Neurological Sciences (2021) 42:835–845

https://doi.org/10.1007/s10072-020-04951-z

REVIEW ARTICLE

Imaging and depression in multiple sclerosis: a historical perspective

Fabio Giuseppe Masuccio 1 & Giulia Gamberini 1 & Massimiliano Calabrese 2 & Claudio Solaro 1

Received: 28 February 2020 / Accepted: 28 November 2020 / Published online: 7 January 2021
# Fondazione Società Italiana di Neurologia 2021

Abstract
Purpose Patients affected with multiple sclerosis suffer from depression more frequently than the general population. Beyond
psychosocial, genetic and immune-inflammatory factors, also the brain damage which is peculiar of multiple sclerosis has been
claimed to have a role in the aetiology of depression in those patients. The study of this interesting relation has been implemented
with both conventional and advanced magnetic resonance imaging techniques. The aim of this review is to provide a historical
perspective on the link between multiple sclerosis-related depression and structural and functional brain damage.
Methods In this review, the results of the MRI studies regarding multiple sclerosis-related brain damage and the presence of
depression are presented.
Results The findings of the reports reveal a link between brain pathology and depressive symptoms or the diagnosis of depression
in multiple sclerosis.
Conclusions Although a multifactorial aetiology has been theorized for depression and depressive symptoms in patients with
multiple sclerosis, this review supports the hypothesis that the structural and functional brain impairment might substantially be
amongst those factors. Thus, depression itself might be a symptom with a neuro-biological basis and not only the consequence of
the disability derived from the neurological impairment.

Keywords Multiple sclerosis . Depression . Magnetic resonance imaging . Depressive symptoms

Introduction respect to the general population [1]. Amongst PwMS, fe-

It is well recognized that patients with multiple sclerosis
(PwMS) experience depression with a higher frequency in
Key points
• Patients with multiple sclerosis experience depression/depressive symp-
toms more frequently than the general population. Depression in patients
with multiple sclerosis has to be considered as a process with a multifac-
torial aetiology (genetic, psychosocial, immune-inflammatory factors,
brain damage).
• Structural and functional brain damage has been consistently associated
with depression and/or depressive symptoms in patients with multiple
sclerosis.
• Lack of a former depression diagnosis according to DSM criteria is
present in several papers. Only few longitudinal studies are available.
• The available studies show high variability in design, inclusion criteria
and sample size, leading to difficulty in drawing conclusion.
* Claudio Solaro
csolaro@libero.it
1
Department of Rehabilitation, C.R.R.F. “Mons. L. Novarese”, Loc.
Trompone, Moncrivello, VC, Italy
2
Multiple Sclerosis Center, Neurology section, Department of
Neurosciences, Biomedicine and Movement Sciences, University of
Verona, Verona, Italy
males and those younger than 45 years appear to be more at
risk for depression. Its prevalence differs within the diverse
forms of multiple sclerosis (MS), being higher in Secondary
Progressive Multiple Sclerosis (SPMS) than in relapsing-
remitting MS (RRMS) and primary progressive MS (PPMS)
[2].
The specific causes underlying the higher rates of depres-
sion in PwMS have not been well acknowledged. The efforts
of the research identified some aspects potentially involved in
the pathogenesis of depression in these individuals. Genetic,
immune-inflammatory and psychosocial factors, as well as the
typical structural brain damage and disease-modifying treat-
ments (i.e. interferon-β), have been claimed to have a part in
this process. Although all these factors have been involved, at
the present time, no evidence is available for a possible role of
disease-modifying therapies in MS-related depression [3].
When considering the demyelinating brain injury, the pos-
sibility that depression and depressive symptoms in PwMS
might be in some way related to the structural brain alterations
has been abundantly debated. If this is the case, depression in
MS might be considered as a condition with a neuro-
biological basis [4], directly deriving from the primary disease
836
and not only the mere consequence of the MS-related disabil-
ity and of the subsequent reduction of social participation. In
order to implement the investigation on this intriguing link,
magnetic resonance imaging (MRI), a fundamental tool in the
diagnosis and monitoring of MS [5], has been used.
In this review, a historical overview of MRI studies dedi-
cated to the examination of MS-related depression is provid-
ed. The included researches are divided on the basis of the
different methodology used, starting from the results of con-
ventional MRI sequences and going through those obtained
with advanced MRI techniques. The studies are classified ac-
cording to the most advanced MRI methodology employed.
In addition, distinguishing the presence of depressive
symptoms from a former diagnosis of depression is mandato-
ry. Clinical scales, useful screening instruments, can only
identify the presence of depressive symptoms. This does not
allow to make a diagnosis of depression, which should be
formulated by a clinician according to the Diagnostic and
Statistical Manual of Mental Disorders (DSM) criteria. In this
review, we use the term “depression” when DSM criteria have
been fulfilled, “depressive symptoms” when only scales have
been utilized. We also pointed out when the authors classified
the individuals as depressed on the sole basis of scales.
References for this review were identified searching in
Medline and PubMed for all the articles in the English lan-
guage published in peer-reviewed journals, using the term
“multiple sclerosis”, “depression” and “magnetic resonance
imaging”. We checked reference lists for other studies of
interest
Conventional MRI
The first step taken was the analysis of the white matter dis-
ruption using conventional MRI sequences, such as T2
weighted (T2W), T1 weighted (T1W) and gadolinium en-
hanced (Gad+). In particular, T2W lesions are hyperintense
in contrast to a low signal background given by the normal
white matter. Conversely, the T1W lesions, the so-called
black holes, appear to be hypointense and represent a more
chronic feature, where major and non-reversible tissue dam-
age is present. The Gad+ lesions consist in areas of contrast
enhancement, namely the expression of the breakdown of the
blood-brain barrier, due to acute inflammation.
The earliest studies attempted to examine the relation be-
tween brain damage and psychiatric conditions in MS using
T2W sequences. A higher lesion volume in the temporal lobe
[6], in the frontal lobe [7] and a higher lesion score in the
temporal-parietal regions [8] were associated with a higher
level of psychiatric symptoms in PwMS, albeit with a small
influence on the feature “depressed mood”. The samples con-
sidered in these studies were not numerous and the authors did
not forthwith focus on depression. However, as a matter of
Neurol Sci (2021) 42:835–845
fact, these findings paved the way for further inquiry and
additional researches flourished on the link between depres-
sion in MS and the structural brain damage assessed through
MRI.
Among the initial researches, two negative findings have to
be reported. The first was conducted in patients in the early
phases of RRMS (ten males and eight females) and detected a
similar T1W and T2W lesion load between depressed and
non-depressed subjects [9]. The second found no differences
in terms of regional distribution of lesions measured through
T1W and T2W sequences in ten non-depressed and ten de-
pressed PwMS, classified according to both Beck Depression
Inventory (BDI) and Hamilton Depression Rating Scale
(HDRS) [10].
In turn, other studies underlined a link between structural
damage and depression in MS. George et al. [11] observed a
positive correlation between left hemisphere T1W lesions and
HDRS in 16 PwMS, in association with a higher number of
these lesions in the whole brain in those who were defined as
depressed. Subsequently, Pujol et al. [12] associated the BDI
score with the topography of the white matter lesions, in par-
ticular in the uncinate fasciculus, in the cingulum and in the
arcuate fasciculus in 45 PwMS. The total T2W lesion load and
the lesion load in the left arcuate fasciculus correlated with
BDI, even if only 17% of the BDI score variance was ex-
plained. In a following report [13], the demyelinating lesions
in the left arcuate fasciculus were related with the BDI cate-
gories “Affective symptoms” and “Somatic complaint”,
predicting the 26% of the score variance of the specific BDI
category. Thereafter, a significant correlation between the
presence of depression and higher T2W lesion load in both
the right temporal and frontal lobe in 31 depressed out of 78
PwMS was underlined [14].
Another hypothesis arose from Bakshi and colleagues [15],
who suggested that white matter damage between cortical and
subcortical areas involved in mood control might represent the
potential origin of depression in PwMS. Forty-eight PwMS
were classified as depressed or non-depressed in agreement
with the DSM-IV criteria for unipolar depression and tested
with Hamilton Depression Inventory (HDI) and BDI, T1W
and T2W, proton density and fast fluid-attenuated inversion
recovery (FLAIR) sequences. The presence of T1W lesions in
superior frontal and superior parietal regions significantly pre-
dicted the diagnosis of depression and, in association with
indirect signs of temporal and frontal atrophy, the severity of
depression, which was assessed through HDI and BDI [15].
It is not easy to interpret the findings when considering
Gad+ imaging. In a work by Moeller et al. [16], depression
was not correlated with either the presence or the absence of
lesions with enhancement. Kallaur et al. [17] examined 150
PwMS, 42 of which classified as being depressed according to
the Hospital Anxiety and Depression Scale (HADS). The se-
verity of depressive symptoms was found to be inversely
Neurol Sci (2021) 42:835–845
associated with Gad+ lesions, postulating that a chronic in-
flammatory stimulus, rather than an acute one, might be the
trigger for MS-related depression.
In a further study [18], 78 out of 405 RRMS patients, de-
fined as being “active” as they had Gad+ lesions, had higher
Beck Depression Inventory-II (BDI-II) scores and higher fre-
quency of clinically relevant depressive symptoms. Amongst
them, 29 were experiencing a clinical relapse and 49 were not.
The “active” patients with or without a clinical relapse did not
differ in terms of BDI-II. After 3 months, 74 “active” patients
underwent MRI and BDI-II determination for a follow-up
evaluation. Lower BDI-II scores were ascertained in concom-
itance with the remission of the clinical relapse, advocating for
the idea that the clinical relapse itself might exert in some way
a depressive influence.
In summary, the majority of the reports documented a cor-
relation between depression and depressive symptoms in
PwMS and T1W and T2W lesions, in particular in the frontal
and temporal lobes. The findings are in disagreement in only
two papers, probably due to the small sample size or to the
short disease duration [9, 10]. In association, the results pro-
vided from Gad+ imaging are clearly contradictory, feasibly
owing to the variability of the samples and of the methods of
determination of depression.
Advanced MRI
Over time, conventional MRI sequences were improved with
quantitative and advanced techniques, such as the measure-
ment of grey and white matter atrophy, diffusion tensor imag-
ing (DTI) and functional MRI (fMRI). These more sophisti-
cated methodologies have helped the comprehension of the
MS pathology, underlining that a more complex damage was
present beyond the white matter involvement.
Atrophy
Brain atrophy is an early and frequent phenomenon in MS,
mostly due to the cortical and deep grey matter damage, the
origin of which is only in part understood [19]. This has been
demonstrated in vivo by several MRI studies which measured
brain volume or cortical thickness applying automatic and
semi-automatic algorithms, mainly using surface- or voxel-
based morphometry on conventional MRI images.
Atrophy was considered as an alteration potentially linked to
depression in PwMS since the study of Bakshi et al. [15], where
signs of temporal and frontal atrophy were correlated to this
pathology. Afterwards, Zorzon et al. [20] selected 95 PwMS,
matching them with 110 healthy controls and with 97 patients
suffering from chronic rheumatic diseases. All the subjects were
evaluated with DSM-IV to diagnose depression and HDRS to
quantify depressive symptoms. For what concerns MRI analysis,
837
T1W, T2W imaging, regional and total lesion load calculation
and brain atrophy measurement were utilized. In PwMS, HDRS
was correlated to reduced total temporal brain volume and right
hemisphere brain volume.
After 2 years, 90 of those PwMS were reassessed [21]. The
presence of depression and the HDRS values resulted to be
correlated with right temporal atrophy. In three patients who
developed depression, additional T2W lesions and atrophy in
the right temporal lobe were found. Although this sample of
newly depressed individuals was very small, the potential
connection between atrophy of this area and the development
of depression was highlighted.
Stronger support for the relation between atrophy and de-
pression was given by Feinstein et al. [22]. In this study,
carried out in 40 PwMS, 21 of which classified as depressed
according to DMS-IV criteria, less grey matter volume and
higher atrophy of the left anterior temporal region were
highlighted. Furthermore, in the case of major depression,
the combination of T2W lesion load in the left medial inferior
frontal area and the cortical atrophy of the left anterior tempo-
ral gyrus explained the 42% of the variance.
Using a cross-sectional and longitudinal design with an
inter-test interval of 17 months, Stuke et al. examined 40 pa-
tients with RRMS and SPMS [23]. The subjects underwent
T1W magnetization-prepared rapid acquisition with gradient
echo (MPRAGE) MRI scans, voxel-based morphometry and
depressive status assessment through BDI. At the baseline, the
association between left temporal lobe atrophy and depressive
symptoms was confirmed. In addition, the longitudinal results
documented a significant relation between the worsening of
depressive symptoms and growing atrophy in both the left thal-
amus and right internal globus pallidus. The deterioration of the
depressive status was also linked to the reduction of grey matter
volume in the right cingulate cortex and frontal lobe.
In 126 PwMS and 59 healthy controls who underwent
T1W three-dimensional (3D) scans and depressive symptoms
assessment with Montgomery and Asberg Depression Scale
(MADRS), the bilateral frontal cortical thinning was a predic-
tor of depression, and MADRS severity was correlated to right
entorhinal cortical thinning [24]. Besides, Nyygard et al. [25]
detected a negative association between depressive symptoms
and cortical surface area and volume of the frontal and parietal
lobes in 61 RRMS patients. T1W MPRAGE imaging, FLAIR
sequences and quantification of cortical grey matter analysis
were utilized.
These data encouraged the hypothesis that the progression
of structural brain damage over time might be the basis of the
MS-related depression.
Hippocampal atrophy
The interest in the hippocampus grew on the basis of the
results on subjects without MS affected with major
838
depression. In those patients, hippocampal atrophy was asso-
ciated with depression [26]. In parallel, also in PwMS, some
studies correlated the presence of atrophy of the hippocampus
with depression. The theory regarding this link postulated that
hippocampal damage might contribute to depression reducing
the peculiar function of this structure, i.e. the decrease of the
stressing response to depressive stimuli.
In 29 RRMS patients vs 20 healthy controls, the T1W 3D
scans and FLAIR imaging demonstrated smaller volumes of
the dentate gyrus of the hippocampus in the depressed PwMS
cohort (BDI-II score > 13; n = 8). In addition, a smaller mean
hippocampal volume was found in PwMS in comparison to
the healthy controls [27].
In a retrospective analysis, Kiy et al. [28] revealed a posi-
tive correlation between larger left temporal horn volume (an
indirect measure of the hippocampal volume) and either the
psychic nor the somatic items of BDI. Sixty-four patients with
RRMS, 8 with PPMS and 16 healthy controls were examined
using 3D T1W MPRAGE imaging and the calculation of
brain and temporal horn volume.
In another study, 123 PwMS (80 RRMS, 18 with a benign
form, 17 SPMS, 8 PPMS) and 90 matched for sex and age
healthy controls were evaluated with T1W, T2W imaging and
voxel-based morphometry [29]. Sixty-nine PwMS were clas-
sified as being depressed on the basis of MADRS.
Noteworthy, depressed and non-depressed PwMS did not dif-
fer in terms of white matter atrophy. On the contrary, the
analysis of grey matter atrophy showed that depression was
selectively related to atrophy of the left middle frontal gyrus
and right inferior frontal gyrus.
Gold et al. [30] considered a large cohort of female PwMS
who underwent T1W, T2W and Gad+ MRI of the brain as
well as the calculation of the hippocampal volume and shape.
The Center for Epidemiologic Studies-Depression scale
(CES-D) was used to divide the patients into two groups, the
first with a low grade of depression (n = 83) and the second
with a high grade of depression (n = 26). The two cohorts were
similar for left and total hippocampal volumes, but the cluster
including highly depressed subjects had significantly smaller
right hippocampal volume. A modestly significant correlation
was detected between depressive affect and right (r = − 0.23,
p = 0.02) and total hippocampal volume (r = − 0.22, p = 0.02)
but not left hippocampus.
Although the lack of standardized protocols and the pres-
ence of white matter lesions might represent potential con-
founding factors in its measurement, brain atrophy has been
consistently associated with depression in MS, especially
when involving the temporal grey matter and hippocampus.
Diffusion tensor imaging
Other MRI studies based on diffusion tensor imaging (DTI)
suggested that the so-called normal-appearing white matter
Neurol Sci (2021) 42:835–845
(NAWM) was far to be “normal”. This technique permitted
the estimation of the micro-pathological changes in normal-
appearing brain tissues, measuring the water proton diffusiv-
ity, which is described by two main indices: the fractional
anisotropy (FA) and the mean diffusivity (MD). High values
of FA, which represents the degree of anisotropy underlying
the water diffusion, characterize the healthy white matter,
whereas a decrease in FA suggests the loss of axons’ integrity.
The MD indicates the rotationally invariant magnitude of wa-
ter diffusion and can be used as a measure of the overall level
of myelination and density of the tissue.
Feinstein et al. [31] screened 62 PwMS for depression, 30
of which classified as being depressed on the basis of a BDI-
revised score greater than 19, and assayed them with T1W,
T2W sequences and DTI. Smaller NAWM in the left superior
frontal region and greater hypointense T1W lesion volume in
the right medial inferior frontal region were detected in de-
pressed patients. Noteworthy, DTI revealed the presence of
lower FA in NAWM and higher MD in normal-appearing
grey matter (NAGM) of the left anterior temporal lobe.
These data asserted that also the micro-pathology of the so-
called NAWM and NAGM might play a crucial role in the
development of depression in MS.
Such a finding was confirmed in a further work [32], show-
ing lower FA in 15 patients affected by RRMS when com-
pared with 15 healthy individuals. In PwMS, the HDRS score
was also positively correlated with FA values in both grey (left
hypothalamus, right posterior middle cingulate gyrus and hip-
pocampus) and white matter lesions (right precentral gyrus
and posterior cingulate gyrus).
A DTI analysis with the study of the whole-brain structural
connectivity was performed by Nigro et al. [33]. An increased
local path between the right hippocampus and amygdala and
an increased shortest distance in the connection between them
and some regions of the prefrontal cortex were detected.
Seemingly, these results proposed that the MS-related struc-
tural brain impairment might interfere with the information
flow between these cerebral structures, namely deputed to
the emotions modulation.
In a larger study [29], frontal and fronto-temporal pathway
damage was linked with depression and fatigue in 147 PwMS
and 90 gender-matched healthy controls. Reduced FA in the
forceps minor was detected in the depressed and fatigued co-
hort (65 PwMS). Tract-based spatial statistics (TBSS) analy-
sis, an automated method to augment the interpretability of the
FA data deriving from different subjects, was used.
In a recent study by Rojas et al. [34], 23 depressed, 22 non-
depressed PwMS (classified on the basis of BDI-II) and 20
healthy controls were compared through T1W, T2W, DTI
sequences and TBSS analysis. PwMS with depressive symp-
toms had similar values of FA and white matter volume in
comparison with both the non-depressed cohort and the
healthy controls.
Neurol Sci (2021) 42:835–845
Most of these advanced studies found a link between de-
creased FA within the white matter (evidence of loss of axons’
integrity) and depression, thus suggesting a disconnection
mechanism at the origin of this disease.
Functional magnetic resonance imaging
Functional magnetic resonance imaging (fMRI) emerged as a
pivotal technique to expand the knowledge on the physiopath-
ological basis of depression in MS as it can map neural activ-
ity in the brain or spinal cord using the blood-oxygen-level-
dependent contrast to detect changes associated with blood
flow. This registration is usually performed during a visual,
cognitive or sensorimotor task.
An easier way of collecting functional data is represented
by resting-state fMRI (rs-fMRI). This type of fMRI is used in
brain mapping to evaluate regional interactions occurring in a
resting state and to study the functional connectivity (FC),
which is a measure of the functional communication between
different areas of the brain. With rs-fMRI, resting-state net-
works (RSN), namely areas with similar functional character-
istics, are recognized.
Using this technique, Rocca et al. [35] investigated hippo-
campal dysfunction in the pathogenesis of MS-related depres-
sion. When compared with the 42 healthy controls, the 69
PwMS had significantly lower resting-state FC in the default
mode network (DMN), a medial cortical RSN involving the
hippocampus and several frontal-parietal areas only at rest. In
PwMS, the bilateral decrease in hippocampal FC was corre-
lated to more severe depressive symptoms (measured with
MADRS). In particular, depressive symptoms were associated
with the reduction of resting-state FC in the orbitofrontal cor-
tex, middle temporal gyrus and precuneus.
In cognitively preserved depressed PwMS, a decrease in
FC was present not only in the DMN but also in other RSNs,
such as the salience network and executive control network,
where both the anterior and posterior cingulate cortex were
involved [36].
In a recent retrospective study, the data of two groups of
PwMS taken from two different studies who underwent rs-
fMRI have been analyzed [37]. Twenty-two PwMS were de-
fined as depressed on the basis of BDI-II and compared with
21 PwMS considered non-depressed on the basis of HADS. In
the depressed group, lower FC between the amygdala and the
frontal regions was reported.
During an emotional processing “face task”, an increased
activation of the right ventrolateral prefrontal cortex was de-
tected in PwMS in respect to healthy controls, as well as in
depressed versus non-depressed PwMS [38]. In addition, FC
was altered between the left amygdala, dorsolateral and ven-
trolateral prefrontal cortex, with a negative correlation be-
tween MS depression scores (BDI-II and BDI-fast screen
(BDI-FS)) and the connectivity among left hippocampus and
839
orbitofrontal and dorsolateral prefrontal cortex, right amygda-
la and left dorsolateral prefrontal cortex.
When considering depression and FC modifications, a de-
crease in FC has been consistently detected. Consequently, it
has been theorized that a causal link between depression and
brain damage in MS might be the interruption of the connec-
tions between similar areas of the brain. In particular, the
impairment of a specific neural network within the limbic
system might represent an important finding in this regard.
Thus, depression might be the effect of a sort of maladaptive
mechanism intervening in these structures, deputed to emo-
tional control.
Other lines of research
Even the hypothalamic-pituitary-adrenal axis (HPAA) has
been involved in the control of the disease process of MS,
considering the role of the structure and the neuroendocrine
alterations in the development of depression in PwMS.
Although the hypothesis of a reduced activation of the
HPAA has been postulated in PwMS (paralleling the findings
in experimental autoimmune encephalomyelitis, the mouse
model of MS), the available data seem to provide evidence
of a chronic activation of this cerebral structure, conceivably
in response to the production of pro-inflammatory cytokines
in MS lesions [39, 40].
Also, the olfactory bulb volume was claimed to have a part
in MS-related depression. In patients with progressive forms
of MS, Yaldizli et al. [41] presented the correlation between
lower olfactory bulb volume and higher depression scores
(assessed with CES-D). These results are in line with previous
studies demonstrating that depressed subjects without MS
have lower olfactory bulb volumes and that PwMS have a
decreased olfactory function in association with higher de-
pressive scores [42]. Although the underlying mechanisms
have not been elucidated yet, the main theory supposes that
the olfactory bulb dysfunction might represent a step towards
depression due to its connection with the amygdala and the
limbic system in general.
Discussion
PwMS are more predisposed to suffer from depression than
the general population [1]. In these subjects, the brain lesions
embody the expression of the inflammation and demyelin-
ation peculiar of a neuro-inflammatory disease such as MS.
In turn, according to the studies considered, brain damage
might represent one of the determinants of depressive status
(see Table 1 and Table 2).
This might be in line with the findings in patients without
MS, where an emerging evidence suggests that chronic low-
840 Neurol Sci (2021) 42:835–845

Table 1 Details of the studies using MRI for the investigation of the relation between depression and brain damage in PwMS

Study Patients MRI Psychometric Results regarding depression

instruments

Honer et al. 8 MS (with psychiatric disorders) vs 8 T2W DSM-III Higher total T2W lesion volume in temporal lobes in
[6] MS (without psychiatric disorders) the psychiatric disorders cohort.
Reischies 46 MS with clinical relapse T2W Ad hoc 7-point Correlation between cerebral lesions and psychological
et al. [7] scale for symptoms score. More severe symptoms are related
symptoms of to periventricular and frontal lesions.
depression
Ron et al. 116 MS vs 48 disabled subjects (with Spin echo, SCID; BDI-II Greater T2W lesion score involvement in the
[8] rheumatologic or other neurologic inversion temporal-parietal regions in those with flattened
disorders) vs 40 HC recovery mood.
Millefiorini 18 RRMS (13 depressed) T1W, T2W SCID Comparable lesion load between the depressed and
et al. [9] non-depressed group.
Moeller 25 MS (6 depressed) T2W, Gad+ SCID, MADRS, Presence or absence of Gad+ lesions not correlating to
et al. [16] HADS depression.
George 16 MS (8 depressed) T1W HDRS Correlation between high cortical white matter left
et al. [11] hemisphere lesions and HDRS score. In the
depressed group: higher number of brain lesions.
Sabatini 20 RRMS (10 depressed) T1W, T2W, BDI-II, HDRS No differences in regional distribution of lesions in
et al. [10] SPECT, proton depressed and non-depressed patients.
density-- Higher cerebral blood flow in the left limbic cortex of
weighted MRI depressed patients vs right limbic cortex of
non-depressed patients.
Correlation between BDI and HDRS and perfusion
asymmetry.
Pujol et al. 45 MS (31 RRMS, 14 PPMS) T2W BDI-II Correlation between demyelinating lesions in the
[12] arcuate fasciculus and BDI.
Pujol et al. 45 MS (31 RRMS, 14 PPMS) T2W BDI-II Demyelinating lesions of the arcuate fasciculus
[13] correlated to “affective symptoms” and “somatic
complaint” of BDI-II.
Berg et al. 78 MS (37 RRMS, 33 SPMS, 8 T2W DSM-IV, HDRS, Correlation between depression and T2W lesion in right
[14] PPMS) MADRS, BfS, temporal lobe; higher T2W lesion load in the right
BDI-II frontal lobe.
Bakshi 48 MS (33 RRMS, 15 SPMS) T1W, T2W, DSM-IV, HDI, Correlation between the presence of depression and
et al. [15] Proton density, BDI-II T1W lesions in superior frontal and superior parietal
FLAIR regions.
T2W lesions did not correlate to depression.
Zorzon 95 MS vs 97 patients with chronic T1W, T2W DMS-IV, HDRS Correlation between right frontal lesions and HDRS
et al. [20] rheumatic disease vs 110 HC score.
HDRS score correlated with the decreased temporal
brain and right hemisphere brain volume.
Zorzon 90 MS (18 depressed) T1W, T2W HDRS HDRS score correlated with right temporal atrophy.
et al. [21]
Feinstein 40 MS (21 depressed vs 19 T1WI, T2W DSM-IV In the depressed cohort, more extensive T2W
et al. [22] non-depressed) hyperintense and T1W hypointense lesions in the left
medial frontal region, less grey matter volume and
atrophy of the left anterior temporal region.
Kallaur 150 MS vs 249 HC T1W, T2W, Gad+ HADS Depression is inversely associated with Gad+ lesions.
et al. [17]
Rossi et al. 405 RRMS (78 “active signal at Gad+ BDI-II “Active” patients had a higher frequency of depressive
[18] MRI”—29 experiencing clinical symptoms; “active” patients with clinical relapse or
relapse—vs 327 “not active at not were comparable in terms of BDI-II.
MRI”) Lower BDI-II scores in concomitance with the
remission of clinical relapse
Feinstein 62 MS (30 depressed) T1W, T2W, DTI BDI-II Depression related to:
et al. [31] - Smaller NAWM in the left superior frontal region.
- Greater hypointense T1W lesion volume in the right
medial inferior frontal region.
Neurol Sci (2021) 42:835–845 841

Table 1 (continued)

Study Patients MRI Psychometric Results regarding depression

instruments

- Lower fractional anisotropy in NAWM and Higher

mean diffusivity in NAGM of the left anterior
temporal lobe.
Shen et al. 15 RRMS vs 15 HC T2W, DTI HDRS Lower fractional anisotropy in the right posterior
[32] middle cingulate gyrus, the right hippocampus, the
left hypothalamus, the right precentral gyrus and the
posterior cingulate correlated with depression.
Nigro et al. 20 depressed MS vs 22 non-depressed DTI DSM-IV, BDI-II In depressed PwMS:
[33] MS vs 16 HC - Increased local path length in the right hippocampus
and right amygdala
- Increased shortest distance between both the right
hippocampus and right amygdala and the
dorsolateral and ventrolateral prefrontal cortex
orbitofrontal cortex, sensory-motor cortices and
supplementary motor area.
Gobbi et al. 147 MS (65 depressed) vs 90 HC T1W 3D MRI, MADRS No differences between depressed and non-depressed
[29] DTI MRI PwMS for T1W and T2W lesions and white matter
atrophy.
Depression selectively related to atrophy of the left
middle frontal gyrus and right inferior frontal gyrus.
Stuke et al. 44 RRMS and PPMS T1W MPRAGE BDI-II Cross-sectional: association between left temporal lobe
[23] (cross-sectional); 40 RRMS and MRI, atrophy and depressive symptoms.
PPMS (longitudinal) voxel-based Longitudinal: significant relation between the
morphometry worsening of depression and growing atrophy in
both the left thalamus and right internal globus
pallidus. Reduction of grey matter in the right
cingulate cortex and frontal lobe.
Pravatà 126 MS (87 RRMS, 21 Benign MS, T1W 3D MADRS Decrease in cortical thickness in bilateral
et al. [24] 14 SPMS, 4 PPMS) vs 59 HC fronto-temporal regions. Frontal cortical thinning
predicted depression. MADRS severity correlated
with right entorhinal cortex thinning.
Nygard 61 RRMS vs 61 HC T1W MRI, FLAIR BDI-II Negative association between depressive symptoms
et al. [25] and cortical surface area and volume of the frontal
and parietal lobes.
Gold et al. 29 RRMS vs 20 HC T1W, FLAIR BDI-II Smaller volumes of the hippocampal dentate gyrus.
[27] Smaller mean hippocampal volume.
Kiy et al. 72 MS (64 RRMS, 8 PPMS) vs 16 HC T1W BDI-II Correlation between larger left temporal horn volume
[28] and psychic and somatic items of BDI-II.
Gold et al. 138 female MS (83 low grade of T1W, T2W, Gad+ CES-D Highly depressed subjects had significantly smaller
[30] depression vs 26 high grade of right hippocampal volume. Significant correlation
depression) between depressive affect and right and total
hippocampal volume but not left hippocampus.
Rocca et al. 69 RRMS vs 42 HC T1W, T2W, MADRS Lower resting-state functional connectivity in the
[35] rs-fMRI DMN. Decrease in the bilateral hippocampal
functional connectivity. Significant association
between depression and reduction of resting-state
functional connectivity in the orbitofrontal cortex,
middle temporal gyrus and precuneus.
Bonavita 16 depressed MS vs 17 non-depressed rs-fMRI CMDI Disequilibrium in the functional connectivity in DMN.
et al. [36] MS vs 17 non-depressed HC vs 15 Decrease in functional connectivity involved both
depressed subjects the anterior and posterior cingulate cortex, salience
network and executive control network.
Riccelli 77 RRMS (22 depressed) vs 20 HC fMRI BDI-II, BDI-FS Increased activation of the right ventrolateral prefrontal
et al. [38] cortex. Altered connectivity between the left
amygdala and dorso- and ventrolateral prefrontal
cortex. Negative association between depression
scores in PwMS and the connectivity among left
hippocampus and orbitofrontal and dorsolateral
842
Table 1 (continued)
Study Patients MRI
Yaldizli 103 RRMS, 35 SPMS, 8 PPMS Spin echo, TW1,
et al. [41] T2W, proton
density
Rojas et al. 45 RRMS (23 depressed) vs 20 HC T1W, T2W, Gad,
[34] MPRAGE 3D,
DTI
van Geest 43 MS (22 depressed) vs 12 HC T1W, T2W, DTI,
et al. [37] rs-fMRI
MS, multiple sclerosis; RRMS, relapsing-remitting MS; PPMS, primary progressive MS; SPMS, secondary progressive MS; HC, healthy controls; MRI,
magnetic resonance imaging; fMRI, functional MRI; rs-fMRI, resting-state fMRI; DTI, diffusion tensor imaging; FLAIR, fluid-attenuated inversion
recovery; DMN, default mode network; BDI-II, Beck Depression Inventory; BDI-FS, BDI-fast; HDRS, Hamilton Depression Rating Scale; HADS,
Hospital Anxiety and Depression Scale; MADRS, Montgomery and Asberg Depression Scale; BfS, Befindlichkeitsskala; CES-D, Center for
Epidemiologic Studies-Depression Scale; CMDI, Chicago Multiscale Depression inventory; SCID, Structured Clinical Interview DSM-IV; NAWM,
normal-appearing white matter; NAGM, normal-appearing grey matter
grade inflammation might contribute to the onset and mainte-
nance of depression [44].
In PwMS, rather than an independent pathology, or a mere
consequence of the disability derived from the primary dis-
ease, depression appears to be intrinsically linked to the struc-
tural and functional brain damage. However, this has not to be
considered as the exclusive cause, but a part of a multifactorial
mechanism in which the MS-related cerebral impairment
plays a proper role.
In fact, increasing signals convey the attention on this in-
teresting hypothesis, although the results of the studies are not
univocal.
White matter lesion load and localization, in particular in
the temporal and frontal lobes, have been correlated to the
presence of depression. When considering the advanced meth-
odologies, atrophy (mainly of the temporal lobe) and the
micro-pathological changes in NAWM and NAGM detected
with DTI were associated with depression.
In regard to fMRI, the findings suggest that a “disconnec-
tion syndrome” exists in MS and that depression might be the
consequence of a sort of maladaptive mechanism.
With respect to the structure involved, the limbic system, in
particular the hippocampus, and the frontal and temporal re-
gions seem to be the brain parts most implicated in depression.
These sites are delegated to emotional control and integrate
different stimuli. It has to be noted that in fMRI studies,
PwMS presented an abnormal communication between these
areas even in the absence of depression. Reflecting on these
premises, the “disconnection syndrome” might feasibly be the
reason for which PwMS result to be more susceptible to
depression.
Neurol Sci (2021) 42:835–845
Psychometric Results regarding depression
instruments
prefrontal cortex, right amygdala and left dorsolateral
prefrontal cortex.
CES-D Correlation between the olfactory bulb volume and
depression scores in progressive MS.
BDI-II Lower total brain volume, normal grey matter volume,
increased T2W lesion load in depressed PwMS. No
differences in global fractional anisotropy and in
white matter volume between groups.
BDI-II, HADS-D Lower white matter volume, decreased fractional
anisotropy of the uncinated fasciculus and lower
functional connectivity between the amygdala and
the frontal regions in the depressed cohort.
However, it remains a challenging matter to definitely
judge whether neurological damage (which should be consid-
ered in any case as a part of a multifactorial physiopatholog-
ical mechanism), is or not the trigger for depression in PwMS.
The reasons behind this concern can be expressly explained
by the methods of the studies (Table 2).
For instance, the diagnosis of depression is not always
reliable in the most part of the works examined. In regard
to depression, the semi-structured interviews and the scor-
ing scales are the main methods of evaluation. Only the
first methodology permits to obtain a diagnosis of depres-
sion and must be administered by a trained physician.
Conversely, the scoring scales are used to score depres-
sive symptoms and might overestimate depression when
used to diagnose it. In fact, a number of items related to
physical symptoms are present [1]. This can be confusing
when evaluating PwMS, often exposed to fatigue, pain
and sleep problems. Thus, the high variability of evalua-
tion instruments negatively influences the reproducibility
and comparability of the studies.
In addition, some of the considered reports have very small
samples, which provide the possibility of a type II error.
Instead, other studies have a retrospective nature, not allowing
to infer direct causality links.
In association with these methodological problems, it must
not be forgotten that depressive symptoms and depression
itself might arise directly from the adverse conditions which
can be encountered in life. This is mainly true in the case of
PwMS who find themselves thrown into an “ab initio” pro-
gressive disease, with all the uncertainties in terms of disabil-
ity, loss of autonomy and of social participation.
Neurol Sci (2021) 42:835–845 843

Table 2 Summary of the positive and negative findings of the studies the formal lack of longitudinal studies, considering the onset
on depression and MRI alterations in PwMS and related comments
and progression of MS and depression in relation to lesion
Positive/negative reports Comments load and brain pathology, does not permit to unquestionably
attribute to brain damage a clear responsibility in depression
T2W lesions 7 studies [11–14, 20, 22, 34] - Small sample [9–11] and depressive symptoms onset and maintenance.
correlated T2W lesions in - Depression diagnosis not
frontal and temporal according to DSM
Another important point is represented by the population
lobes with depression or [9–11, 29, 34] recruited for the studies, which might lead to ponder a possible
depressive symptoms/5 - MS forms not specified selection bias. As a matter of fact, the samples are often com-
studies [9, 10, 15, 29, 35] [11] posed of individuals with different MS forms (see Table 1 for
did not. - Study in the early phases
of MS [9]
details), which have a diverse incidence of depression and
- No sub-analysis for MS dissimilar MRI pattern, as it has been already acknowledged.
forms [12–15, 22, 43] Unfortunately, these problems are sometimes the results of the
T1W lesions 3 studies [15, 22, 31] - Depression diagnosis not high costs of these kinds of studies, mainly deriving from MRI
correlated T1W lesions in according to DSM [31, utilization.
frontal lobes with 34]
depression or depressive - No sub-analysis for MS
In addition, when specifically reflecting on atrophy, the
symptoms. forms [15, 22] findings of the studies are not comparable as no unified pro-
1 study [34] did not. tocol is available, augmenting the variability in the method of
Gad + lesions 1 study [18] correlated - Small sample [16] assessment.
Gad+ lesions to - Depression diagnosis not
depression/1 study [17] according to DSM [17,
However, apart from the techniques used, it has to be pon-
inversely correlated Gad+ 18] dered that it might be easier to correlate depression with the
lesions to depression/1 - MS forms not specified lesion load in a specific area of the brain, than with the lesion
study [16] did not. [16, 17] load in the whole brain, for the simple reason that the study of
Atrophy 11 studies correlated - Small sample [22, 27] a more specific area has been performed.
atrophy in temporal [20, - Depression diagnosis not
21, 23, 28], according to DSM [21,
The lifetime prevalence of depression in MS is about 50%,
frontal-temporal lobes 24, 25, 27, 28, 37] and its social burden represents a matter of paramount impor-
[22, 24, 25] and hippo- - Use of two different tance, underlining the necessity for further strategies in the
campus [27, 30, 37, 43] to scales for determination prevention and treatment of cerebral damage. At the present
depressive symptoms or of depressed and
time, although an antidepressant function might have been
depression/0 study did non-depressed PwMS
not. [37] attributed to some disease-modifying drugs, and some antide-
DTI 4 studies correlated lower - Small sample [32] pressants have been proposed as potential “neuroprotectors”,
fractional anisotropy in - Depression diagnosis not the data available are not so comforting [45].
the temporal lobe [31] according to DSM [31,
and hippocampus [32, 33, 32, 34, 37]
37]/1 study did not [34]. - Use of two different
scales for determination Conclusions
of depressed and
non-depressed PwMS Taken together, all the observations suggest that depression in
[37]
PwMS might not be a simple consequence of disability but, at
fMRI 4 studies [35–38] correlated - Depression diagnosis not
least in part, a result of the multi-faceted brain injury charac-
a decrease of functional according to DSM
connectivity to [35–38] terizing MS. This damage includes focal damage, diffuse im-
depressive symptoms or - Use of two different pairment of white and grey matter and the subtle harm in
depression/0 study did scales for determination normal-appearing brain tissues, as well as the reduction in
not. of depressed and
FC. Currently, we do not fully comprehend the authentic re-
non-depressed PwMS
[37] lation between depression and the neuro-inflammatory disease
which is MS, although it appears to be a complex multifacto-
rial mechanism in which brain damage seems to play an im-
portant role.
Thus, there is an urgent need for studies which might boost
It is also well known that the individual cognitive styles our understanding of the matter, in particular with a prospec-
might exert a pivotal role in the aetiology of depression [1]. If tive longitudinal design. The advantage of this type of re-
this is the case, the lack of consideration of the premorbid search would consist of a more precise analysis of the clinical
personality traits and of the possibility of a pre-existing psy- and MRI changes over time. Short- and long-term longitudi-
chiatric morbidity might constitute a matter of crucial interest nal studies involving selected PwMS and considering the
in evaluating the results of the studies. On the other hand, also, premorbid personality traits and depressive status might
844
enhance the opportunity to understand if depression might be
a consequence of the disease since the onset. Furthermore,
studies taking into account the lesions’ evolution and the de-
pressive status onset or worsening might explore whether a
direct link between brain pathology and depression might be
present or not. In addition, the use of high-field MRI might
enhance the possibility to understand properly the relation
between the damage of the central nervous system in MS
and depression.
Another piece of the jigsaw is represented by the research
on new prospective neuroprotective and anti-inflammatory
pharmacological strategies. Hypothetically, these drugs might
be associated with the conventional antidepressants, potential-
ly preventing the structural and functional disease-related in-
jury and decreasing the disability deriving from MS itself and
from depression in PwMS. However, despite these promising
theories, the destination of the journey towards the definitive
knowledge of these issues is still out of sight.
Compliance with ethical standards
Conflict of interest Fabio Giuseppe Masuccio and Giulia Gamberini
have nothing to disclose. Massimiliano Calabrese received honoraria
for research or speaking from Sanofi-Genzyme, Merck-Serono, Biogen
Idec, Teva, Novartis Pharma and Roche and funds for travel from Sanofi-
Genzyme, Merck-Serono, Biogen Idec, Teva, Novartis Pharma and
Roche. Claudio Solaro served on the advisory boards of Biogen Idec
and Merck Serono. He received speaking honoraria from Bayer
Schering, Biogen Idec, Merck Serono, Almirall, Teva and Genzyme.
He received research grants and support from the Italian MS Society
Research Foundation (Fondazione Italiana Sclerosi Multipla).
Ethical approval None.
Informed consent None.
References
1. Solaro C, Gamberini G, Masuccio FG (2018) Depression in multi-
ple sclerosis: epidemiology, aetiology, diagnosis and treatment.
CNS Drugs 32:117–133. https://doi.org/10.1007/s40263-018-
0489-5
2. Solaro C, Trabucco E, Signori A, Martinelli V, Radaelli M,
Centonze D, Rossi S, Grasso MG, Clemenzi A, Bonavita S,
D’Ambrosio A, Patti F, D’Amico E, Cruccu G, Truini A (2016)
Depressive symptoms correlate with disability and disease course
in multiple sclerosis patients: an Italian multi-center study using the
Beck Depression Inventory. PLoS One 11:1–9. https://doi.org/10.
1371/journal.pone.0160261
3. Alba Palé L, León Caballero J, Samsó Buxareu B, Salgado Serrano
P, Pérez Solà V (2017) Systematic review of depression in patients
with multiple sclerosis and its relationship to interferonβ treatment.
Mult Scler Relat Disord 17:138–143. https://doi.org/10.1016/j.
msard.2017.07.008
4. Feinstein A, Magalhaes S, Richard J et al (2014) The link between
multiple sclerosis and depression. Nat Publ Gr 10:507–517. https://
doi.org/10.1038/nrneurol.2014.139
Neurol Sci (2021) 42:835–845
5. Wattjes MP, Rovira À, Miller D et al (2015) Evidence-based guide-
lines: MAGNIMS consensus guidelines on the use of MRI in mul-
tiple sclerosis - establishing disease prognosis and monitoring pa-
tients. Nat Rev Neurol 11:597–606. https://doi.org/10.1038/
nrneurol.2015.157
6. Honer WG, Hurwitz T, Li DKB, Palmer M, Paty DW (1987)
Temporal lobe involvement in multiple sclerosis patients with psy-
chiatric disorders. Arch Neurol 44:187–190. https://doi.org/10.
1001/archneur.1987.00520140053017
7. Reischies FM, Baum K, Bräu H et al (1988) Cerebral magnetic
resonance imaging findings in multiple sclerosis. Arch Neurol 8:
8–10
8. Ron MA, Logsdail SJ (1989) Psychiatric morbidity in multiple
sclerosis: a clinical and MRI study. Psychol Med 19:887–895.
https://doi.org/10.1017/S0033291700005602
9. Millefiorini E, Paduvani A, Pozzili C et al (1992) Depression in the
early phase of MS: influence of functional disability, cognitive
impairment and brain abnormalities. Acta Neurol Scand 86:354–
358
10. Sabatini U, Pozzilli C, Pantano P, Koudriavtseva T, Padovani A,
Millefiorini E, Biasi CD, Gualdi GF, Salvetti M, Lenzi GL (1996)
Involvement of the limbic system in multiple sclerosis patients with
depressive disorders. Biol Psychiatry 39:970–975. https://doi.org/
10.1016/0006-3223(95)00291-X
11. George MS, Kellner CH, Bernstein H, Goust JM (1994) A magnet-
ic resonance imaging investigation into mood disorders in multiple
sclerosis: a pilot study. J Nerv Ment Dis 182:410–412
12. Pujol J, Bello J, Deus J, Marti-Vilalta JL, Capdevila A (1997)
Lesions in the left arcuate fasciculus region and depressive symp-
toms in multiple sclerosis. Neurology 49:1105–1110. https://doi.
org/10.1212/WNL.49.4.1105
13. Pujol J, Bello J, Deus J, Cardoner Ń, Martı-Vilalta JL, Capdevila A
(2000) Beck Depression Inventory factors related to demyelinating
lesions of the left arcuate fasciculus region. Psychiatry Res -
Neuroimaging 99:151–159. https://doi.org/10.1016/S0925-
4927(00)00061-5
14. Berg D, Supprian T, Thomae J, Warmuth-Metz M, Horowski A,
Zeiler B, Magnus T, Rieckmann P, Becker G (2000) Lesion pattern
in patients with multiple sclerosis and depression. Mult Scler 6:
156–162. https://doi.org/10.1177/135245850000600304
15. Bakshi R, Czarnecki CAD, Shaikh ZA et al (2000) Brain MRI
lesions and atrophy are related to depression in multiple sclerosis.
Neuroreport 11:1153–1158. https://doi.org/10.1097/00001756-
200004270-00003
16. Möller A, Wiedemann G, Rohde U, Backmund H, Sonntag A
(1994) Correlates of cognitive impairment and depressive mood
disorder in multiple sclerosis. Acta Psychiatr Scand 89:117–121.
https://doi.org/10.1111/j.1600-0447.1994.tb01497.x
17. Kallaur AP, Lopes J, Oliveira SR, Simão ANC, Reiche EMV, de
Almeida ERD, Morimoto HK, de Pereira WLCJ, Alfieri DF,
Borelli SD, Kaimen-Maciel DR, Maes M (2016) Immune-
inflammatory and oxidative and nitrosative stress biomarkers of
depression symptoms in subjects with multiple sclerosis : increased
peripheral inflammation but less acute neuroinflammation. Mol
Neurobiol 53:5191–5202. https://doi.org/10.1007/s12035-015-
9443-4
18. Rossi S, Studer V, Motta C, Polidoro S, Perugini J, Macchiarulo G,
Giovannetti AM, Pareja-Gutierrez L, Calò A, Colonna I, Furlan R,
Martino G, Centonze D (2017) Neuroinflammation drives anxiety
and depression in relapsing-remitting multiple sclerosis. Neurology
89:1338–1347. https://doi.org/10.1212/WNL.0000000000004411
19. Calabrese M, Magliozzi R, Ciccarelli O, Geurts JJG, Reynolds R,
Martin R (2015) Exploring the origins of grey matter damage in
multiple sclerosis. Nat Rev Neurosci 16:147–158. https://doi.org/
10.1038/nrn3900
Neurol Sci (2021) 42:835–845
20. Zorzon M, de Masi R, Nasuelli D, Ukmar M, Pozzi Mucelli R,
Cazzato G, Bratina A, Zivadinov R (2001) Depression and anxiety
in multiple sclerosis. A clinical and MRI study in 95 subjects. J
Neurol 248:416–421. https://doi.org/10.1007/s004150170184
21. Zorzon M, Zivadinov R, Nasuelli D et al (2002) Depressive symp-
toms and MRI changes in multiple sclerosis. Eur J Neurol 9:491–
496. https://doi.org/10.1046/j.1468-1331.2002.00442.x
22. Feinstein A, Roy P, Lobaugh N et al (2004) Structural brain abnor-
malities in multiple sclerosis patients with major depression.
Neurology 62:586–590. https://doi.org/10.1212/01.wnl.
0000110316.12086.0c
23. Stuke H, Hanken K, Hirsch J et al (2016) Cross-sectional and lon-
gitudinal relationships between depressive symptoms and brain at-
rophy in MS patients. 10:1–8. https://doi.org/10.3389/fnhum.2016.
00622
24. Pravatà E, Rocca MA, Valsasina P et al (2017) Gray matter
trophism, cognitive impairment, and depression in patients with
multipls sclerosis. Mult Scler J 23:1864–1874. https://doi.org/10.
1177/2040620717690316
25. Nygaard GO, Walhovd KB, Sowa P, Chepkoech JL, Bjørnerud A,
Due-Tønnessen P, Landrø NI, Damangir S, Spulber G, Storsve AB,
Beyer MK, Fjell AM, Celius EG, Harbo HF (2015) Cortical thick-
ness and surface area relate to specific symptoms in early relapsing–
remitting multiple sclerosis. Mult Scler J 21:402–414. https://doi.
org/10.1177/1352458514543811
26. Koolschijn PCMP, Van Haren NEM, Lensvelt-Mulders GJLM et al
(2009) Brain volume abnormalities in major depressive disorder: a
meta-analysis of magnetic resonance imaging studies. Hum Brain
Mapp 30:3719–3735. https://doi.org/10.1002/hbm.20801
27. Gold SM, Kern KC, O’Connor MF et al (2010) Smaller cornu
ammonis 23/dentate gyrus volumes and elevated cortisol in multi-
ple sclerosis patients with depressive symptoms. Biol Psychiatry
68:553–559. https://doi.org/10.1016/j.biopsych.2010.04.025
28. Kiy G, Lehmann P, Hahn HK, Eling P, Kastrup A, Hildebrandt H
(2011) Decreased hippocampal volume, indirectly measured, is as-
sociated with depressive symptoms and consolidation deficits in
multiple sclerosis. Mult Scler J 17:1088–1097. https://doi.org/10.
1177/1352458511403530
29. Gobbi C, Rocca MA, Pagani E, Riccitelli GC, Pravatà E, Radaelli
M, Martinelli-Boneschi F, Falini A, Copetti M, Comi G, Filippi M
(2014) Forceps minor damage and co-occurrence of depression and
fatigue in multiple sclerosis. Mult Scler J 20:1633–1640. https://
doi.org/10.1177/1352458514530022
30. Gold SM, Connor MO, Gill R et al (2014) Detection of altered
hippocampal morphology in multiple sclerosis-associated depres-
sion using automated surface mesh modeling. 37:30–37. https://doi.
org/10.1002/hbm.22154
31. Feinstein A, O’Connor P, Akbar N et al (2010) Diffusion tensor
imaging abnormalities in depressed multiple sclerosis patients.
Mult Scler 16:189–196. https://doi.org/10.1177/
1352458509355461
32. Shen Y, Bai L, Gao Y, Cui F, Tan Z, Tao Y, Sun C, Zhou L (2014)
Depressive symptoms in multiple sclerosis from an in vivo study
with TBSS. 2014: https://doi.org/10.1155/2014/148465
33. Nigro S, Passamonti L, Riccelli R, Toschi N, Rocca F, Valentino P,
Nisticò R, Fera F, Quattrone A (2015) Structural ‘connectomic’
845
alterations in the limbic system of multiple sclerosis patients with
major depression. Mult Scler J 21:1003–1012. https://doi.org/10.
1177/1352458514558474
34. Rojas JI, Sanchez F, Patrucco L, Miguez J, Besada C, Cristiano E
(2017) Brain structural changes in patients in the early stages of
multiple sclerosis with depression. Neurol Res 39:596–600.
https://doi.org/10.1080/01616412.2017.1298279
35. Rocca MA, Pravatà E, Valsasina P, Radaelli M, Colombo B,
Vacchi L, Gobbi C, Comi G, Falini A, Filippi M (2015)
Hippocampal-DMN disconnectivity in MS is related to WM le-
sions and depression. Hum Brain Mapp 36:5051–5063. https://
doi.org/10.1002/hbm.22992
36. Bonavita S, Sacco R, Esposito S et al (2017) Default mode network
changes in multiple sclerosis : a link between depression and cog-
nitive impairment?:27–36. https://doi.org/10.1111/ene.13112
37. van Geest Q, Boeschoten RE, Keijzer MJ, Steenwijk MD, Pouwels
PJW, Twisk JWR, Smit JH, Uitdehaag BMJ, Geurts JJG, van
Oppen P, Hulst HE (2019) Fronto-limbic disconnection in patients
with multiple sclerosis and depression. Mult Scler J 25:715–726.
https://doi.org/10.1177/https
38. Riccelli R, Passamonti L, Cerasa A, Nigro S, Cavalli SM, Chiriaco
C, Valentino P, Nisticò R, Quattrone A (2016) Individual differ-
ences in depression are associated with abnormal function of the
limbic system in multiple sclerosis patients. Mult Scler J 22:1094–
1105. https://doi.org/10.1177/1352458515606987
39. Fassbender K, Schmidt R, Mo R, Ku J (2017) Mood disorders and
dysfunction of the hypothalamic-pituitary-adrenal axis in multiple
sclerosis. Arch Neurol 55:66–72
40. Kantorová E, Poláček H, Bittšanský M, Baranovičová E, Hnilicová
P, Čierny D, Sivák Š, Nosáľ V, Zeleňák K, Kurča E (2017)
Hypothalamic damage in multiple sclerosis correlates with disease
activity, disability , depression , and fatigue. Neurol Res 6412:1–8.
https://doi.org/10.1080/01616412.2016.1275460
41. Yaldizli PIK, Yonekawa T et al (2016) The association between
olfactory bulb volume, cognitive dysfunction, physical disability
and depression in multiple sclerosis. Eur J Neurol 23:510–519.
https://doi.org/10.1111/ene.12891
42. Zivadinov R, Zorzon M, Bragadin LM et al (1999) Olfactory loss in
multiple sclerosis. J Neurol Sci 168:127–130
43. Gobbi C, Rocca M, Riccitelli G, Pagani E, Messina R, Preziosa P,
Colombo B, Rodegher M, Falini A, Comi G, Filippi M (2014)
Influence of the topography of brain damage on depression and
fatigue in patients with multiple sclerosis. Mult Scler J 20:192–
201. https://doi.org/10.1177/1352458513493684
44. Leonard BE (2018) Inflammation and depression: a causal or coin-
cidental link to the pathophysiology ? Acta Neuropsychiatr 30:1–
16. https://doi.org/10.1017/neu.2016.69
45. Grech LB, Butler E, Stuckey S, Hester R (2019) Neuroprotective
benefits of antidepressants in multiple sclerosis: are we missing the
mark? J Neuropsychiatry Clin Neurosci. https://doi.org/10.1176/
appi.neuropsych.18070164
Publisher’s note Springer Nature remains neutral with regard to jurisdic-
tional claims in published maps and institutional affiliations.