharacteristicsM (SD)M (SD)M(SD)
Age in years
Score GDS (max 15)b
Score CCI (max 19)c
Age at depression onset (years)
Duration of depressive
Gender (% women)
aNumber of years of education completed.
bGeriatric Depression Scale.
cCharlson Comorbidity Index.
White matter hyperintensity scores for EOD patients, LOD patients and controls.
M (SD)M (SD)M(SD)
Sheltens' scale scores
WMH in deep
0.83(1.05)0.80(0.96) 2.45 (1.44)13.000.002
2.87(4.99)3.23(4.13)6.45(6.30) 5.39 0.068
Mean normalized volumes of the brain region of interestb
Entorhinal cortex Total1.17
(0.26)1.69(0.25)1.63 (0.18) 0.340.519
aStatistical comparisons were made using Kruskal–Wallis test/analysis of variance
bMean normalized volumes: [(absolute volume of ROI in mm3/intracranial volume
C. Delaloye et al. / Journal of the Neurological Sciences 299 (2010) 19–23
using the RI-48 item test that controls for cognitive processing during
encoding. In contrast to free recall tests, this procedure involves cued
recall and minimises the effect of impaired attention, inefficient
strategies or reduced processing capacity that could result from these
factors . Moreover, the episodic memory impairment cannot be
attributed to the co-existenceof dementiasince we carefullyexcluded
all of these cases. In line with our results, two recent studies found a
reduction of episodic memory capacity in LOD patients compared to
controlsandEODpatients[2,16].Moreover, Thomaset al. observed
that none of their LOD patients developed dementia during a four-
year follow-up suggesting that incipient Alzheimer's disease is not the
explanation ofthelargerverbalmemoryimpairmentidentifiedin LOD
group. The significant increase of periventricular hyperintensities in
our LOD group may partly explain the observed cognitive deficits. In
fact, recent contributions pointed to a possible vascular origin of
episodic memory impairment both in population-based samples and
LOD cohorts [5,17].
To our knowledge, this is the only study that compared
concomitantly the cognitive profile and the brain structural char-
acteristics of elderly patients who recovered from EOD or LOD.
Additional strengths of the present study include the careful exclusion
of lifetime psychiatric comorbidities as well as physical burden,
control for demographic variables and duration of depression,
detailed assessment of cognitive performances, volumetric analyses
as well as assessment of WMH. Two limitations should also be taken
into account. First, we cannot exclude that the lack of significant
between-group differences on some measures may be due to an
insufficient sample size. Second, the effect of additional clinical
parameters such as the number of previous episodes and history of
psychotic symptoms was not addressed.
In conclusion, our observations support the distinction between
the aetiological mechanisms implicated in the pathogenesis of LOD
and EOD . While LOD may be more driven by acquired pathology
such as vascular burden, genetic background and personality dimen-
sions may be the most important determinants of EOD.
5. Conflict of interest
This research was supported by the Swiss National Science
Foundation (SNSF grant no 3200BO-112018). The SNSF had no further
role in study design; in collection, analysis and interpretation of the
data; in the writing of the report; and in the decision to submit the
paper for publication. All authors declare that they have no conflict of
The authors thank Abba Moussa, Corina Meiler-Mititelu, Karsten
Ebbing, Montserrat Mendez Rubio, Françoise Lanet and Reto Meuli for
their contribution to this work. This work was also supported by the
Center for Biomedical Imaging (CIBM) of the Geneva-Lausanne
Universities and the EPFL.
 Herrmann LL, Goodwin GM, Ebmeier KP. The cognitive neuropsychology of
depression in the elderly. Psychol Med 2007;37(12):1693–702.
 Thomas AJ, Gallagher P, Robinson LJ, Porter RJ, Young AH, Ferrier IN, et al. A
comparison of neurocognitive impairment in younger and older adults with major
depression. Psychol Med 2009;39(5):725–33.
 Bhalla RK, Butters MA, Mulsant BH, Begley AE, Zmuda MD, Schoderbek B, et al.
Persistence of neuropsychologic deficits in the remitted state of late-life
depression. Am J Geriatr Psychiatry 2006;14(5):419–27.
 Brodaty H, Luscombe G, Anstey KJ, Cramsie J, Andrews G, Peisah C. Neuropsycho-
logical performance and dementia in depressed patients after 25-year follow-up: a
controlled study. Psychol Med 2003;33(7):1263–75.
 Takahashi K, Oshima A, Ida I, Kumano H, Yuuki N, Fukuda M, et al. Relationship
between age at onset and magnetic resonance image-defined hyperintensities in
mood disorders. J Psychiatr Res 2008;42(6):443–50.
 Kohler S, Thomas AJ, Barnett NA, O'Brien JT. The pattern and course of cognitive
impairment in late-life depression. Psychol Med 2009:1–12.
 Herrmann LL, Le Masurier M, Ebmeier KP. White matter hyperintensities in late
life depression: a systematic review. J Neurol Neurosurg Psychiatry 2008;79(6):
 Campbell S, Marriott M, Nahmias C, MacQueen GM. Lower hippocampal volume in
patients suffering from depression: a meta-analysis. Am J Psychiatry 2004;161(4):
 Scheltens P, Barkhof F, Leys D, Pruvo JP, Nauta JJ, Vermersch P, et al. A
semiquantative rating scale for the assessment of signal hyperintensities on
magnetic resonance imaging. J Neurol Sci 1993;114(1):7–12.
 Watson C, Andermann F, Gloor P, Jones-Gotman M, Peters T, Evans A, et al.
Anatomic basis of amygdaloid and hippocampal volume measurement by
magnetic resonance imaging. Neurology 1992;42(9):1743–50.
 Bernasconi N, Bernasconi A, Andermann F, Dubeau F, Feindel W, Reutens DC.
Entorhinal cortex in temporal lobe epilepsy: a quantitative MRI study. Neurology
 Sassi RB, Brambilla P, Hatch JP, Nicoletti MA, Mallinger AG, Frank E, et al. Reduced
left anterior cingulate volumes in untreated bipolar patients. Biol Psychiatry
Neuropsychological performances in EOD patients, LOD patients and controls.
Cognitive function TaskM(SD)M (SD)M (SD)
Processing speedSimple reaction timeb
Latencies in ms
CTM part 1b,c
Latencies in sec.
Cued recall 48
Cerad 10 words:
Total recall score
336.92 (74.77)299.72 (48.48)312.00 (54.22)2.260.112
Working memory2.27 (0.54)2.40 (0.59)2.04 (0.38)1.770.178
10.40 (2.72)9.90 (1.83)8.55(2.16)4.92 0.085
Executive function2.00 (0.55)2.45 (0.59)2.10 (0.55) 2.480.091
1.04 (0.59)0.91 (0.55)1.33 (0.66) 3.66 0.161
Episodic memory 29.07(5.65) 29.10(5.34)23.55 (7.59)4.16 0.020
22.17(3.74) 22.70(2.98) 17.73(4.47)8.26 0.001
aStatistical comparisons were made using Kruskal–Wallis test/Analysis of variance (ANOVA).
bData were transformed prior to analyses (logarithmic or power 2 transformation).
cCTM=Color Trail Making Test.
C. Delaloye et al. / Journal of the Neurological Sciences 299 (2010) 19–23
 Mechelli A, Price CJ, Friston KJ, Ashburner J. Voxel-based morphometry of the
human brain: methods and applications. Curr Med Imaging Rev 2005;1:105–13.
 Chanraud S, Reynaud M, Wessa M, Penttila J, Kostogianni N, Cachia A, et al.
Diffusion tensor tractography in mesencephalic bundles: relation to mental
flexibility in detoxified alcohol-dependent subjects. Neuropsychopharmacology
 Ivanoiu A, Adam S, Van der Linden M, Salmon E, Juillerat AC, Mulligan R, et al.
Memory evaluation with a new cued recall test in patients with mild cognitive
impairment and Alzheimer's disease. J Neurol 2005;252(1):47–55.
 Dillon C, Allegri RF, Serrano CM, Iturry M, Salgado P, Glaser FB, et al. Late- versus
early-onset geriatric depression in a memory research center. Neuropsychiatr Dis
 Potter GG, McQuoid DR, Steffens DC, Welsh-Bohmer KA, Krishnan KR. Neuropsy-
chological correlates of magnetic resonance imaging-defined subcortical ischemic
depression. Int J Geriatr Psychiatry 2009;24(3):219–25.
C. Delaloye et al. / Journal of the Neurological Sciences 299 (2010) 19–23