brain networks: foundations and futures in bipolar disorder
TRANSCRIPT
Brain networks: Foundations and futures in bipolardisorder
LENA PALANIYAPPAN1 & DAVID ANDREW COUSINS2
1Department of Psychiatry, Institute of Neuroscience, Newcastle University, Newcastle upon Tyne, and2Newcastle Magnetic Resonance Centre, Campus for Ageing and Vitality, Newcastle University,
Newcastle upon Tyne, UK
AbstractBackground: Bipolar affective disorder is a common psychiatric illness with an often episodic nature,the neurobiological basis of which remains elusive. Symptom clusters in bipolar disorder can beconceptualized in terms of disordered brain networks, and doing so may aid our understanding of thevaried presentations, differing illness courses and treatment responses.Aims: To review the rationale behind proposed disordered brain network function in bipolar disorderand the evidence of network dysfunction from imaging studies together with an overview of more noveltechniques pertinent to this field.Methods: Medline databases were searched using the terms bipolar disorder, imaging, connectivityand brain networks. Relevant articles were reviewed and bibliographic cross-referencing was used tofocus on key areas of interest, supplemented by additional Medline searches as required.Results: Structural and functional imaging studies support the concept of brain network dysfunctionin bipolar disorder. Novel techniques such as diffusion tensor imaging and resting state networkanalysis can assess such dysfunction more directly, but there are few studies specific to bipolardisorder.Conclusions: Brain network dysfunction is a useful framework for considering the varied presentationsof bipolar disorder. Advanced imaging techniques are increasingly available, with the potential toprovide insights into this important area.
Keywords: Mood disorders, brain imaging, networks
Introduction
Bipolar disorder is a common psychiatric disorder in which abnormalities of mood are found
in association with changes in biological rhythms, cognitive functions and behaviours. As
with most psychiatric disorders, the diagnosis is syndromal and based on clinical observation
and history. There is an implicit assumption in biological psychiatry that disorders of the
mind have a physical basis in brain dysfunction, and consequently brain structure and
function in bipolar disorder has been contrasted with other psychiatric illnesses and
normality. Numerous lines of research support the reduction of clinical syndromes to
neurobiological dysfunction. Brain imaging research is arguably the most accessible to the
clinical psychiatrist, both conceptually and in terms of interpretation. The findings of
Correspondence: Lena Palaniyappan, BA, MBBS, MRCPsych, University Department of Psychiatry, Leazes Wing, Royal Victoria
Infirmary, Queen Victoria Road, Newcastle upon Tyne NE1 4LP, UK. Tel: þ44 (0)191 282 5660. Fax: þ44(0)191 222 6162.
E-mail: [email protected]
Journal of Mental Health,
April 2010; 19(2): 157–167
ISSN 0963-8237 print/ISSN 1360-0567 online � Informa UK Ltd.
DOI: 10.3109/09638230903469129
J M
ent H
ealth
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Mcg
ill U
nive
rsity
on
11/0
4/14
For
pers
onal
use
onl
y.
imaging studies can now be integrated into wider illness models, synthesizing observed
changes in brain structure and function with neurochemical and psychopharmacological
theories. However, as the imaging field has advanced, more complex scanning techniques
have been applied which themselves need consideration before the data can be integrated
into our conceptual frameworks. This paper will approach the imaging field of bipolar
disorder from its foundations in structural work, progressing to functional studies and later
to more advanced techniques. The findings will be considered within our concepts of the
illness and conclude with an examination of novel techniques and possible future
developments.
Clinical features
Bipolar disorder is classically described as a mood disorder in which mania and/or
hypomania alternates with episodes of depression, interspersed with periods of return to
normal function. Mania is characterized by elated, expansive or irritable mood accompanied
by increased drive, engaging in pleasurable high risk activities, motor hyperactivity and
alterations in cognitive function, speech, appetite, libido and sleep patterns. It can be
associated with psychosis. Conversely, depression manifests as persistently low mood, loss of
enjoyment from pleasurable activities and reduced energy levels, together with disturbed
sleep, appetite, libido, and hopelessness and suicidal ideation. In practice there are a great
range of presentations of bipolar disorder, both in terms of episode characteristics and illness
course. Episodic presentation with return to normal functioning can occur, but many
patients experience chronic ill health with prolonged periods of mood disturbance and/or
persistent subsyndromal symptomatology. Episodes themselves can be diverse, including
mixed states and psychotic presentations in addition to classical descriptions of mania and
depression.
It is increasingly recognized that cognitive dysfunction is a feature of bipolar disorder
during both syndromal mood disturbance and remission, with a pronounced effect on
psychosocial functioning. Dysfunction has been reported in a number of cognitive domains
in patients with bipolar disorder and working memory has been investigated in some detail
(Ferrier & Thompson, 2002). Working memory is a hypothetical construct that ‘‘refers to a
brain system that provides temporary storage and manipulation of the information necessary
for such complex cognitive tasks as language comprehension, learning and reasoning’’
(Baddely, 1986). The neural substrates of working memory are well understood and have a
bearing on models of psychiatric illnesses. Manic subjects perform poorly on tests assessing
executive function (Albus et al., 1996; Clark et al., 2001; Martinez-Aran et al., 2004;
McGrath et al., 1997; Morice, 1990; Murphy et al., 1999; Seidman et al., 2002; Sweeney
et al., 2000), verbal and spatial working memory (Clark et al., 2001; Murphy et al., 1999;
Sweeney et al., 2000), and sustained attention. When the latter is investigated, manic
patients are found to have faster reaction times but make more errors, resulting in impaired
overall performance (Clark et al., 2001; Sax et al., 1999). Depressed patients also perform
poorly on such tasks, though not necessarily similarly, and there are differences when
unipolar and bipolar subjects are compared (Wolfe et al., 1987). Bipolar disorder patients
are impaired on other tests of executive function such as verbal fluency (Calev et al., 1989),
and in assessments of verbal memory (Martinez-Aran et al., 2004; Wolfe et al., 1987).
Impairments may persist despite clinical resolution of the episode, and although some of
these deficits may be accounted for by residual symptomatology, abnormalities persist when
these are controlled for (Ferrier et al., 1999). Euthymic bipolar patients perform poorly
compared to controls on tasks of executive function (Thompson et al., 2005; Zubieta et al.,
158 L. Palaniyappan & D. A. Cousins
J M
ent H
ealth
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Mcg
ill U
nive
rsity
on
11/0
4/14
For
pers
onal
use
onl
y.
2001), verbal memory and learning (Cavanagh et al., 2002; van Gorp et al., 1998) and non-
verbal memory in some, but not all studies (Clark et al., 2001; Ferrier et al., 1999; van Gorp
et al., 1998). Deficits in attention are apparent in some tests such as trail making (El-Badri
et al., 2001), but inconsistently so when specifically examined in continuous performance
tasks (Addington & Addington, 1997; Clark et al., 2002; Dixon et al., 2005).
It would seem reasonable to localize many of these symptoms to prefrontal cortical
regions, yet the diversity and complexity of the presentations of bipolar disorder makes it
implausible to identify dysfunctional regions in isolation and conceptually challenging to
construct valid neurobiological illness models. The apparently opposing mood states of
mania and depression, together with their cognitive, emotional and vegetative symptom
clusters seem to involve numerous and seemingly disparate brain regions. Indeed, imaging
studies have implicated a myriad of brain structures and highlighted many regions as
potential sites of derangement using functional techniques (Strakowski et al., 2005). It is
increasingly common to discuss the origins of symptoms in bipolar disorder as having a basis
in dysfunction in interconnected neural systems or networks (Adler et al., 2006). Brain
networks are invoked to explain complex functions such as emotions and cognitions,
behaviours and drives. Reciprocal loops connecting cortical areas, typically via subcortical
relays, have an established anatomical basis and appear to be valid substrates of higher brain
functions on empirical testing (Cummings, 1993). Network dysfunction is a useful way of
conceptualizing bipolar disorder, providing an interim step on the reduction of clinical
features to brain function. Disordered networks arguably reduce the plurality in
explanations of disordered brain structure in bipolar disorder, and may inform our
understanding of basic science data and treatment effects, additionally offering a framework
for future investigations.
The components of key networks will be considered, highlighting important findings from
the structural imaging field, followed by evidence of functional derangement. The models
will briefly be overlaid onto neurochemical and treatment theories before discussing the
steps necessary to consolidate these models.
Structural volumetric changes in bipolar disorder
Volumetric imaging studies in bipolar disorder have recently been subjected to meta-
analysis (Kempton et al., 2008; McDonald et al., 2004) and, although increased ventricle
volume and reduced corpus callosum area were shown to be significant, a consistent pattern
of abnormality in other brain structures is not clear. This may in part be due to study
heterogeneity as medication effects, variable mood states and the clinical course of the
subjects investigated are difficult to control for. Further, advanced image analysis tools such
as voxel based morphometry (VBM) do not lend themselves to inclusion in a meta-analysis,
preventing a comprehensive synthesis of that body of data (Kempton et al., 2008). Caveats
aside, there is a consensus that the hippocampus is of normal volume in bipolar disorder, in
contrast to the volume loss in unipolar depression. There are mixed reports of increased and
decreased amygdala volumes, which may be sensitive to the age (Doty et al., 2008) or
developmental status of the subjects investigated. Additionally, differences in amygdala
volume across age groups may be a consequence of the stress of the illness process, as the
amygdala and hippocampus may be particularly susceptible to the effects of repeated
elevations of cortisol (Rainnie et al., 2004). Amygdala glial cell loss is more clearly
established in unipolar disorder than bipolar disorder (Bowley et al., 2002; Doty et al.,
2008), possibly because of the proposed neuroprotective effect of drugs such as lithium and
valproate that are commonly used in the latter. The prefrontal cortex has been shown to be
Foundations and futures in bipolar disorder 159
J M
ent H
ealth
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Mcg
ill U
nive
rsity
on
11/0
4/14
For
pers
onal
use
onl
y.
smaller in bipolar disorder compared to controls, though not consistently so. The
contradictions may be due to comparisons of global regions, and analysis of specific regions
may be more informative (Strakowski et al., 2000). The subgenual and anterior cingulate
have been shown to be prone to cortical thinning in bipolar disorder and decreased grey
matter in the left anterior cingulate has been observed (Drevets et al., 1998). However, the
cingulate as a whole may be of normal volume (Formto et al., 2008), and there are reports of
increased grey matter (Adler et al., 2005). Subcortical structures have also received
attention, forming as they do an important part of the proposed network systems. The
volume of the caudate in bipolar disorder in unclear as both normal (Aylward et al., 1994;
Brambilla et al., 2001; Strakowski et al., 1993, 2002; Swayze et al., 1992) and increased
volumes have been observed (Aylward et al., 1994; Noga et al., 2001; Sax et al., 1999), with
a further suggestion that volume reduction is a feature of older and late onset groups (Beyer
et al., 2004). The putamen volume may be state dependent, being largely normal in
euthymia and increased during mania (Aylward et al., 1994; Brambilla et al., 2001; Swayze
et al., 1992).
Functional imaging studies in bipolar disorder
The functional imaging studies of bipolar patients have been reviewed and caveats such as
differences in imaging modality, task design, rest conditions and medication effects have
been discussed (Strakowski et al., 2005). Many authors conclude that candidate areas in
brain networks show evidence of dysfunction during illness, though the exact nature of the
changes will require further investigation (Adler et al., 2006). There are mixed reports of
prefrontal cortex dysfunction in bipolar depression and mania. Increased signal in the
frontal lobes of depressed subjects has been reported and conversely a reduction in mania
(Blumberg et al., 2003). Presumably, this reduction in activity translates as a reduction in
frontal inhibition of behaviour in mania (Foland et al., 2008) and the reverse in depression.
Although subject to some inconsistencies, results from functional studies generally suggest
increased limbic activation in the depressed phase (Adler et al., 2006). Subcortical
structures are also functionally implicated, with an increased blood flow to the basal ganglia
seen in mania (O’Connell et al., 1995). This may be related to motor presentations in
addition to cognitive dysfunction.
Demonstrating causality
The validity of network models of bipolar disorder can be considered from a number of
angles. The models hold face validity, since the networks proposed are thought to govern the
normal emotional and cognitive processes that are then disturbed in the illness (Cummings,
1993; Phillips, 2006). There are structural changes demonstrable by imaging and
pathological investigation in the brain areas thought to give rise to complex emotions and
cognitions (Haldane & Frangou, 2004). As discussed, functional changes are apparent in the
areas structurally implicated, with opposing abnormalities in different poles of the illness.
Further, neuronal transmission in the proposed networks is conducted through brain
chemicals such as dopamine, serotonin, GABA and glutamate. The monoamine
neurotransmitter systems have long been implicated in the genesis of bipolar symptomatol-
ogy, and are common sites of action of mood stabilizers such as lithium, valproate,
carbemazepine and the atypical antipsychotics (Cousins & Young, 2007). Dysfunction in
excitatory neurotransmitter systems may underlie cell loss and disturbed synaptic plasticity
in key components of brain networks, such as the amygdala and hippocampus (Doty et al.,
160 L. Palaniyappan & D. A. Cousins
J M
ent H
ealth
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Mcg
ill U
nive
rsity
on
11/0
4/14
For
pers
onal
use
onl
y.
2008; Rainnie et al., 2004). The structure and function of the oligodendrocytes that govern
network myelination, and so neuronal conduction, are disrupted by the loss of NRG1 in
animal models (Fields, 2008; Roy et al., 1989). Dopaminergic networks are disrupted in
such models, and it is notable that NRG1 is a replicated susceptibility gene for bipolar
disorder (Georgieva et al., 2008) with a potential impact on white matter integrity
(McIntosh et al., 2008).
However, despite such strands of supporting evidence, the argument that network
dysfunction drives the presentation of bipolar disorder is largely one of inference. It seems
intuitive that structural and functional changes in candidate regions should be linked, but
the accumulated evidence thus far is not direct. A number of recent advances in imaging
technology have occurred that allow the integrity of networks to be visualized more directly,
and the nature of the pathology to be gauged.
White matter structural changes denoting a pathological process
Brain networks can be considered as cortical and subcortical regions interconnected by
specific white matter tracts. The integrity of white matter is crucial to network function, and
is itself amenable to investigation in patients with bipolar disorder. Hyperintense areas in the
white matter are sometimes seen in patient groups, being particularly well visualized on T2-
weighted images. Such hyperintensities could be due to neuronal damage related to defects
in blood flow as cerebrovascular risk factors (smoking, age, hypertension and diabetes)
correlate positively with the presence of such lesions. One of the most consistent findings in
the imaging field of bipolar disorder is that patients generally have more hyperintensities in
subcortical regions and the deep white matter than controls (2.5 times more likely)
(Kempton et al., 2008). Some predominance in frontal and parietal regions has been noted,
suggesting a disruption in fronto-limbic or fronto-parietal networks in bipolar disorder. Of
note is that white matter hyperintensities are a nonspecific finding observed in a variety of
conditions including major depressive disorder, vascular dementia and Alzheimer’s disease.
Preliminary findings suggest that the predominance of these hyperintensities in bipolar
patients compared to schizophrenia or controls may appear as early as adolescence (Pillai
et al., 2002) though in a group of first episode psychosis patients, no such difference was
noted between affective vs. non affective group (Zanetti et al., 2008).
Arnone et al. (2008) present an argument outlining the importance of the corpus callosum
in bipolar disorder, highlighting its role in inter-hemispheric communication and the
integration of cognitive and emotional processes. Building on the findings of previous
reviews (Kempton et al., 2008), their extended meta-analysis showed a reduction in the area
of the corpus callosum in bipolar subjects compared to controls. The validity of this
observation is strengthened by the lack of heterogeneity and publication bias in the studies
reviewed. A strong case can therefore be made for investigating the structure and integrity of
white matter tracts in bipolar disorder, and a number of MRI techniques now exist to enable
this.
Diffusion tensor imaging to study structural connectivity
Diffusion tensor imaging is a relatively new technique that allows a more direct investigation
of structural white matter integrity. DTI is based on the principle that diffusion of water
molecules in white matter tracts has a directional bias, being greatest along the direction of
the tracts. Damage to the white matter disrupts this directionality, as water can move more
freely out of neurons at sites of pathology (Le Bihan, 2007). Two terms are used to describe
Foundations and futures in bipolar disorder 161
J M
ent H
ealth
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Mcg
ill U
nive
rsity
on
11/0
4/14
For
pers
onal
use
onl
y.
this phenomenon; fractional anisotropy (FA) is a measure of preferential diffusion of water
along axons, and mean diffusivity describes overall water displacement. Using this diffusion
tensor imaging techniques, it is possible to detect subtle white matter changes before
alterations in volume and structure become apparent on standard T1- and T2-weighted
images. Several DTI studies have been conducted in bipolar disorder and altered white
matter integrity has been reported. The major findings include reduced FA in prefrontal
white matter (Adler et al., 2004), higher mean diffusivity in orbitofrontal areas (Beyer et al.,
2005), reduced FA in temporal white matter and increased mean diffusivity in prefrontal
areas (Bruno et al., 2008), increased FA in genu of corpus callosum (Yurgelun-Todd et al.,
2007) and decreased FA in anterior cingulum (Wang et al., 2008).
Tractography is a more sophisticated analysis technique using the data obtained from
DTI, wherein the actual white matter tracts can be plotted out in three dimensions using
probabilistic methods. Charting such fibre tracts could reveal any loss of integrity relative to
each other as well as the site of such defects along a specified tract. Studies investigating
white matter tracts using tractography in bipolar disorder have revealed disturbances in
cingulate-amygdala-hippocampal connections (Houenou et al., 2007), uncinate fasciculus
and anterior thalamic radiation (McIntosh et al., in press).
Analysing functional connectivity
There are different techniques available for studying connectivity using functional magnetic
resonance imaging (fMRI) data (Rogers et al., 2007), in which the image contrast arises
from changes in the blood oxygen level dependent (BOLD) signal. Correlation between
BOLD signal of one region of interest (called ‘‘seed’’) and various other brain areas
(represented as voxels in MRI image) can be used to create functional connectivity maps.
Another method, structural equation modelling, tests prior hypotheses of interconnections
between designated regions of interest using time-series data obtained from fMRI.
Functional connectivity can be investigated using resting state fMRI, though this is a
controversial technique and consensus has not been reached regarding its utility and
interpretation. In a resting fMRI investigation, the subjects are usually asked to lie on the
scanner without performing any mental or physical activity, typically instructed to ‘‘let their
mind go blank’’. It is debatable whether reliable resting states could be achieved inside the
scanner, or indeed if ‘‘resting’’ is the most appropriate term to use for this sort of state. The
evolution of the concept of resting state scanning roughly follows the observation of what
seemed to be background noise in the BOLD signal detected during the rest periods of task
related fMRI. Such ‘‘noise’’ appeared to fluctuate over time and persisted when scanner
variables were controlled for. The most obvious interpretation of such signal patterns was
that they were related to changes in blood flow, with rhythmical patterns arising from cardiac
and respiratory cycles. However, when these are monitored and filtered from the data, low
frequency fluctuations in the signal are still detected and appear to show a degree of
synchronization between distinct brain areas. Brain regions that are synchronized in terms of
these BOLD signal fluctuations are proposed to have functional connectivity. Default
coherent networks have been proposed in the somatosensory, visual, and language
processing regions (Fox & Raichle 2007). Independent component analysis is a technique
whereby resting fMRI data can be separated according to source of the signals, irrespective
of prior knowledge (i.e., not according to expected BOLD signal variation from
hypothesized connections). This allows one to overlay observed baseline connectivity data
onto brain structure to see if functional patterns match anatomical pathways. Compared to
controls, patients with bipolar disorder showed decreased ventral anterior cingulate
162 L. Palaniyappan & D. A. Cousins
J M
ent H
ealth
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Mcg
ill U
nive
rsity
on
11/0
4/14
For
pers
onal
use
onl
y.
connectivity to the amygdalae, thalami and striatum; all key components of mood regulating
circuits (Anand et al., 2008). In a later subset of patients, this connectivity appeared to be
increased from initial readings following 8 weeks treatment with lithium. Bipolar patients
have also been found to have dysregulation of the resting state signal in the medial prefrontal
cortex and hypothalamus (Greenhouse et al., 2006). The prefrontal cortical changes support
the network systems discussed, but interestingly the hypothalamic changes may relate to the
core disturbances of circadian rhythms and drives seen in affective disorders.
Functional connectivity analysis techniques are not restricted to studies of the ‘‘resting
state’’ and are increasingly being applied to data sets acquired during task-related fMRI.
Rich et al. (2008) examined data from a study of childhood bipolar disorder in which
subjects were performing a task of face emotion identification. fMRI time series analysis was
used to study frequencies pertinent to connectivity and the left amygdala was chosen a priori
as a seed region, with connectivity to regions such as the temporal association cortex being
of particular interest. Compared to controls, bipolar subjects were found to have reduced
connectivity between the left amygdala and small regions bordering the posterior cingulate/
precuneus and fusiform gyrus/parahippocampal gyrus on the right side of the brain. This
study arguably provides a functional correlate to the structural studies of amygdala volume
in which developmental abnormalities of maturation have been proposed, and raises the
concern that dysfunction in neural circuitry may be present from the outset of the condition,
or even precede its presentation.
Consolidating findings related to networks
Reductions in activation in a particular brain region observed during an fMRI task may be
due to actual defects at a cortical level in that area, or due to disturbances in the networks for
which that brain region served as a nodal point. Differentiating these possibilities is
problematic, but studies investigating functional and structural connectivity simultaneously
using DTI & fMRI may be the most appropriate route to follow. Altered structure-function
relationship has been shown in one such study of schizophrenia (Schlosser et al., 2007), but
not yet in bipolar disorder.
Connectivity refers to spatial and/or temporal correspondence among brain regions.
Studying temporal or spatial connectivity is important but does not give a complete picture,
providing little information about ‘‘effective connectivity’’ – the influence of one region on
the other. One way of studying this is to consider the brain as a system of multiple inputs and
outputs in which unobserved coupling exists between various regions. When the dynamics
of such a system is perturbed using a psychological task, the metrics of influence of one
region over another can be measured from the response. This approach, called dynamic
causal modelling or DCM (Friston et al., 2003) has been used to analyse fMRI data from
patients with schizophrenia. The results indicated an impaired interregional coupling
between superior temporal cortex and anterior cingulate cortex. The effect of self versus
external speech on this connectivity was reversed in those with auditory hallucinations
compared to those without and to controls (Mechelli et al., 2007). In major depressive
disorder DCM has been used to demonstrate impairments in coupling between dorsal and
rostral anterior cingulate cortex (Schlosser et al., in press). Such models have yet to be
applied to data from patients with bipolar disorder.
It is important to consider potential confounding variables before firm conclusions can be
reached about network abnormalities in bipolar disorder. Medications are likely to have an
effect on assessments of brain networks, and presumably on the functional activity of the
networks themselves. Lithium has already been investigated in this regard (Anand et al.,
Foundations and futures in bipolar disorder 163
J M
ent H
ealth
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Mcg
ill U
nive
rsity
on
11/0
4/14
For
pers
onal
use
onl
y.
2008). Medication effects can be estimated by studying medication-naive patients or
patients who undergo a washout period before the observations. The latter may be ethically
unadvisable while both designs are practically very difficult to conduct. A review of
medication effects on neuroimaging findings in bipolar disorder found variable results with
respect to the influence of medications on functional imaging (Phillips et al., 2008). The
influence of aging on anatomical integrity and functional connectivity of different networks
is yet to be investigated in detail, but could be a potential confounder. Longitudinal studies
on normal healthy subjects to investigate the effect of brain maturation and aging on
neuroimaging parameters are required as a first step. A related variable must also be the
duration of illness and number of episodes experienced, since recurrent stress can have a
deleterious effect on key brain regions within the networks (Bremner, 1999).
Conclusion
Bipolar disorder comprises the seemingly contrasting states of depression and mania,
together with a predisposition to fluctuate between such states. This clinical presentation
lends itself to an aetiological model based on neural circuitry wherein defined structural and
functional abnormalities generate different clinical states that fluctuate and progress on a
longitudinal timeframe. This review has considered some of the evidence for disordered
brain network function in bipolar disorder, with a focus on brain imaging techniques. There
is mounting evidence that the structural and functional integrity of key brain networks is
disrupted in patients with bipolar disorder, and that this may underlie the various
symptomatic components of the condition.
Network models may advance our understanding of bipolar disorder in a number of ways,
ranging from new insights in structural studies to more valid interpretations of complex
functional work. For example, specific examination of white matter tract integrity is coming
to the fore, and techniques such as DTI could aid our interpretation of well documented but
poorly understood abnormalities such as deep white matter hyperintensities. Future work
exploring the temporal patterns of network transmission may unveil areas that are primarily
affected in bipolar disorder, and guide future therapeutic interventions. Potential areas for
further research include exploring the contribution of cortical (top-down) and sub-cortical
(bottom-up) defects and the role of primary dysfunction in fibre tracts themselves.
Such enquiries may help delineating sub-groups with better response to therapeutic
interventions and establish likely neuroprotective effects of some drugs in re-establishing
network function.
The limitations of this traditional and selective review need to be recognized, and future
work utilizing the techniques of meta-analysis may produce new insights. Nevertheless, on
the basis of our brief overview it seems plausible that studying network dysfunction in
bipolar disorder holds the promise of advancing our understanding of this important
condition, with the potential to highlight factors that influence treatment response and
possibly produce suitable biomarkers of this illness.
Declaration of interest: The authors report no conflict of interest. The authors alone are
responsible for the content and writing of the paper.
References
Addington, J., & Addington, D. (1997). Attentional vulnerability indicators in schizophrenia and bipolar disorder.
Schizophrenia Research, 23(3), 197–204.
164 L. Palaniyappan & D. A. Cousins
J M
ent H
ealth
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Mcg
ill U
nive
rsity
on
11/0
4/14
For
pers
onal
use
onl
y.
Adler, C.M., DelBello, M.P., et al. (2006). Brain network dysfunction in bipolar disorder. CNS Spectrums, 11(4),
312–320.
Adler, C.M., Holland, S.K., et al. (2004). Abnormal frontal white matter tracts in bipolar disorder: A diffusion
tensor imaging study. Bipolar Disorders, 6(3), 197–203.
Adler, C.M., Levine, A.D., et al. (2005). Changes in gray matter volume in patients with bipolar disorder. Biological
Psychiatry, 58(2), 151–157.
Albus, M., Hubmann, W., et al. (1996). Contrasts in neuropsychological test profile between patients with first-
episode schizophrenia and first-episode affective disorders. Acta Psychiatrica Scandinavica, 94(2), 87–93.
Anand, A., Wang, Y., et al. (2008). Novel brain imaging studies of resting state mood regulating circuit connectivity
in unmedicated bipolar disorder and effects of lithium treatment. Bipolar Disorders, 10, 6.
Arnone, D., McIntosh, A.M., et al. (2008). Meta-analysis of magnetic resonance imaging studies of the corpus
callosum in bipolar disorder. Acta Psychiatrica Scandinavica, 118(5), 357–362.
Aylward, E.H., Robertstwillie, J.V., et al. (1994). Basal ganglia volumes and white-matter hyperintensities in
patients with bipolar disorder. American Journal of Psychiatry, 151(5), 687–693.
Baddely, A. (1986). Working memory. Oxford: Clarendon Press.
Beyer, J.L., Kuchibhatla, M., et al. (2004). Caudate volume measurement in older adults with bipolar disorder.
International Journal of Geriatric Psychiatry, 19(2), 109–114.
Beyer, J.L., Taylor, W.D., et al. (2005). Cortical white matter microstructural abnormalities in bipolar disorder.
Neuropsychopharmacology, 30(12), 2225–2229.
Blumberg, H.P., Leung, H.C., et al. (2003). A functional magnetic resonance imaging study of bipolar disorder -
State- and trait-related dysfunction in ventral prefrontal cortices. Archives of General Psychiatry, 60(6), 601–
609.
Bowley, M.P., Drevets, W.C., et al. (2002). Low glial numbers in the amygdala in major depressive disorder.
Biological Psychiatry, 52(5), 404–412.
Brambilla, P., Harenski, K., et al. (2001). Anatomical MRI study of basal ganglia in bipolar disorder patients.
Psychiatry Research-Neuroimaging, 106(2), 65–80.
Bremner, J.D. (1999). Does stress damage the brain? Biological Psychiatry, 45(7), 797–805.
Bruno, S., Cercignani, M., et al. (2008). White matter abnormalities in bipolar disorder: A voxel-based diffusion
tensor imaging study. Bipolar Disorders, 10(4), 460–468.
Calev, A., Nigal, D., et al. (1989). Retrieval from semantic memory using meaningful and meaningless constructs
by depressed, stable bipolar and manic patients. British Journal of Clinical Psychology, 28, 67–73.
Cavanagh, J.T.O., Van Beck, M., et al. (2002). Case-control study of neurocognitive function in euthymic patients
with bipolar disorder: An association with mania. British Journal of Psychiatry, 180, 320–326.
Clark, L., Iversen, S.D., et al. (2001). A neuropsychological investigation of prefrontal cortex involvement in acute
mania. American Journal of Psychiatry, 158(10), 1605–1611.
Clark, L., Iversen, S.D., et al. (2002). Sustained attention deficit in bipolar disorder. British Journal of Psychiatry,
180, 313–319.
Cousins, D.A., & Young, A.H. (2007). The armamentarium of treatments for bipolar disorder: A review of the
literature. International Journal of Neuropsychopharmacology, 10(3), 411–431.
Cummings, J.L. (1993). Frontal-subcortical circuits and human behavior. Archives of Neurology, 50(8), 873–880.
Dixon, A.L., Prior, M., et al. (2005). Dopamine antagonist modulation of amphetamine response as detected using
pharmacological MRI. Neuropharmacology, 48(2), 236–245.
Doty, T.J., Payne, M.E., et al. (2008). Age-dependent reduction of amygdala volume in bipolar disorder. Psychiatry
Research: Neuroimaging, 163(1), 84–94.
Drevets, W.C., Ongur, D., et al. (1998). Neuroimaging abnormalities in the subgenual prefrontal cortex:
implications for the pathophysiology of familial mood disorders. Molecular Psychiatry, 3(3), 220–226.
El-Badri, S.M., Ashton, C.H., et al. (2001). Electrophysiological and cognitive function in young euthymic patients
with bipolar affective disorder. Bipolar Disorders, 3(2), 79–87.
Ferrier, I.N., Stanton, B.R., et al. (1999). Neuropsychological function in euthymic patients with bipolar disorder.
British Journal of Psychiatry, 175, 246–251.
Ferrier, I.N., & Thompson, J.M. (2002). Cognitive impairment in bipolar affective disorder: Implications for the
bipolar diathesis. British Journal of Psychiatry, 180, 293–295.
Fields, R.D. (2008). White matter in learning, cognition and psychiatric disorders. Trends in Neurosciences, 31(7),
361–370.
Foland, L.C., Altshuler, L.L., et al. (2008). Evidence for deficient modulation of amygdala response by prefrontal
cortex in bipolar mania. Psychiatry Research-Neuroimaging, 162(1), 27–37.
Formto, A., Malhi, G.S., et al. (2008). Anatomical abnormalities of the anterior cingulate and paracingulate cortex
in patients with bipolar I disorder. Psychiatry Research-Neuroimaging, 162(2), 123–132.
Foundations and futures in bipolar disorder 165
J M
ent H
ealth
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Mcg
ill U
nive
rsity
on
11/0
4/14
For
pers
onal
use
onl
y.
Fox, M.D., & Raichle, M.E. (2007). Spontaneous fluctuations in brain activity observed with functional magnetic
resonance imaging. Nature Reviews – Neuroscience, 8(9), 700–711.
Friston, K.J., Harrison, L., et al. (2003). Dynamic causal modelling. NeuroImage, 19(4), 1273–1302.
Georgieva, L., Dimitrova, A., et al. (2008). Support for neuregulin 1 as a susceptibility gene for bipolar disorder and
schizophrenia. Biological Psychiatry, 64(5), 419–427.
Greenhouse, I., Cohen, B.M., et al. (2006). Combined group ICA and GLM analysis of the resting state network in
bipolar disorder using fMRI. Society for Neuroscience Abstracts. Georgia: World Congress Center.
Haldane, M., & Frangou, S. (2004). New insights help define the pathophysiology of bipolar affective disorder: Neuro-
imaging and neuropathology findings. Progress in Neuropsychopharmacology & Biological Psychiatry, 28(6), 943–960.
Houenou, J., Wessa, M., et al. (2007). Increased white matter connectivity in euthymic bipolar patients: Diffusion
tensor tractography between the subgenual cingulate and the amygdalo-hippocampal complex. Molecular
Psychiatry, 12(11), 1001–1010.
Kempton, M.J., Geddes, J.R., et al. (2008). Meta-analysis, database, and meta-regression of 98 structural imaging
studies in bipolar disorder. Archives in General Psychiatry, 65(9), 1017–1032.
Le Bihan, D. (2007). The ‘wet mind’: Water and functional neuroimaging. Physics in Medicine and Biology, 52(7),
R57–R90.
Martinez-Aran, A., Vieta, E., et al. (2004). Cognitive function across manic or hypomanic, depressed, and euthymic
states in bipolar disorder. American Journal of Psychiatry, 161(2), 262–270.
McDonald, C., Zanelli, J., et al. (2004). Meta-analysis of magnetic resonance imaging brain morphometry studies
in bipolar disorder. Biological Psychiatry, 56(6), 411–417.
McGrath, J., Scheldt, S., et al. (1997). Performance on tests sensitive to impaired executive ability in schizophrenia,
mania and well controls: Acute and subacute phases. Schizophrenia Research, 26(2–3), 127–137.
McIntosh, A.M., Maniega, S.M., et al. (2008). White matter tractography in bipolar disorder and schizophrenia.
Biological Psychiatry, 64(12), 1088–1092.
McIntosh, A.M., Moorhead, T.W.J., et al. (2008). The effects of a neuregulin 1 variant on white matter density and
integrity. Molecular Psychiatry, 13(11), 1054–1059.
Mechelli, A., Allen, P., et al. (2007). Misattribution of speech and impaired connectivity in patients with auditory
verbal hallucinations. Human Brain Mapping, 28(11), 1213–1222.
Morice, R. (1990). Cognitive inflexibility and prefrontal dysfunction in schizophrenia and mania. British Journal of
Psychiatry, 157, 50–54.
Murphy, F.C., Sahakian, B.J., et al. (1999). Emotional bias and inhibitory control processes in mania and
depression. Psychological Medicine, 29(6), 1307–1321.
Noga, J.T., Vladar, K., et al. (2001). A volumetric magnetic resonance imaging study of monozygotic twins
discordant for bipolar disorder. Psychiatry Research-Neuroimaging, 106(1), 25–34.
O’Connell, R.A., Vanheertum, R.L., et al. (1995). Single-photon emission computed-tomography of the brain in
acute mania and schizophrenia. Journal of Neuroimaging, 5(2), 101–104.
Phillips, M.L. (2006 ). The neural basis of mood dysregulation in bipolar disorder. Cognitive Neuropsychiatry, 11(3),
233–249.
Phillips, M.L., Travis, M.J., et al. (2008). Medication effects in neuroimaging studies of bipolar disorder. American
Journal of Psychiatry, 165(3), 313–320.
Pillai, J.J., Friedman, L., et al. (2002). Increased presence of white matter hyperintensities in adolescent patients
with bipolar disorder. Psychiatry Research: Neuroimaging, 114(1), 51–56.
Rainnie, D.G., Bergeron, R., et al. (2004). Corticotrophin releasing factor-induced synaptic plasticity in the
amygdala translates stress into emotional disorders. Journal of Neuroscience, 24(14), 3471–3479.
Rich, B.A., Fromm, S.J., et al. (2008). Neural connectivity in children with bipolar disorder: impairment in the face
emotion processing circuit. Journal of Child Psychology & Psychiatry, 49(1), 88–96.
Rogers, B.P., Morgan, V.L., et al. (2007). Assessing functional connectivity in the human brain by fMRI. Magnetic
Resonance Imaging, 25(10), 1347–1357.
Roy, A., De Jong, J., et al. (1989). Cerebrospinal fluid monoamine metabolites and suicidal behavior in depressed
patients. A 5-year follow-up study. Archives of General Psychiatry, 46(7), 609–612.
Sax, K.W., Strakowski, S.M., et al. (1999). Frontosubcortical neuroanatomy and the continuous performance test
in mania. American Journal of Psychiatry, 156(1), 139–141.
Schlosser, R.G., Nenadic, I., et al. (2007). White matter abnormalities and brain activation in schizophrenia: a
combined DTI and fMRI study. Schizophrenia Research, 89(1–3), 1–11.
Schlosser, R.G. M., Wagner, G., et al. (in press). Fronto-cingulate effective connectivity in major depression: A
study with fMRI and dynamic causal modeling. NeuroImage, 43(3), 645–655.
Seidman, L.J., Kremen, W.S., et al. (2002). A comparative profile analysis of neuropsychological functioning in
patients with schizophrenia and bipolar psychoses. Schizophrenia Research, 53(1–2), 31–44.
166 L. Palaniyappan & D. A. Cousins
J M
ent H
ealth
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Mcg
ill U
nive
rsity
on
11/0
4/14
For
pers
onal
use
onl
y.
Strakowski, S.M., DelBello, M.P., et al. (2000). Neuroimaging in bipolar disorder. Bipolar Disorders, 2(3), 148–
164.
Strakowski, S.M., Delbello, M.P., et al. (2005). The functional neuroanatomy of bipolar disorder: a review of
neuroimaging findings. Molecular Psychiatry, 10(1), 105–116.
Strakowski, S.M., DelBello, M.P., et al. (2002). Ventricular and periventricular structural volumes in first- versus
multiple-episode bipolar disorder. American Journal of Psychiatry, 159(11), 1841–1847.
Strakowski, S.M., Wilson, D.R., et al. (1993). Structural brain abnormalities in 1st-episode mania. Biological
Psychiatry, 33(8–9), 602–609.
Swayze, V.W., Andreasen, N.C., et al. (1992). Subcortical and temporal structures in affective-disorder and
schizophrenia – a magnetic-resonance-imaging study. Biological Psychiatry, 31(3), 221–240.
Sweeney, J.A., Kmiec, J.A., et al. (2000). Neuropsychologic impairments in bipolar and unipolar mood disorders
on the CANTAB neurocognitive battery. Biological Psychiatry, 48(7), 674–684.
Thompson, J.M., Gallagher, P., et al. (2005). Neurocognitive impairment in euthymic patients with bipolar
affective disorder. British Journal of Psychiatry, 186, 32–40.
van Gorp, W.G., Altshuler, L., et al. (1998). Cognitive impairment in euthymic bipolar patients with and without
prior alcohol dependence – A preliminary study. Archives of General Psychiatry, 55(1), 41–46.
Wang, F., Jackowski, M., et al. (2008). Abnormal anterior cingulum integrity in bipolar disorder determined
through diffusion tensor imaging. British Journal of Psychiatry, 193(2), 126–129.
Wolfe, J., Granholm, E., et al. (1987). Verbal memory deficits associated with major affective-disorders – a
comparison of unipolar and bipolar patients. Journal of Affective Disorders, 13(1), 83–92.
Yurgelun-Todd, D.A., Silveri, M.M., et al. (2007). White matter abnormalities observed in bipolar disorder: A
diffusion tensor imaging study. Bipolar Disorders, 9(5), 504–512.
Zanetti, M.V., Schaufelberger, M.S., et al. (2008). White-matter hyperintensities in first-episode psychosis. British
Journal of Psychiatry, 193(1), 25–30.
Zubieta, J.K., Huguelet, P., et al. (2001). Cognitive function in euthymic Bipolar I Disorder. Psychiatry Research,
102(1), 9–20.
Foundations and futures in bipolar disorder 167
J M
ent H
ealth
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Mcg
ill U
nive
rsity
on
11/0
4/14
For
pers
onal
use
onl
y.