606987

research-article2015

MSJ0010.1177/1352458515606987Multiple Sclerosis JournalRiccelli et al.

MULTIPLE
SCLEROSISMSJ
JOURNAL

Original Research Paper

Individual differences in depression are

associated with abnormal function of the limbic
system in multiple sclerosis patients

Multiple Sclerosis Journal

112
DOI: 10.1177/
1352458515606987
The Author(s), 2015.
Reprints and permissions:
http://www.sagepub.co.uk/
journalsPermissions.nav

Roberta Riccelli, Luca Passamonti, Antonio Cerasa, Salvatore Nigro, Salvatore Maria Cavalli,
Carmelina Chiriaco, Paola Valentino, Rita Nistic and Aldo Quattrone

Abstract
Background: Depression is common in patients with multiple sclerosis (MS), although the brain mechanisms of this psychiatric condition in MS are poorly understood. Specifically, it remains to be determined
whether depression in MS is related to altered activity and functional connectivity patterns within limbic
circuits.
Methods: Seventy-seven MS patients with variable levels of depression (as assessed via the Beck
Depression Inventory) underwent functional magnetic resonance imaging while performing an
emotional processing task. To conduct the functional connectivity analyses, the bilateral amygdala
and hippocampus, two areas critically involved in the pathophysiology of depression, were chosen as
seed regions. Multiple regression models were used to assess how depression in MS patients was
correlated with the activity and functional connectivity patterns within the limbic system.
Results: Depression scores in MS patients were negatively correlated: (1) with the activity in the
subgenual cingulate cortex; (2) with the functional connectivity between the hippocampus and orbitofrontal
cortex as well as the dorsolateral prefrontal cortex, and (3) with the functional connectivity between the
amygdala and dorsolateral prefrontal cortex.
Conclusions: Our study showed that individual differences in depression in MS patients were significantly associated with altered regional activity and functional connectivity patterns within the limbic
system.

Correspondence to:
Luca Passamonti
Department of Clinical
Neurosciences, University of
Cambridge, Herchel Smith
Building, Robinson Way,
Cambridge, CB2 0SZ, UK.
lp337@medschl.cam.ac.uk

Keywords: Depression, multiple sclerosis, limbic system, prefrontal cortex, functional connectivity

Luca Passamonti
Institute of Bioimaging
and Molecular Physiology,
National Research Council,
Catanzaro, Italy/Department
of Clinical Neurosciences,
University of Cambridge, UK

Date received: 19 April 2015; accepted: 11 August 2015

Introduction
It has been estimated that a significant proportion of
patients with multiple sclerosis (MS) display variable
levels of depression during their life span,1 although
the brain mechanisms underlying depression in MS
remain unclear. Increasing our knowledge of the brain
basis of depression in MS is important as it can have
implications for developing novel diagnostic markers
in a common psychiatric condition in MS.
There is accumulating evidence that depression in
MS is mediated by specific neurobiological factors.
The first, although indirect, support for this notion
came from past clinical studies which found a weak

Roberta Riccelli
Salvatore Nigro
Paola Valentino
Salvatore Maria Cavalli
Department of Medical and
Surgical Sciences, University
Magna Graecia, Catanzaro,
Italy
Antonio Cerasa
Carmelina Chiriaco
Rita Nistic
Institute of Bioimaging
and Molecular Physiology,
National Research Council,
Catanzaro, Italy
Aldo Quattrone
Department of Medical and
Surgical Sciences, University
Magna Graecia, Catanzaro,
Italy/Institute of Bioimaging
and Molecular Physiology,
National Research Council,
Catanzaro, Italy

relationship between depression and MS diseases

severity, suggesting that depression in MS may not
be necessarily linked to neurological disability.2,3,4
Furthermore, peripheral immune markers (e.g. stimulated production of interferon-gamma by T-cells)
have been associated with mood disturbances
in MS.5 Earlier neuroimaging studies have also
revealed that focal damage in frontal and temporal
lobes may be important in the pathogenesis of
depression in MS,6,7,8 although they did not assess
whether changes in fronto-temporal connectivity
were linked to depression in MS. This is likely as MS
is a demyelinating disorder that typically affects
large-scale neural networks.9 Consistent with this

http://msj.sagepub.com 1

Downloaded from msj.sagepub.com at CILEA Bibliosan on October 9, 2015

Multiple Sclerosis Journal

hypothesis, our recent diffusion tensor imaging
(DTI) study has revealed abnormal structural connectomic measures within limbic fronto-temporal
circuits in MS patients with major depression relative
to MS patients with no depression.10 However, it
remains to be determined whether altered functional
connectivity patterns are also present over and above
the structural connectivity abnormalities that we have
recently observed in MS patients with depression.10
This is an important issue as functional connectivity
indices can offer, relative to structural measures,
different information to describe the brain dynamics
underlying depression in MS. There is also evidence
that functional magnetic resonance imaging (fMRI)
may be a sensible tool for characterizing the neuroplasticity mechanisms that have been consistently
reported in MS patients while executing a broad
range of fMRI tasks.11
To our knowledge, this is the first study which has
assessed changes in brain activity and functional connectivity patterns in a large and homogenous sample
of MS patients with variable levels of depression
while executing a face task, a paradigm with strong
ecological validity for probing brain emotional
regions like the amygdala, hippocampus and prefrontal cortex (PFC).12 Overall, we expected greater
amygdala and reduced hippocampal response to emotional faces in MS patients as a function of progressively increasing depression scores. This is because
the amygdala is known to promote the viscerohormonal and behavioural responses that characterize
depression,13 while the hippocampus is thought to
exert the opposite function (i.e. reduce the stress
response associated with depression).14
We also predicted that depression in MS may be
associated with abnormally enhanced activity in a
series of PFC regions that play a key role in the
pathophysiology of depression such as the orbitofrontal cortex (OFC), ventrolateral PFC (VLPFC),
dorsolateral PFC (DLPFC) and subgenual cingulate
cortex.15,16 This is because a number of past studies
have shown that MS patients, relative to controls,
exhibit greater activity in PFC regions while performing motor, cognitive and emotional tasks.17,18 This
consistent finding has been interpreted as reflecting
adaptive changes, that is, the brain would react to
MS-related pathological changes via compensatory
mechanisms that tend to limit the behavioural deficits despite the damage.18 Nonetheless, increased
brain activity is not expected when the task complexity and/or the burden of the disease increase to a certain level and when reduced or maladaptive plasticity
is associated with impaired function and exhaustion

of the functional reserve.19 Finally, we tested whether

depression in MS influenced the functional connectivity patterns across the above-mentioned areas (i.e.
amygdala, hippocampus and PFC areas). In this case,
depression levels in MS patients were expected to
correlate with altered connectivity patterns within
limbic PFC-medial temporal lobe circuits.20
Materials and methods
Participants
Patients were recruited from the Neurological Unit of
the University Magna Graecia of Catanzaro. We
enrolled 77 patients with clinically definite relapsing
remitting MS who met the following criteria: (1) no
history of traumatic brain injury, past or current history of substance abuse, or other coexisting medical
conditions; (2) no clinical relapses for at least six
months prior to study entry; (3) no assumption of steroids, or disease-modifying therapies in the three
months before recruitment; (4) Expanded Disability
Status Scale (EDSS) ranging from 0 to 4.5. This
EDSS range was chosen as it reflected a disease
severity level that was unlikely to prevent the correct
execution of the fMRI task. Forty-two MS patients
also participated in our recent DTI study.10
MS patients were either antidepressant medication
nave or had undergone a washout period of at least
five half-lives of a previously prescribed antidepressant medication. This procedure has been used in
previous studies to reduce the risk that the data are
contaminated with non-specific drugs effects.21,22
Depression levels were assessed using the Beck
Depression Inventory (BDI-II) and then BDI-Fast
Screen (BDI-FS) scores were extracted from the
total BDI-II scores as in previous research.23 This is
because some of the somatic symptoms of depression typically assessed via the BDI-II may overlap
with classic MS symptoms (e.g. fatigue) so the
BDI-FS is thought to represent a purer measure of
depression in MS than the BDI-II.24 MS patients
were also interviewed using the mood disorder section of the Structured Clinical Interview of the
Diagnostic and Statistical Manual of Mental
Disorders to evaluate the presence of major depression. Fatigue levels were assessed via the Fatigue
Severity Scale (FSS).
A group of n=20 healthy controls (with a normal MRI
brain scan) was included to test for group differences in
brain activations independent of depression. Finally, all
participants completed neuropsychological tests that
assessed language skills, verbal and spatial memory,

2 http://msj.sagepub.com

Downloaded from msj.sagepub.com at CILEA Bibliosan on October 9, 2015

R Riccelli, L Passamonti et al.

information processing speed, executive control and
ability to recognize facial identities (see Supplementary
Material online).

correction. Finally, for the sake of completeness, we

also reported the results that met a borderline threshold of p<0.001 uncorrected.

A written informed consent was obtained from all

participants and the study was approved by the Ethical
Committee of the University Magna Graecia of
Catanzaro, according to the Helsinki Declaration
(http://www.wma.net/e/policy/b3.htm).

See the Supplementary Material online for further

details about preprocessing and first level analysis.

Demographic, neuroimaging, neuropsychological

assessment
See the Supplementary Material online.
fMRI task
Participants categorized the sex of grey-scale photographs of angry, sad and neutral faces posed by 30
different actors (half were female). Additional tasks
details are provided in the Supplementary Material.

fMRI connectivity analysis: psycho-physiological

interaction analyses
A psycho-physiological interaction (PPI) represents
the change in connectivity (i.e. correlation) between
pairs of regions (i.e. a seed and the rest of the brain)
that is induced by a specific psychological context
(e.g. task A versus task B).27 The right and left amygdala and hippocampus were chosen as seed regions
because they have been consistently implicated in the
pathophysiology of depressive disorders.15 See the
Supplementary Material for further details.
Results

fMRI preprocessing and analysis of regional

effect
fMRI data were preprocessed using SPM8 (www.fil.
ion.ucl.ac.uk/spm). To investigate how depression
modulated the neural activity in the MS group, a general linear model (GLM) was employed.
To test for differences in brain response between controls and MS patients a full factorial analysis including two groups and three facial expressions (i.e.
angry, sad and neutral faces) as main factors was conducted. We first compared the sample of healthy controls with the whole sample of MS patients (77
patients). We then restricted the analyses to those 56
MS patients who showed no or minimal depression
(Table S1 online).
To evaluate how depression modulated brain responses
to emotional stimuli in MS patients, subject-specific
BDI-FS scores and contrast images were entered in
second level regression models. Two main approaches
were employed to threshold second-level maps. First,
we used a region of interest (ROI) approach based on
an anatomical definition of brain regions, as previously recommended (i.e. p<0.05, family-wise error
(fwe) correction for multiple comparisons using a
small volume correction (svc)).25 The following bilateral ROIs, extracted from the probabilistic Harvard
Oxford atlas, were used:26 VLPFC, DLPFC, OFC,
subgenual cingulate, amygdala, and hippocampus.
Second, we reported all the other brain regions that
met a threshold of p<0.05, fwe, whole-brain

Participants
Clinical, neuropsychological and demographic data in
MS patients and healthy controls are summarized in
Table 1. Fifty-five MS patients (i.e. 71.4%) showed
no or minimal signs of depression (i.e. BDI-FS scores:
03), while the remaining 22 MS patients (i.e. 28.6%)
displayed mild to severe forms of depression (i.e.
BDI-FS scores: 416). The presence of major depressive symptoms was confirmed in 15 MS patients (i.e.
19.4%) via the DSM-IV interview.
No significant correlations were found, within the
MS group, between BDI-FS scores and: (1) overall
neurological disability (EDSS) (R=0.14 p=0.22); (2)
disease duration (R=0.10, p=0.38); (3) total greymatter volume (GMV) (R= 0.18, p=0.11); (4) total
white-matter volume (WMV) (R= 0.07, p=0.55);
(5) total brain volume (R= 0.17, p=0.14), and (6)
total lesion load (R=0.02, p=0.86). Conversely, as
expected from previous research,28 depression scores
were positively associated with: (1) trait anxiety
scores (R=0.70, p<0.0001), and (2) fatigue levels
(R=0.45, p<0.0001) (Table 1). No significant correlations were found between depression and attentional, memory and facial identity recognition scores
(ps>0.3).
fMRI task
No significant correlations were found between percentage of correct answers and depression in MS
patients (R=0.10, p=0.39). Additional information

http://msj.sagepub.com 3

Downloaded from msj.sagepub.com at CILEA Bibliosan on October 9, 2015

Multiple Sclerosis Journal

Table 1. Demographic, clinical, neuroimaging and neuropsychological data of the sample.

Demographic and clinical data

Sex, F/M
Age, years [range]
Disease duration, years [range]
EDSS, median [range]
Fatigue Severity Scale [range]
Neuroimaging data
Total lesion volume, ml
Total grey matter volume, ml
Total white matter volume, ml
Total brain volume, ml
Neuropsychological data
Long-term storage
Consistent long-term storage
Selective Reminding Test Delayed (SRT-D)
Spatial Recall Test-Delayed
Spatial Recall Test-Immediate
Symbol Digit Modality Test (SDMT)
Word List Generation (WLG)
Stroop Colour-Word Test
Benton Face Test
Beck Depression Inventory (BDI) [range]
Beck Depression Inventory Fast Screen
(BDI-FS) [range]
Beck Depression Inventory Somatic
Subscale (BDI-S) [range]
State Trait Anxiety Inventory (STAY-II)

Healthy controls
(N=20)

Multiple sclerosis
patients (N=77)

t/2-values

Correlation
coefficient

10/10
36.4 (8.3) [2454]

46/31
34.0 (8.1) [1951]
6.8 (5.8) [0.221]
2 [14.5]
3.62 (1.83) [17]

0.39
t(95)=1.2ns

r = 0.23ns
r = 0.10ns
r = 0.14ns
r = 0.45*

17.84 (17.39)
606.3 (67.3)
477.0 (57.3)
1319.2 (123.8)

t(95)=0.8ns
t(95)=2.2ns
t(95)=0.7ns

r = 0.02ns
r = 0.18ns
r = 0.07ns
r = 0.17ns

t(95)=0.5ns
t(95)=0.7ns
t(95)=0.3 NS
t(95)=1.8ns
t(95)=1.5ns
t(95)=3.1*
t(95)=4.5*
t(95)=1.9ns
t(95)=2.6*
t(95)=1.6ns
t(95)=1.4ns

r = 0.10ns
r = 0.05ns
r = 0.03ns
r = 0.13ns
r = 0.05ns
r = 0.25ns
r = 0.25ns
r = 0.26ns
r = 0.21ns

t(95)=1.7ns

t(95)=2.3ns

r =0.70*

619.6 (51.0)
508.4 (57.7)
1340.1 (120.7)
34.79 (11.95)
25.94 (9.70)
6.36 (2.89)
6.94 (2.06)
19.74 (4.35)
48.10 (9.62)
20.05 (3.91)
21.37 (7.92)
48.2 (3.2)
7.35 (5.08) [016]
1.94 (1.52) [05]

37.55 (13.40)
28.49 (13.73)
6.25 (2.66)
5.86 (2.35)
17.97 (4.58)
38.11 (13.39)
15.61 (3.93)
18.29 (6.04)
45.3 (4.8)
11.10 (9.92) [045]
3.12 (3.55) [016]

5.16 (3.38) [010]

7.64 (6.15) [027]

34.45 (5.44)

40.74 (11.71)

Data are expressed as mean (SD) except for EDSS, which is expressed as median (range). Minimum and maximum values for demographic and clinical values
(age, disease duration and EDSS), BDI and its subscales are also provided. Correlation coefficients between depression severity (BDI-FS scores) and
demographic, clinical, neuropsychological and MRI data are referred to the MS group.
*p<0.01.
ns: not significant; r: Pearson correlation coefficient; MRI: magnetic resonance imaging; EDSS: Expanded Disability Status Scale.

on the fMRI behavioural data are reported in the

Supplementary Material Figure S1 and Tables S5
and S6.

Brain regional effects

Consistent with our previous work,18 we found
increased activation in the right VLPFC in MS
patients relative to controls (Table 2 and Figure 1(a))
when studying the main effect of the group. Similar
results were found when restricting the analyses to
those MS patients showing no or minimal depression
(see Table S1 in the Supplementary Material).
Depression scores in MS patients were negatively correlated with the activity in the right subgenual

cingulate cortex (Table 3 and Figure 2) for the contrast

all faces>rest. When assessing each facial expression
separately (i.e. angry>rest, neutral>rest and
sad>rest) similar results were obtained in the right
subgenual cingulate cortex for the contrast angry>rest
(i.e. p=0.02, fwe, svc) although the effect for the
neutral>rest and sad>rest comparison was statistically
borderline (i.e. p<0.001 uncorrected). No significant
findings were found for the contrasts angry>neutral
and sad>neutral. See the Supplementary Material for
additional analyses.

Functional connectivity results

Similar to our previous results,18 an altered pattern of
connectivity between the left amygdala and left

4 http://msj.sagepub.com

Downloaded from msj.sagepub.com at CILEA Bibliosan on October 9, 2015

R Riccelli, L Passamonti et al.

Table 2. Results of the full factorial analysis with two groups (healthy controls and MS patients) and three facial
expressions (angry, sad and neutral faces) for brain response and functional connectivity.
Side
A. Factorial analysis for brain response
Main effect of group
VLPFC
R
DLPFC
R
Main effect of facial expression
VLPFC
L

R
DLPFC
R
Amygdala
L

R
Hippocampus
L

4.09**
3.60*

44
24

2
40

28
44

4.53**
4.26**
4.03*
4.13**
3.63**
3.59*
4.32**

56
58
42
24
26
18
14

22
32
14
8
12
28
30

28
14
24
22
10
6
8

38

36

14

32
18
6

34
6
18

26
34

10
10

16
0
26

24
32
24

30
22

12
36

26

14

28

B. Full factorial analysis for psychophysiological interaction (seed: left amygdala)

Main effect of group
L
3.14*
28
OFC
Main effect of expression
L
4.87**
20
Hippocampus
Interaction group by expression
L
3.7*
44
VLPFC/DLPFC
L
3.25*
42
OFC
R
3.08*
24
C. Full factorial analysis for psychophysiological interaction (seed: right amygdala)
Main effect of group
OFC
L
3.52*
22
R
3.38*
18
Main effect of expression
L
5.7***
36
Hippocampus
R
3.96**
26
L
4.2**
52
VLPFC
Interaction group by expression
OFC
R
3.82*
24
VLPFC
L
3.03*
36
D. Full factorial analysis for psychophysiological interaction (seed: left hippocampus)
Main effect of group
OFC
L
3.33*
30
Main effect of expression
VLPFC
R
3.22*
28
Interaction group by expression
None
E. Full factorial analysis for psycho-physiological interaction (seed: right hippocampus)
Main effect of group
VLPFC
R
3.15*
46
OFC
R
3.88**
14
Main effect of expression
Hippocampus
L
3.33*
32
R
3.04*
36
Interaction group by expression
VLPFC
L
3.5*
40
R
3.01*
48

32
36

14
14

12
10

24
18

24
14

40
42

**p<0.05 fwe, small volume corrected.

*p<0.1 fwe, small volume corrected or p<0.001 uncorrected.
Z: Z-score: L: left hemisphere; R: right hemisphere; x, y, z: MNI coordinates of local maxima in millimetres; VLPFC: ventrolateral
prefrontal cortex; DLPFC: dorsolateral prefrontal cortex; OFC: orbitofrontal cortex; fwe: family-wise error.

http://msj.sagepub.com 5

Downloaded from msj.sagepub.com at CILEA Bibliosan on October 9, 2015

Multiple Sclerosis Journal

Figure 1. Differences between controls and MS patients in brain regional responses and functional connectivity patterns
arising from the left amygdala seed. An increased activation in MS patients relative to controls was found in the right
VLPFC/DLPFC (a) for the main effect of group. Moreover, an altered pattern of functional connectivity between the left
amygdala and the VLPFC was found for the group task interaction (b) Colour Bars represent F-statistic.
MS: multiple sclerosis; HCs: healthy controls; VLPFC: ventrolateral prefrontal cortex; R: right.

Table 3. Effect of depression levels in MS patients on brain regional responses. Correlation between depression severity
(expressed as BDI-FS scores) and brain responses in the MS patients for the contrast all faces>null event.
Correlation with BDI-FS score

Positive

Side

DLPFC

R
L

3.52*
3.34*

18
30

24
12

52
52

Subgenual cingulate
cortex (ACC)
Amygdala

Negative
Side

R
L
L

3.56**
3.12*
3.48*

14
14
16

14
16
10

8
10
16

**p<0.05 fwe, small volume corrected.

*p<0.1 fwe, small volume corrected or p<0.001 uncorrected.
Z: Z-score; L: left hemisphere; R: right hemisphere; x, y, z: MNI coordinates of local maxima in millimetres; DLPFC: dorsolateral
prefrontal cortex; BDI-FS: Beck Depression Inventory-Fast Screen; fwe: family-wise error; ACC: anterior cingulate cortex.

VLPFC/DLPFC was found when exploring the main

effect of the group (i.e. MS patients and controls)
(Table 2 and Figure 1(b)).

When assessing for the effect of depression on the

functional connectivity patterns in MS patients, we
found significantly negative correlations between

6 http://msj.sagepub.com

Downloaded from msj.sagepub.com at CILEA Bibliosan on October 9, 2015

R Riccelli, L Passamonti et al.

Figure 2. Effect of depression severity on regional brain activations during the faces processing task. Correlation
between depression severity and brain regional response in MS patients, for the contrast all faces> rest. Depression scores
were negatively correlated with the activity in the subgenual anterior cingulate cortex (ACC). Colour bar represents
T-statistic.
R: right.

depression scores and the connectivity between the

left hippocampus and OFC as well as DLPFC (for the
contrast all faces>rest) (Figure 3 and Table 4).
Likewise, a negative correlation between depression
scores in MS patients and the connectivity change
between the right amygdala and left DLPFC was identified (Figure 4 and Table 4). See the Supplementary
Material for further connectivity analyses.
Discussion
To our knowledge, this is the first fMRI study showing that depression in MS patients was associated
with changes in regional activity and functional connectivity patterns within key limbic fronto-temporal
circuits. Specifically, depression in MS patients was
negatively correlated with the activity in the subgenual cingulate cortex and with functional connectivity
abnormalities between the hippocampus and the OFC
and DLPFC as well as with functional connectivity
alterations between the amygdala and DLPFC.
Notably, no significant correlations were found
between depression and a series of clinical and neuroimaging measures that reflect overall neurological
disability as well as global brain damage in MS
(i.e. EDSS score, disease duration, total GMV, total
WMV, and total lesion load). Considered together,
these results provide support to the notion that depression in MS is associated with dysfunctions within
specific neural networks rather than being primarily
linked to the overall neurological disability.
Contrary to our predictions, we did not find significant associations between depression and regional
responses in the amygdala and hippocampus, although

dysfunctions in these regions may still represent an

important pathophysiological mechanism of depression in MS.29,30 There are at least two possible explanations for this null result. First, regional effects
within the amygdala and hippocampus in MS patients
with depression may emerge using other emotional
tasks like those using affective pictures or fearconditioning paradigms. A not mutually exclusive
possibility is that normal amygdala and hippocampal
responses in MS patients with depression may be
driven by enhanced top-down PFC control over
these sub-cortical areas. Increased PFC control over
the amygdala and hippocampus was indeed suggested
by the greater response of the DLPFC that we found
as a function of higher depression scores in MS
patients, although this result was statistically borderline. Nevertheless, increased activity in different parts
of the PFC has been consistently observed in MS
patients during the execution of a broad range of
fMRI tasks and this is thought to reflect adaptive
mechanisms to maintain normal levels of behavioural
performances.17 Consistent with this view, we found
increased activation in the right VLPFC, a key PFC
region at the interface between cognitive and emotional systems, when comparing MS patients with
controls. Likewise, depression scores in MS patients
were not associated with behavioural performances
during the execution of the face task, which suggests
the presence of adaptive mechanisms that compensate
for MS-related damage. Furthermore, it is important
to highlight that we found decreased activity in the
subgenual cingulate cortex as a function of higher
depression scores in MS patients and this results
diverged from some previous studies in patients with
major depression which found increased and not

http://msj.sagepub.com 7

Downloaded from msj.sagepub.com at CILEA Bibliosan on October 9, 2015

Multiple Sclerosis Journal

Figure 3. Effect of depression severity on brain functional connectivity patterns arising from the left hippocampal
seed. Psycho-physiological interaction analyses revealed an altered connectivity pattern between the left hippocampus
and the bilateral OFC and left DLPFC as a function of depression severity in MS patients. Colour bar represents
T-statistic.
OFC: orbitofrontal cortex; DLPFC: dorsolateral prefrontal cortex; L: left.

decreased activity in the same PFC area.31 We also

assessed whether the reduced response in the subgenual cingulate cortex in MS patients with depression
was driven by partial volume effects and we found
that this was not the case (i.e. no relationship between
grey-matter volume in the subgenual cingulate cortex
and depression was found; see the Supplementary
Material). The reduced subgenual cingulate cortex
activity in depressed MS patients may thus reflect
specific MS-related changes within the limbic system and may represent an important biomarker of a
common psychiatric disorder in MS, although further
studies including groups of patients with mood disorders are necessary to verify this hypothesis.
Another key finding of this study was represented by
the altered patterns of connectivity that we found
between the amygdala and DLPFC as well as between

the hippocampus and DLPFC as a function of higher

depression scores in MS patients. Although the DLPFC
sends very limited outputs to either the amygdala or
hippocampus and vice versa, several indirect neuronal
pathways enable a robust cross-talk between these
regions.32 Most of these routes involve the OFC, a PFC
region that showed a significant modulation of its connectivity pattern with the hippocampus as a function of
higher depression scores in MS patients. Therefore, it is
possible that the amygdalaDLPFC and hippocampus
DLPFC connectivity results were mediated via intermediate areas such as the OFC, which is also consistent
with the notion that PPI analyses are able to identify
connectivity effects between pairs of brain areas that
are not necessarily linked via direct anatomical pathways.27 Furthermore, we would like to emphasize that
structural damage of the white matter fibres that drive
the information flow throughout the limbic system may

8 http://msj.sagepub.com

Downloaded from msj.sagepub.com at CILEA Bibliosan on October 9, 2015

R Riccelli, L Passamonti et al.

Table 4. Effect of depression levels in MS patients on brain connectivity measures. Correlation of coupling (between
seed and target areas) and depression severity (expressed as BDI-FS scores) in the MS patients for the contrast all
faces>null event.
Correlation with BDI score

Positive

Side

Negative
Z

A. Left hippocampus seed

OFC

DLPFC

VLPFC

Hippocampus
B. Right amygdala seed
DLPFC
VLPFC

OFC
C. Right hippocampus seed
DLPFC
OFC

VLPFC
D. Left amygdala seed
VLPFC

Side

L
R
R
L
R
L
R

4.59***
3.79*
4.88***
4.47**
3.87*
3.86*
3.84**

26
34
10
20
44
48
22

24
14
26
28
42
42
22

10
10
66
62
0
0
8

L
R
L
R

4.40**
3.82*
3.69*
3.72*

28
32
50
54

6
22
42
36

66
22
0
2

L
R
L
L

3.91*
3.54*
3.50*
3.42*

28
30
24
56

8
40
22
16

64
4
12
32

3.14*

40

26

26

Seed regions are the left hippocampus (A), the right amygdala (B), the right hippocampus (C) and the left amygdala (D). Target
areas are those areas that showed a significant dependence of coupling on depression scores in the psycho-physiological interaction
analysis.
***p<0.05 fwe, whole brain corrected.
**p<0.05 fwe, small volume corrected.
*p<0.1 fwe, small volume corrected or p<0.001 uncorrected.
Z: Z-score; L: left hemisphere; R: right hemisphere; x, y, z: MNI coordinates of local maxima in millimetres; BDI: Beck Depression
Inventory; BDI-FS: BDI-Fast Screen; OFC: orbitofrontal cortex; DLPFC: dorsolateral prefrontal cortex; fwe: family-wise error.

be the underlying cause of the altered functional

connectivity patterns that we observed between the
hippocampus, amygdala and DLPFC. This hypothesis
is corroborated by our recent DTI study in which we
found reduced communicability across the amygdala,
hippocampus and a series of PFC areas including the
OFC, VLPFC and DLPFC.10
Taken together, our findings are in general agreement
with current neurobiological models of depression
which emphasize the critical role of the hippocampus,
amygdala and different parts of the PFC in mediating
the hormonal and viscero-motor responses that characterize mood disturbances.16 More specifically, the
imbalanced cross-talk between the hippocampus,
amygdala and PFC may be responsible for the
increased cortisol response that has been reported in
MS patients with depression,30 although further
studies simultaneously assessing hormonal and brain

connectivity measures are necessary to verify this

hypothesis. It would also be worth testing whether
peripheral measures of inflammation influence brain
activity and connectivity patterns in relation to depression in MS. This is because there is emerging evidence that peripheral inflammation may influence the
brain circuits associated with emotional behaviours
including depression.5
A further issue is whether the present results extend to
other emotions that have not been directly tested in
our experiment (e.g. fear, happiness, surprise and disgust) or whether, alternatively, depression in MS
modulates brain regional responses and functional
connectivity patterns only when processing angry, sad
or neutral faces. Although our study was not able to
address this issue, we speculate that current results
may generalize to other emotional and non-emotional
stimuli as previous studies have found that patients

http://msj.sagepub.com 9

Downloaded from msj.sagepub.com at CILEA Bibliosan on October 9, 2015

Multiple Sclerosis Journal

Figure 4. Effect of depression severity on brain functional connectivity patterns arising from the right amygdala
seed. Psycho-physiological interaction analyses revealed an altered connectivity pattern between the right amygdala and
the left DLPFC as a function of depression severity in MS patients. Although at a lower statistical threshold (i.e. p<0.001,
uncorrected), altered patterns of connectivity were also found between the right amygdala and the orbitofrontal cortex
and ventrolateral prefrontal cortex. Colour bar represents T-statistic.
DLPFC: dorsolateral prefrontal cortex; L: left; R: right.

with depressive disorders can display an overall negative bias for a number of stimuli.33
Another limitation of our study was represented by its
cross-sectional nature, which did not allow us to infer
whether the changes in brain function that we
observed were actually causing depression in MS or
not all we can say is that depression was associated
with altered patterns of brain activity and functional
connectivity within limbic circuits. This also implies
that longitudinal studies using multiple brain scans
over time would be fundamental to further characterize the brain determinants of depression in MS.
Lastly, it remains to be tested whether our findings in
MS patients with the relapsingremitting form of the
disease extend to other MS phenotypes (e.g. primaryor secondary-progressive MS) or other mood disturbances that have been described in MS (e.g. bipolar
disorder).

Conclusion
To summarize, our study provided compelling new
evidence that depression in MS patients was significantly associated with altered regional activity and
functional connectivity patterns within the limbic system, a group of brain areas that are fundamental for
the survival of human beings.
Acknowledgements
The authors are grateful to the multiple sclerosis
patients and controls who kindly participated in the
experiment. RR performed the main analyses and
wrote the manuscript. LP was responsible for the
study concept and design, contributed to data acquisition and drafted the manuscript. AC implemented the
fMRI task and provided critical revision to the manuscript. SN performed structural brain imaging analyses. SC provided critical revision to the manuscript.
CC administered the neuropsychological tests to all
participants. PV and RN evaluated clinical signs and

10 http://msj.sagepub.com

Downloaded from msj.sagepub.com at CILEA Bibliosan on October 9, 2015

R Riccelli, L Passamonti et al.

symptoms in multiple sclerosis patients. AQ was
responsible for the study concept and design, and provided critical revision to the manuscript. All of the
authors reviewed the content of the manuscript and
approved its final version for submission. RR and LP
contributed equally to the work.
Conflict of interest
The authors declare that there is no conflict of
interest.

multiple sclerosis patients with major depression.

Mult Scler 2015; 21: 10031012.
11. Pelletier J, Audoin B, Reuter F, et al. Plasticity in
MS: From functional imaging to rehabilitation. Int
MS J 2009; 16: 2631.
12. Fusar-Poli P, Placentino A, Carletti F, et al.
Functional atlas of emotional faces processing: A
voxel-based meta-analysis of 105 functional magnetic
resonance imaging studies. J Psychiatry Neurosci
2009; 34: 418432.

Funding
This work was supported by The National Research
Council.

13. Davis M. The role of the amygdala in fear and

anxiety. Annu Rev Neurosci 1992; 15: 353375.

References
1. Feinstein A, Magalhaes S, Richard JF, et al. The link
between multiple sclerosis and depression. Nat Rev
Neurol 2014; 10: 507517.

15. Stuhrmann A, Suslow T and Dannlowski U. Facial

emotion processing in major depression: A systematic
review of neuroimaging findings. Biol Mood Anxiety
Disord 2011; 1: 10.

2. McIvor GP, Riklan M and Reznikoff M. Depression

in multiple sclerosis as a function of length and
severity of illness, age, remission and perceived social
support. J Clin Psychol 1984; 40: 10281033.

16. Drevets WC, Price JL and Furey ML. Brain structural

and functional abnormalities in mood disorders:
Implications for neurocircuitry models of depression.
Brain Struct Funct 2008; 213: 93118.

3. Millefiorini E, Padovani A, Pozzilli C, et al.

Depression in the early phase of MS: Influence of
functional disability, cognitive impairment and brain
abnormalities. Acta Neurol Scand 1992; 86: 354358.

17. Staffen W, Mair A, Zauner H, et al. Cognitive

function and fMRI in patients with multiple
sclerosis: Evidence for compensatory cortical
activation during an attention task. Brain 2002;
125: 12751282.

4. Minden SL, Orav J and Reich P. Depression in

multiple sclerosis. Gen Hosp Psychiatriy 1987;
9: 426434.
5. Pokryszko-Dragan A, Frydecka I, Kosmaczewska
A, et al. Stimulated peripheral production of
interferon-gamma is related to fatigue and
depression in multiple sclerosis. Clin Neurol
Neurosurg 2012; 114: 11531158.
6. Zorzon M, de Masi R, Nasuelli D, et al. Depression
and anxiety in multiple sclerosis. A clinical and
MRI study in 95 subjects. J Neurol 2001; 248:
416421.
7. Feinstein A, Roy P, Lobaugh N, et al. Structural brain
abnormalities in multiple sclerosis patients with major
depression. Neurology 2004; 62: 586590.
8. Feinstein A, OConnor P, Akbar N, et al. Diffusion
tensor imaging abnormalities in depressed multiple
sclerosis patients. Mult Scler 2010; 16: 189196.
9. Rocca MA, Valsasina P, Martinelli V, et al. Largescale neuronal network dysfunction in relapsingremitting multiple sclerosis. Neurology 2012; 79:
14491457.
10. Nigro S, Passamonti L, Riccelli R, et al. Structural
connectomic alterations in the limbic system of

14. Herman JP, Schafer MK, Young EA, et al. Evidence

for hippocampal regulation of neuroendocrine
neurons of the hypothalamo-pituitary-adrenocortical
axis. J Neurosci 1989; 9: 30723082.

18. Passamonti L, Cerasa A, Liguori M, et al.

Neurobiological mechanisms underlying emotional
processing in relapsingremitting multiple sclerosis.
Brain 2009; 132: 33803391.
19. Rocca MA and Filippi M. Functional MRI in multiple
sclerosis. J Neuroimaging 2007; 17(Suppl. 1):
36S41S.
20. Mayberg HS. Limbic-cortical dysregulation: A
proposed model of depression. J Neuropsychiatry
Clin Neurosci 1997; 9: 471481.
21. Korgaonkar MS, Fornito A, Williams LM, et al.
Abnormal structural networks characterize major
depressive disorder: A connectome analysis. Biol
Psychiatry 2014; 76: 567574.
22. Berney A, Nishikawa M, Benkelfat C, et al. An index
of 5-HT synthesis changes during early antidepressant
treatment: alpha-[11C]methyl-L-tryptophan PET
study. Neurochem Int 2008; 52: 701708.
23. Beck AT, Steer RA and Brown GK. Manual for the
BDIFast Screen for medical patients. San Antonio,
TX: Psychological Corporation, 2000.
24. Benedict RH, Fishman I, McClellan MM, et al.
Validity of the Beck Depression InventoryFast

http://msj.sagepub.com 11

Downloaded from msj.sagepub.com at CILEA Bibliosan on October 9, 2015

Multiple Sclerosis Journal

Screen in multiple sclerosis. Mult Scler 2003; 9:
393396.
25. Friston KJ. Testing for anatomically specified
regional effects. Hum Brain Mapp 1997; 5: 133136.
26. Desikan RS, Segonne F, Fischl B, et al. An automated
labeling system for subdividing the human cerebral
cortex on MRI scans into gyral based regions of
interest. Neuroimage 2006; 31: 968980.
27. Friston KJ, Buechel C, Fink GR, et al.
Psychophysiological and modulatory interactions in
neuroimaging. Neuroimage 1997; 6: 218229.

Visit SAGE journals online

http://msj.sagepub.com

SAGE journals

30. Gold SM, Kern KC, OConnor MF, et al. Smaller

cornu ammonis 2-3/dentate gyrus volumes and
elevated cortisol in multiple sclerosis patients with
depressive symptoms. Biol Psychiatry 2010; 68:
553559.
31. Gotlib IH, Sivers H, Gabrieli JD, et al. Subgenual
anterior cingulate activation to valenced emotional
stimuli in major depression. Neuroreport 2005; 16:
17311734.

28. Beiske AG, Svensson E, Sandanger I, et al.

Depression and anxiety amongst multiple sclerosis
patients. Eur J Neurol 2008; 15: 239245.

32. Phillips ML, Ladouceur CD and Drevets WC.

A neural model of voluntary and automatic
emotion regulation: Implications for understanding
the pathophysiology and neurodevelopment of
bipolar disorder. Mol Psychiatry 2008; 13: 829,
3357.

29. Gold SM, OConnor MF, Gill R, et al. Detection of

altered hippocampal morphology in multiple sclerosisassociated depression using automated surface mesh
modeling. Hum Brain Mapp 2014; 35: 3037.

33. Leppanen JM, Milders M, Bell JS, et al.

Depression biases the recognition of emotionally
neutral faces. Psychiatry Res 2004; 128:
123133.

12 http://msj.sagepub.com

Downloaded from msj.sagepub.com at CILEA Bibliosan on October 9, 2015