Review

The Emerging Neurobiology of Bipolar

Disorder
Paul J. Harrison,1,2,* John R. Geddes,1,2 and Elizabeth M. Tunbridge1,2

Bipolar disorder (BD) is a leading cause of global disability. Its biological basis Trends
is unknown, and its treatment unsatisfactory. Here, we review two recent areas BD is highly heritable and mostly attri-
butable to common variants of small
of progress. First, the discovery of risk genes and their implications, with a
effect. Several risk genes and gene
focus on voltage-gated calcium channels as part of the disease process and as networks have been identified.
a drug target. Second, facilitated by new technologies, it is increasingly appar-
Calcium signalling is prominent among
ent that the bipolar phenotype is more complex and nuanced than simply one of the genetic risk pathways, and cur-
recurring manic and depressive episodes. One such feature is persistent mood rently appears to have the greatest
instability, and efforts are underway to understand its mechanisms and its therapeutic traction.

therapeutic potential. BD illustrates how psychiatry is being transformed by Digital technologies and sophisticated
contemporary neuroscience, genomics, and digital approaches. mathematical and computational ana-
lyses are being used to quantify and
understand BD.

Bipolar Disorder and the New Psychiatry These new methods reflect, and are
Psychiatry still relies largely on 19th-Century diagnostic categories. These are based on promoting, reconceptualisation of BD
clusters of symptoms rather than biological markers, and are treated with drugs discovered as a chronic instability of mood and
neural circuitry.
serendipitously several decades ago. BD typifies this unsatisfactory state of affairs. Although its
name has changed [it was formerly known as manic depression (see Glossary)], its cardinal Stem cells are becoming an integral
features, and how it is assessed and treated (Box 1) have barely altered. An important reason part of the approaches to understand-
ing BD and its pharmacotherapy.
for this stagnation has been the lack of any real traction on its causes and underlying biology,
beyond its well-established high heritability [1]. Although there is evidence for altered structural New experimental medicine models
and functional brain connectivity [2–4], and changes in markers of oxidative stress [5], mito- are being applied to identify and rapidly
chondrial function [6], inflammation [7], circadian rhythms [8], and dopamine [9], it remains test potential mood-stabilising
treatments.
difficult to integrate these diverse findings, and to disentangle causative changes from those
that are secondary to the disorder and its treatment.

The situation is belatedly improving. While optimism must be tempered by appreciation of the
many complexities, there are realistic prospects for a transformation in our understanding of BD
and how it is diagnosed and treated. Here, we highlight two areas of current interest: the
discovery of the first BD risk genes and their implications, and the application of novel
technologies with the potential to refine, or redefine, the BD phenotype. These developments
exemplify how genomics, neuroscience, and digital technologies are heralding a new era for
psychiatry. For broader reviews of BD, see [10,11]. 1
Department of Psychiatry, University
of Oxford, Warneford Hospital,
The Genomics of Bipolar Disorder Oxford, OX3 7JX, UK
2
Oxford Health NHS Foundation Trust,
A child of an affected parent has about a tenfold increased risk of developing BD, and twin Warneford Hospital, Oxford, OX3 7JX,
studies estimate a heritability of 0.7–0.8 [1]. There is no evidence for Mendelian inheritance or UK
for genes of major effect. Instead, as with most psychiatric disorders, there are multiple
susceptibility loci, each of small effect, which genome-wide association studies (GWAS) are
*Correspondence:
beginning to identify. Several GWAS, and meta-analyses thereof, have been carried out since paul.harrison@psych.ox.ac.uk
2007; Table 1 lists the loci and implicated genes that have emerged to date. The combined (P.J. Harrison).

18 Trends in Neurosciences, January 2018, Vol. 41, No. 1 https://doi.org/10.1016/j.tins.2017.10.006

© 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
 Box 1. Bipolar Disorder: A Clinical Primer
Glossary
The classic picture of BD is like a modified sine wave, with mood fluctuating between episodes of mood elevation
Affect(ive): essentially synonymous
(mania) and depression, interspersed with periods of euthymia. The number of episodes in each mood phase, and their
with mood. BD is sometimes known
duration, varies markedly between individuals, but for most patients the depressive episodes are more prolonged and
as ‘bipolar affective disorder’.
are responsible for much of the morbidity of the disorder.
Bipolar I disorder: the ‘textbook’
form of the disorder, with full manic
The depressive episodes of BD are broadly similar in nature and severity to those of ‘ordinary’ depression. A manic episodes.
episode includes not only significant elevation of mood, but also related changes in behaviour, such as a reduced need Bipolar II disorder: a form of the
for sleep, increased energy, grandiose thoughts and beliefs, rapid speech, increased libido, and reckless behaviour (e. disorder in which manic episodes are
g., spending excessively). In severe episodes, psychotic symptoms (delusions and hallucinations) may also be present; milder or briefer, and depression
for example, the person may believe, or hear voices telling them, that they have superpowers (and they may then act predominates.
accordingly). ‘Hypomania’ refers to a milder and less prolonged form of mania. The exact criteria depend upon the Bipolar spectrum disorder:
classificatory system used (ICD-10 or DSM-5); the latter subdivides BD into bipolar I and bipolar II. Although not part of includes bipolar I and II, and other
the diagnostic criteria, cognitive impairment is a notable aspect of BD; it is present at first episode and persists during cases not meeting criteria for those
euthymia. Attention, processing speed, and verbal learning and fluency are the domains most affected. At least half of disorders.
patients with BD also have an anxiety disorder or substance use disorder. Patients are typically diagnosed during their Cyclothymia: a personality type
20s, following a long prodrome, and actual onset is often in adolescence. The lifetime prevalence of BD is approximately characterised by ups and downs of
1%, rising to 4% if a broader definition of bipolar spectrum disorder is used. The major risk factors are genetic (see the mood, but lacking the severity or
main text), but environmental factors, including childhood adversity, also have a role. Of patients with BD, 10% die by functional impact of BD.
suicide and this, coupled with an excess mortality from natural causes, shortens average life expectancy by approxi- DSM-5: the fifth edition of the
mately 15 years. Diagnostic and Statistical Manual of
Mental Disorders. Includes formal
The mainstay of BD treatment is pharmacological, with lithium salts or the antiepileptic drug sodium valproate used for diagnostic criteria for BP. Published
prophylaxis. Antipsychotics, antidepressants, and antiepileptic drugs are given to treat the mood episodes. Psychoe- by the American Psychiatric
ducation and psychological treatments also have an important supporting role. All current treatments have limited Association in 2013 and widely used
efficacy, and the drugs can have serious adverse effects. in research.
Euthymia (or euthymic): normal
For an introduction to clinical aspects of BD, see [10,95]. mood, neither manic nor depressed.
ICD-10: the tenth edition of the
International Classification of Mental
and Behavioural Disorders was
sample sizes remain small by GWAS standards, and more loci remain to be identified; indeed, published by the World Health
the forthcoming Psychiatric Genomics Consortium analysis, comprising over 20 000 BD cases Organisation in 1992. ICD-10
describes the clinical features of BD
and 30 000 controls, identifies 19 significant loci, including 12 novel ones. Initial exome and
but is rarely used in research since
genome sequencing data suggest that rare deleterious variants also have a role in some BD DSM-5 criteria are operationalised
cases, but their identity and overall contribution to the disorder remain unclear [12–14]. Within and more detailed. ICD-11 will be
BD, there is modest clinicogenetic heterogeneity, for example, based on the predominant released in 2018.
Manic depression (or manic
symptoms, or between bipolar I and bipolar II subtypes [15]. However, there is little evidence
depressive psychosis): an older
for BD-specific genes; joint GWAS analyses show substantial commonalities in risk loci for BD term for BD.
and schizophrenia [16], as well as significant overlap with other major psychiatric disorders [17] Mixed affective states: when
and with intermediate phenotypes, including circadian traits [18,19]. One distinction between depressive and manic features are
present simultaneously.
schizophrenia and BD is that copy number variation is much less prominent in the latter [20].
Psychotic BD: a severe form of BD,
in which psychotic symptoms
Although the genomics of BD are in their infancy, efforts have begun to understand the (delusions and hallucinations) occur
biological basis for the associations identified to date. Interest has centred on two genes during manic and/or depressive
episodes.
(CACNA1C and ANK3) because of what was already known of their functions. CACNA1C is Rapid cycling: occurrence of four or
discussed in detail below. ANK3 encodes ankyrin G, which couples axonal voltage-gated more episodes (or mania or
sodium channels to the cytoskeleton and also has roles in dendrites and glia; another risk gene, depression) within a year.
TRANK1, contains multiple ankyrin repeat domains, suggesting some shared functions.
Complementing the focus on specific risk genes, the first attempts have been made to identify
the pathways that they influence. Using data from four of the GWAS, Nurnberger et al. [21]
reported six pathways that showed replicable association with BD, involving glutamate and
calcium signalling, second messengers, and hormones. Together, these findings support the
possibility that BD is, at least in part, an ion channelopathy [22], in which aberrant calcium
signalling is important [23].

Trends in Neurosciences, January 2018, Vol. 41, No. 1 19

Table 1. GWAS Hits for BDa
Locus Implicated gene(s)

2q11.2 LMAN2L

2q32.1 ZNF804A

3p22.2 TRANK1 (LBA1)

5p15.31 ADCY2

6q16.1 MIR2113, POU3F2 (OTF7)

6q25.2 SYNE1

7p22.3 MAD1L1

9p21.3 Intergenic

10q21.2 ANK3

11q14.1 TENM4 (ODZ4)

12p13.3 CACNA1C

12q13.1 DDN

17q12 ERBB2

a
Data from [98,117].

Calcium Signalling in Bipolar Disorder: Linking Genetics, Pathophysiology,

and Therapeutics
Calcium dysregulation has long been implicated in BD, based primarily on ex vivo studies in
cells taken from patients and controls. The findings are disparate, but on balance indicate that
measures of intracellular calcium signalling are increased in BD, especially after stimulation
(reviewed in [23,24]). The abnormalities appear largely independent of current mood state (i.e.,
they are trait rather than state related). Moreover, they are attenuated by lithium, which is used
in the treatment of the disorder (Box 1). Despite the many uncertainties, these findings led to L-
type voltage-gated calcium channel (VGCC) antagonists, with existing indications in angina and
hypertension, being evaluated for the treatment of BD [25]. Antiepileptic drugs, such as
pregabalin, which act via VGCC a2d subunits (Box 2) have also been tested [26], and
lamotrigine, another antiepileptic drug that may block calcium channels, among its various
actions [27], is an effective treatment for bipolar depression [28].

The results of the recent genomic studies strongly suggest that the involvement of calcium
signalling in BD is at least partly causal [29], and have rekindled attempts to explain more
precisely the nature of the alterations, not least because this may provide clues to more-
effective and tolerable drug strategies to normalise them [30]. However, the discovery of
genetic variants is only the first step, and provides many more questions than answers.
Calcium signalling offers an informative exemplar to highlight the opportunities and complexi-
ties associated with moving from psychiatric genomic discoveries to pathophysiological
insights and therapeutic advances [31,32].

Genomic data provide a starting point to identify the molecules involved in the core ‘calcium
pathophysiology’ of BD. They focus attention on the VGCCs, especially of the L-type, and their
accessory subunits (encoded by the CACNx genes; Box 2). As indicated above, the best
evidence is for CACNA1C (encoding the a1 subunit of Cav 1.2), but pathway analysis also
suggests a role for CACNA1D and CACNB3 [29,33], and other VGCC genes are implicated in

20 Trends in Neurosciences, January 2018, Vol. 41, No. 1

Box 2. VGCC Genes, Their Isoforms, and Relevance in Bipolar Disorder
Identifying the specific VGCCs most relevant for BD is a significant challenge, because their genes give rise to a vast
diversity of functional channels (named Cav channels) [96,97]. VGCCs comprise multiple subunits, each encoded by
one of a subfamily of separate genes. The properties of the ten distinct a1 subunits (encoded by the CACNA1- gene
family) depend on the accessory subunits to which it is bound (Figure I). The main accessory subunits are the b (encoded
by CACNB1–4) and a2d (encoded by CACNA2D1–4) subunits, which are obligate in most cases [98]. Current VGCC
antagonists block the L-type channels; the anticonvulsant/analgesic drugs pregabalin and gabapentin are a2d ligands
[30].

Further channel diversity arises because each gene gives rise to multiple isoforms. The human CACNA1C mRNA has at
least 50 exons and over 40 predicted isoforms (arising from transcriptional and splicing mechanisms). CACNA1C
splicing gives rise to channel isoforms that are differentially expressed in brain compared with heart, and which differ in
their biophysical properties, including voltage-gating characteristics [97]. Another feature affected by splicing is the
isoform sensitivity to existing VGCC antagonists [98]. This suggests that it might be possible to selectively target splice
variants that mediate disease risk and/or are preferentially expressed in the brain, compared with peripheral tissues
(particularly the cardiovascular system, where VGCCs are also abundant), thereby maximising their therapeutic potential
and tolerability in BD [25].

Given these considerations, defining the repertoire of VGCCs present in different human tissues is important, as is
identifying which ones are impacted by the BD-associated risk variants or by BD itself. However, information on the
transcript diversity of human VGCC subunits is sparse, particularly in brain. Furthermore, because VGCC subunit genes
are large (full-length CACNA1C mRNA, for instance, is over 10 kb long), the transcript structure of most isoforms
remains unclear. Characterising the profile of full-length VGCCs isoforms in the human brain, compared with other
tissues, and assessing which are altered in association with genetic risk for BD, are critical first steps in translating the
VGCC genomic findings into pathophysiological insights and novel treatment targets. The availability of large, high-
quality human postmortem brain series and technological advances in the field of RNA sequencing make this goal
achievable.

(A) α2δ
α2
α1 δ

(B)
Cav1 Cav2 Cav3

Gene (CACNA1-): S C D F A B E G H I
Protein (Cav-): 1.1 1.2 1.3 1.4 2.1 2.2 2.3 3.1 3.2 3.3
Subtype: L P/Q N R T

Figure I. The Voltage-Gated Calcium Channel (VGCC) Family of Proteins. (A) Structure of VGCCs, showing the
transmembrane topology of the a1 subunit, its long intracellular C terminus and interactions with accessory (b and a2d)
subunits. (B) Dendrogram and nomenclature of the VGCC family.

Trends in Neurosciences, January 2018, Vol. 41, No. 1 21

BD by rare variant studies [13]. Of note, apart from BD, CACNA1C is associated with schizo-
phrenia [34] and major depression [35], and CACNB2 confers susceptibility to multiple psychiatric
disorders [17]. Involvement of VGCC genes has also been reported in large-scale genomic studies
of BD-relevant phenotypes, such as working memory performance and the associated patterns of
brain activation [36], as well as in general cognitive functioning [37]. There are also smaller
candidate gene studies that suggest effects of CACNA1C genetic variation on brain imaging
phenotypes [38,39] and on cognitive domains, such as reward responsiveness [40].

Identification of the molecular basis for disease associations is a key step in understanding the
mechanisms linking VGCC genes with BD. The VGCC loci revealed by GWAS are noncoding,
and while large-scale exome studies may identify rare variants that disrupt the coding sequence
of VGCCs, it is unlikely that the BD-associated GWAS loci tag as-yet-unidentified, disease-
causing mutations. Instead, they probably act by influencing aspects of gene expression,
including methylation [41], alternative promoter usage, and RNA splicing [42]. In the case of
CACNA1C, the index risk polymorphism for BD (rs1006737) is located in the third intron. On
balance, the available data indicate that the risk allele is associated with enhanced expression
and activity of the gene product [43,44], but there are conflicting findings [45,46], precluding
firm conclusions. Some of the variability in these results may be due to differences in the effect
of rs1006737 on CACNA1C expression between brain regions. Inconsistencies may also result
from the risk single nucleotide polymorphism (SNP) differentially altering the abundance of
particular splice variants, as has been observed for other BD-relevant genes, including ANK3
[47] and ZNF804A [48]. In the case of ANK3, there is evidence that the shift in isoform ratios has
functional consequences for neuronal physiology [49]. Efforts to identify and understand
whether altered splicing is also relevant for CACNA1C and other VGCCs are hampered by
the limited information about their isoform profile in the human brain (Box 2). This information is
critical since splicing patterns are poorly conserved and the brain shows one of the greatest
diversities of alternative splicing [50], meaning that the current data, which pertain to other
species and tissues, are insufficient.

While the identification of specific VGCC subunits and isoforms that mediate BD risk is
important, better understanding of the underlying biology is also crucial. This requires the
use of appropriate cellular and animal model systems, as well as in vivo approaches in humans.
For cellular analyses, in addition to standard cell lines, which are useful for studying the function
of individual genes, induced pluripotent stem cells (iPSCs) may prove valuable. Indeed, iPSC-
derived neurons have already provided intriguing data to support the presence of cellular BD-
related phenotypes, including alterations in calcium signalling (Box 3).

The molecular findings can also help guide the development of rodent models overexpressing
or lacking specific VGCC genes [51–53] or splice variants thereof [54], in which their functional
impact can be studied in the intact animal. For example, embryonic deletion of Cacna1c from
forebrain glutamatergic neurons in mice produced BD-relevant behavioural and cognitive
effects and an increased susceptibility to stress, whereas the same deletion during adulthood
caused a lesser and, in some instances, opposite phenotype [51]. This effect of developmental
stage on the phenotype is intriguing, given the characteristic early adulthood age of BD onset,
and its childhood antecedents (Box 1). In a separate study, the phenotype of Cacna1c-deficient
mice could be rescued using a small-molecule inhibitor of the translation initiation factor eIF2a
[52], giving clues as to the possible intervening biochemical mechanisms.

In parallel with these various genetically driven approaches, pharmacological investigations of

VGCCs in humans are possible because of the existing L-type VGCC antagonists, which can

22 Trends in Neurosciences, January 2018, Vol. 41, No. 1

 Box 3. Stem Cells and the Calcium Pathophysiology of Bipolar Disorder
The potential of iPSCs for studying cellular phenotypes of psychiatric disorders such as BD is considerable, and the
approach has already been exploited in several studies. These have been of three designs: comparison of BD with
healthy patients; contrasting lithium-responsive versus lithium-unresponsive BD cases; or evaluating the effect of BD
risk genotypes (reviewed in [99]).

Mertens et al. [100] generated dentate gyrus-like neurons and showed that cells from patients with BD were
hyperexcitable, with differences in several electrophysiological and transcriptional parameters; moreover, the excitability
was normalised by lithium, but only in cells derived from patients who had responded clinically to the drug. Notably,
these findings were largely replicated in a separate cohort and using a different methodology [101]. Taking a genetic
strategy, Yoshimizu et al. [44] made induced neurons from 24 people, genotyped for the CACNA1C rs1006737
polymorphism, and showed that CACNA1C gene expression and calcium current density were greater in risk allele
homozygotes compared with nonrisk carriers.

3D differentiation approaches are now being used to produce ‘brain spheroids’. Although no studies of this kind have
yet been reported using iPSCs from patients with BD, Birey et al. [102] reported that, intriguingly, spheroids made with
iPSCs from patients with Timothy syndrome (caused by a gain-of-function coding CACNA1C mutation) exhibited
interneuron migratory abnormalities that could be normalised by a VGCC antagonist. The finding draws attention to a
possible role for early developmental events and specific interneuron populations in mediating the role of calcium
signalling in the pathogenesis of BD. Studies using spheroids in BD may be valuable in helping identify circuit-level
phenotypes, while CRISPR techniques to selectively manipulate disease-relevant VGCC isoforms and variants could aid
the identification of the key molecular mechanisms.

These findings, and others (e.g., [103]), illustrate how iPSCs are providing new clues regarding the neuronal phenotypes
of BD, and support an involvement of calcium signalling in these processes. The results encourage further efforts to
extend and scale up the work [104].

be used as experimental tools and for proof-of-principle studies. Their availability is a distinct
advantage compared with most other genetically supported therapeutic targets in BD, for
which no such drugs are available. To date, the clinical trial data linking L-type VGCC blockade
with therapeutic outcome in BD (and, indeed, their psychiatric effects more generally) are wholly
inconclusive [25]. Hence, a priority is to investigate in detail the impact of brain-penetrant VGCC
blockers on BD-relevant phenotypes, including detailed measures of mood, cognition, sleep,
and brain activity [55], and with the incorporation of genotype as a factor [56]. The potential
psychiatric effects of VGCC antagonists can also be assessed using routinely collected clinical
data; for example, a study of electronic medical records in Scotland reported higher hospital
admission rates for depression in patients given VGCCs compared with those prescribed other
antihypertensives [57].

In summary, a range of methods are needed to make the most of the genomic discoveries in
BD, to understand their molecular mechanisms and implications for cellular and systems
functioning, and to evaluate their therapeutic potential. The conceptual approach is illustrated
in Figure 1 (Key Figure), which uses VGCCs as the exemplar.

New Approaches to the Bipolar Phenotype

Digital Psychiatry
Complementing the developments from genomics-driven discovery science, novel methods
are being applied to characterise the BD phenotype, and thereby provide new perspectives on
its key elements.

Conventional psychiatric assessments rely on cross-sectional, retrospective analysis of the

pattern of symptoms over weeks or months. Being based on a patient’s or informant’s recall,
this approach is subject to inherent biases and unreliability, especially when, as in BD, the key

Trends in Neurosciences, January 2018, Vol. 41, No. 1 23

Key Figure
Genomics, Neuroscience, and Treatment Innovation in Bipolar Disorder
(BD)

How do
BD-relevant
isoforms alter
neuron and
circuit funcon?
How do How do
BD-relevant VGCC risk alleles
isoforms alter and drugs affect
intracellular emoon,
signalling? cognion and
Cells Networks behaviour?

Cascades Behaviour

Molecules Trials What is the

Which VGCC therapeuc
isoforms are potenal of
relevant for VGCC drugs
BD? for BD?

Big data
Genomics Detailed phenotyping

Figure 1. The study of voltage-gated calcium channels (VGCCs) in BD exemplifies the potential and the challenges of
translating genomic discoveries into pathophysiology and therapeutics. Although sometimes portrayed as a one-way,
bottom-up, process from genes to treatments, the wheel-like figure emphasises that it is iterative. Underpinning this
approach is the increasing wealth of ‘big data’, since it is the scale of both the genomics and the contemporary digital
approaches to phenotyping that have driven the developments discussed in this review. Looking ahead, an integrative,
collaborative approach will be essential to link the power of big data with the more-focussed, hypothesis-driven studies
that are required for most of the intervening segments.

feature is not simply a symptom (mood), but its profile of variation over time. Several
approaches have been used to try to improve the reliability and validity of BD assessments
[58,59]. The main advances are emerging from developments in information technology,
including the increasingly widespread use of connected and wearable devices: we are entering
an era of ‘digital phenotyping’ in psychiatry [60], with BD, we would argue, at the forefront [61].

24 Trends in Neurosciences, January 2018, Vol. 41, No. 1

One early example is the True Colours platform, which allows patients to submit (by text, email,
web, or app) their ratings of depression, mania, and other symptoms, in response to a weekly or
daily prompt, resulting in a longitudinal and graphical representation of symptom course [62].

Importantly, smartphones and other devices allow the remote capture of not only contempo-
raneous self-reporting of symptoms, but also behavioural, cognitive and physiological mea-
sures, such as heart rate, activity, geolocation, speech, and environmental interactions [63–68].
In turn, digital methods can facilitate a change from entirely symptom-based characterisation of
psychiatric disorders, such as BD, towards a more multimodal, biologically informed one. Thus,
they raise the prospect of more-objective, data-driven diagnoses, and ultimately personalised
predictions of illness and treatment response. However, this promise has yet to be realised; the
initial hype about digital phenotyping, including its application to BD, has been tempered by
increasing awareness of the significant technical, analytical, and other issues involved (Box 4).

Persistent Mood Instability

Notwithstanding the many challenges, the longitudinal, long-term collection of data using digital
approaches has already contributed to a renewed focus on the ‘real-world’ phenotype of BD.
One example is the appreciation that many patients have chronic mood instability (also called
affective lability), persisting during euthymia. This contrasts with the simplistic textbook
description of BD as comprising periods of normal, stable mood in between the episodes
of depression and mania (Box 1). Although this reality was already appreciated by experienced
clinicians [69,70], it is the use of digital methods that has led to a greater awareness of mood
instability, which in turn has encouraged research to understand its origins and significance.
Here, we briefly review some of these implications.

Mood (in)stability is a continuous variable, and is present to varying extents in healthy individuals
[71,72]. Therefore, it is a useful phenotype for studies seeking to identify mechanisms under-
lying mood (dys)regulation per se, and a range of experimental approaches are being taken. For
example, it can be viewed from a computational perspective, with models showing how mood

Box 4. Digital and Mathematical Approaches to Bipolar Disorder

Collection of multidimensional data, as in the remote monitoring of symptoms, behaviours, and physiology, has
considerable potential in the study and treatment of BD and other disorders, but also throws up several significant
challenges [105,106].

First, there are many technical and practical issues to overcome, ranging from ensuring the compatibility of data
collection between operating systems and software versions, to maintaining compliance over long time periods. There
are also privacy, acceptability, and engagement issues raised by the recording and storage of personal data [107].

Second, the resulting data sets are large and complex, and the extraction, analysis, and interpretation of information are
not straightforward [108]. One key question to consider is the mathematical approach chosen for analysing the data.
Options include linear and nonlinear time series methods [109–111] as well as more-flexible and advanced techniques,
such as relaxation oscillator frameworks [112] and machine-learning methods [90]. Another mathematical technique of
interest is rough paths theory [113]. By taking account of lead-lag relationships, rough paths can reduce time-stamped
data sets from complex interacting nonlinear systems to their critical information content or ‘signature’. This trans-
formation provides a structured sequence of low-dimensional summaries of the primary data that completely char-
acterise their complexity but are amenable to linear analysis. This greatly facilitates efforts to use the data for
classification and prediction. For example, a rough path signature, based on daily measures of mood instability,
can differentiate BD from borderline personality disorder [114].

Progress in this developing field will need to be interdisciplinary, combining advanced mathematical analyses with digital
phenotyping technologies that produce robust data and that are sufficiently acceptable, and rewarding, to patients to
secure their long-term engagement [115,116].

Trends in Neurosciences, January 2018, Vol. 41, No. 1 25

instability interacts with reward sensitivity to alter performance [73]. One can also ask how Outstanding Questions
mood instability relates to variation in neural activity across a range of temporal resolutions, What is the overall genetic architecture
using functional MRI [74] and magnetoencephalography [75]. In addition, mathematical anal- of BD? What are the main pathways
impacted by the genes?
yses, such as those outlined in Box 4, can help explain the relationships between fluctuations in
mood and other parameters, ranging from environmental exposures to behavioural variables How will genetics alter the diagnosis of
[66,67]. BD and its relationship to other
disorders?
Mood instability is prominent not only in BD, but also in several other psychiatric disorders, such
What are the causes and mechanisms
as attention deficit disorder, borderline personality disorder, and schizophrenia [76,77]. It is
of altered calcium signalling in BD?
therefore a trans-diagnostic construct, compatible with the NIMH Research Domain Criteria
project [78]. Thus, understanding the origins and mechanisms of mood instability, and the What is the core neurobiology of BD, at
biological and behavioural sequelae of interventions which modify it – is therefore likely to have systems and cellular levels?
value beyond BD. Indeed, mood instability is a moderately heritable trait in the general
population, and the first genome-wide significant loci have been reported [79]. Apart from Will genetic findings and novel pheno-
typing lead to improved therapies for
studying the implicated molecules and pathways in their own right, it will be of interest to
BD?
investigate the extent to which the risk loci for mood instability overlap with those for BD
specifically, and other disorders involving mood instability. Moreover, the specific character-
istics of mood instability (e.g., its frequency, amplitude, or impact on behaviour), while partially
overlapping, may differ sufficiently between disorders [80] to help gain traction on the underlying
neural mechanisms, providing further clues to the biological bases of these illnesses and
leading to refinements in classification.

Finally, mood instability has potential immediate value as a therapeutic target. First, it may
provide a way to screen novel treatments for BD more rapidly and cost-effectively than
conventional clinical trials, in which treatment or prevention of depressive or manic episodes
are the usual outcomes, requiring prolonged periods of observation. In this sense, early mood
stabilisation could prove to predict therapeutic efficacy in the way that rapid effects on
emotional processing predict subsequent therapeutic response in unipolar depression [81].
Moreover, the rapid effects of antidepressants on emotional processing are seen in euthymic
subjects as well as in those with depression; similarly, new treatments for BD, such as novel
VGCC-acting drugs, could be screened in patients who have unstable mood but who do not
have a formal diagnosis. Second, stabilisation of mood may be beneficial in its own right, above
and beyond simply preventing clinical episodes of depression and mania, because mood
instability independently predicts poor functional recovery in BD, and worse outcomes of
various kinds [76,82–84]. In addition, since mood instability often occurs before the onset of BD
[85,86], such treatments might have a preventative role.

In summary, the emergence of persistent mood instability as a construct illustrates how digital
methods can highlight neglected or hard-to-characterise psychiatric phenomena, and stimu-
late a range of studies into their origins, mechanisms, and therapeutic implications. Compara-
ble approaches are being taken to other components of BD, such as circadian dysregulation
and reward sensitivity. Although the work is at an early stage, it is likely ultimately to alter the
view of the core phenotype of BD, with a greater role for biological and quantitative data in BD
diagnosis, prognosis, and management.

Concluding Remarks
Our understanding of BD remains frustratingly limited. It continues to be a descriptive syn-
drome, since we lack sufficient knowledge to allow its characterisation or conceptualisation
based on aetiology or mechanism. Certainly, many questions remain (see Outstanding Ques-
tions). However, there are reasons for optimism. First, the discovery of some of the BD risk

26 Trends in Neurosciences, January 2018, Vol. 41, No. 1

genes has the potential to revolutionise our understanding of its pathogenesis and neurobiol-
ogy. Second, the use of digital technologies and remote sensors, coupled with advanced
analyses of the resulting data, is already allowing a more-quantitative, longitudinal approach to
the BD phenotype. This raises the potential for better prediction of an individual’s clinical
course, and also provides a more-sophisticated phenotype for behavioural and biological
studies. In both of these domains, a common feature is the increasing importance of ‘big data’,
whether in terms of the genomic studies, or the multidimensional data streams captured by
digital devices. Third, although not discussed here, structural and functional brain imaging is
helping to identify the key neural circuits of BD and may also have diagnostic and prognostic
value [2–4,87–90].

The ultimate goal of these contemporary approaches is to allow a more scientifically informed,
evidence-based approach to how BD is classified, measured, and treated [91]. Possible
changes include redrawing or removing the diagnostic boundaries between BD and other
disorders involving lability of mood, emotion, and behaviour; the inclusion of behavioural and
physiological correlates of mood and mood instability (or other symptoms) into clinical
practice; a focus on identifying interventions that can stabilise mood independent of the
underlying diagnosis; and new, genetically informed treatments, such as those targeting
VGCCs.

It is not always appreciated that BD causes a global health burden comparable with that of
schizophrenia [92], yet it has attracted considerably less research funding and interest from
policy makers [93]. We would argue that, for various reasons, BD currently has a greater
potential for transformative advances. More generally, BD is a useful case study for illustrating
how psychiatric disorders are belatedly embarking on the journey from being descriptive
syndromes towards more neurobiologically grounded, quantitative, and digital phenotypes
[94]. Despite the many difficulties this process entails, it is not unreasonable to hope that it will
be successful, and accompanied by the development of more rational, effective, and person-
alised treatments.

Acknowledgments
J.R.G. is a National Institute of Health Research (NIHR) Senior Investigator. E.M.T. is a Royal Society University Research
Fellow. Research supported by Wellcome Trust strategic awards102616 and 09846, Medical Research Council grant
P026028/1, and by the Oxford Health NIHR Biomedical Research Centre. The views expressed are those of the authors
and not necessarily those of the National Health Service, NIHR or the Department of Health. We thank Kia Nobre for the
idea for Figure 1 and Ruth Abrahams and Noel Buckley for supplying images. We apologise to the many authors whose
studies could not be cited due to reference restrictions.

References
1. Craddock, N. and Sklar, P. (2013) Genetics of bipolar disorder. 6. Andreazza, A.C. et al. (2017) Bipolar disorder as a mitochondrial
Lancet 381, 1654–1662 disease. Biol. Psychiatry Published online September 28, 2017.
2. Phillips, M.L. and Swartz, H.A. (2014) A critical appraisal of http://dx.doi.org/10.1016/j.biopsych.2017.09.018
neuroimaging studies of bipolar disorder: toward a new concep- 7. Goldsmith, D.R. et al. (2016) A meta-analysis of blood cytokine
tualization of underlying neural circuitry and a road map for future network alterations in psychiatric patients: comparisons
research. Am. J. Psychiatry 171, 829–843 between schizophrenia, bipolar disorder and depression. Mol.
3. Wise, T. et al. (2016) Voxel-based meta-analytical evidence of Psychiatry 21, 1696–1709
structural disconnectivity in major depression and bipolar disor- 8. Melo, M.C.A. et al. (2017) Chronotype and circadian rhythm in
der. Biol. Psychiatry 79, 293–302 bipolar disorder: a systematic review. Sleep Med. Rev. 34, 46–58
4. Hibar, D.P. et al. (2017) Cortical abnormalities in bipolar disor- 9. Ashok, A.H. et al. (2017) The dopamine hypothesis of bipolar
der: an MRI analysis of 6503 individuals from the ENIGMA affective disorder: the state of the art and implications for treat-
Bipolar Disorder Working Group. Mol. Psychiatry Published ment. Mol. Psychiatry 22, 666–679
online May 2, 2017. http://dx.doi.org/10.1038/mp.2017.73 10. Grande, I. et al. (2016) Bipolar disorder. Lancet 387, 1561–1572
5. Brown, N.C. et al. (2014) An updated meta-analysis of oxida- 11. Harrison, P.J. et al. (2016) Innovative approaches to bipolar
tive stress markers in bipolar disorder. Psychiatry Res. 218, disorder and its treatment. Ann. N.Y. Acad. Sci. 1366,
61–68 76–89

Trends in Neurosciences, January 2018, Vol. 41, No. 1 27

12. Goes, F.S. et al. (2016) Exome sequencing of familial bipolar 35. Rao, S.Q. et al. (2016) Common variants in CACNA1C and MDD
disorder. JAMA Psychiatry 73, 590–597 susceptibility: a comprehensive meta-analysis. Am. J. Med.
13. Ament, S.A. et al. (2015) Rare variants in neuronal excitability Genet. Neuropsychiatr. Genet. 171, 896–903
genes influence risk for bipolar disorder. Proc. Natl. Acad. Sci. 36. Heck, A. et al. (2014) Converging genetic and functional brain
U. S. A. 112, 3576–3581 imaging evidence links neuronal excitability to working memory,
14. Kataoka, M. et al. (2016) Exome sequencing for bipolar disorder psychiatric disease, and brain activity. Neuron 81, 1203–1213
points to roles of de novo loss-of-function and protein-altering 37. Smeland, O.B. et al. (2017) Identification of genetic loci jointly
mutations. Mol. Psychiatry 21, 885–893 influencing schizophrenia risk and the cognitive traits of verbal-
15. Charney, A.W. et al. (2017) Evidence for genetic heterogeneity numerical reasoning, reaction time, and general cognitive func-
between clinical subtypes of bipolar disorder. Transl. Psychiatry tion. JAMA Psychiatry 74, 1065–1075
7, e993 38. Dixson, L. et al. (2014) Identification of gene ontologies linked to
16. O’Donovan, M.C. and Owen, M.J. (2016) The implications of the prefrontal-hippocampal functional coupling in the human brain.
shared genetics of psychiatric disorders. Nat. Med. 22, Proc. Natl. Acad. Sci. U. S. A. 111, 9657–9662
1214–1219 39. Ou, X. et al. (2015) CACNA1C rs1006737 genotype and bipolar
17. Cross-Disorder Group of the Psychiatric Genomics Consortium disorder: focus on intermediate phenotypes and cardiovascular
(2013) Genetic relationship between five psychiatric disorders comorbidity. Neurosci. Biobehav. Rev. 55, 198–201
estimated from genome-wide SNPs. Nat. Genet. 45, 984–994 40. Lancaster, T.M. et al. (2014) CACNA1C risk variant affects
18. Fears, S.C. et al. (2014) Multisystem component phenotypes of reward responsiveness in healthy individuals. Transl. Psychiatry
bipolar disorder for genetic investigations of extended pedi- 4, e461
grees. JAMA Psychiatry 71, 375–387 41. Starnawska, A. et al. (2016) CACNA1C hypermethylation is
19. Pagani, L. et al. (2016) Genetic contributions to circadian activity associated with bipolar disorder. Transl. Psychiatry 6, e831
rhythm and sleep pattern phenotypes in pedigrees segregating 42. Xiao, X. et al. (2017) Molecular mechanisms underlying noncod-
for severe bipolar disorder. Proc. Natl. Acad. Sci. U. S. A. 113, ing risk variations in psychiatric genetic studies. Mol. Psychiatry
E754–E761 22, 497–511
20. Green, E.K. et al. (2016) Copy number variation in bipolar 43. Bigos, K.L. et al. (2010) Genetic variation in CACNA1C affects
disorder. Mol. Psychiatry 21, 89–93 brain circuitries related to mental illness. Arch. Gen. Psychiatry
21. Nurnberger, J.I. et al. (2014) Identification of pathways for bipo- 67, 939–945
lar disorder: a meta-analysis. JAMA Psychiatry 71, 657–664 44. Yoshimizu, T. et al. (2015) Functional implications of a psychiat-
22. Gargus, J.J. (2006) Ion channel functional candidate genes in ric risk variant within CACNA1C in induced human neurons. Mol.
multigenic neuropsychiatric disease. Biol. Psychiatry 60, Psychiatry 20, 162–169
177–185 45. Gershon, E.S. et al. (2014) A rare mutation of CACNA1C in a
23. Berridge, M.J. (2014) Calcium signalling and psychiatric dis- patient with bipolar disorder, and decreased gene expression
ease: bipolar disorder and schizophrenia. Cell Tissue Res. associated with a bipolar-associated common SNP of CAC-
357, 477–492 NA1C in brain. Mol. Psychiatry 19, 890–894

24. Warsh, J.J. (2004) Role of intracellular calcium signalling in the 46. Eckart, N. et al. (2016) Functional characterization of schizo-
pathophysiology and pharmacotherapy of bipolar disorder: cur- phrenia-associated variation in CACNA1C. PLoS One 11,
rent status. Clin. Neurosci. Res. 4, 201–213 e0157086

25. Cipriani, A. et al. (2016) A systematic review of calcium channel 47. Hughes, T. et al. (2016) A loss-of-function variant in a minor
antagonists in bipolar disorder and some considerations for their isoform of ANK3 protects against bipolar disorder and schizo-
future development. Mol. Psychiatry 21, 1324–1332 phrenia. Biol. Psychiatry 80, 323–330

26. Houghton, K.T. et al. (2017) Biological rationale and potential 48. Tao, R. et al. (2014) Expression of ZNF804A in human brain and
clinical use of gabapentin and pregabalin in bipolar disorder, alterations in schizophrenia, bipolar disorder, and major depres-
insomnia and anxiety: protocol for a systematic review and sive disorder: a novel transcript fetally regulated by the psycho-
meta-analysis. BMJ Open 7, e013433 sis risk variant rs1344706. JAMA Psychiatry 71, 1112–1120

27. Ketter, T.A. et al. (2003) Potential mechanisms of action of 49. Lopez, A.Y. et al. (2017) Ankyrin-G isoform imbalance and
lamotrigine in the treatment of bipolar disorders. J. Clin. Psy- interneuronopathy link epilepsy and bipolar disorder. Mol. Psy-
chopharmacol. 23, 484–495 chiatry 22, 1464–1472
28. Geddes, J.R. et al. (2016) Comparative evaluation of quetiapine 50. Barbosa-Morais, N.J. et al. (2012) The evolutionary landscape of
plus lamotrigine combination versus quetiapine monotherapy alternative splicing in vertebrate species. Science 338,
(and folic acid versus placebo) in bipolar depression (CEQUEL): 1587–1593
a 2 x 2 factorial randomised trial. Lancet Psychiatry 3, 31–39 51. Dedic, N. et al. (2017) Cross-disorder risk gene CACNA1C
29. Heyes, S. et al. (2015) Genetic disruption of voltage-gated differentially modulates susceptibility to psychiatric disorders
calcium channels in psychiatric and neurological disorders. during development and adulthood. Mol. Psychiatry Published
Prog. Neurobiol. 134, 36–54 online July 11, 2017. http://dx.doi.org/10.1038/mp.2017.133

30. Zamponi, G.W. et al. (2015) The physiology, pathology, and 52. Kabir, Z.D. et al. (2017) Rescue of impaired sociability and
pharmacology of voltage-gated calcium channels and their anxiety-like behavior in adult cacna1c–deficient mice by
future therapeutic potential. Pharmacol. Rev. 67, 821–870 pharmacologically targeting eIF2a. Mol. Psychiatry 22,
1096–1109
31. Harrison, P.J. (2015) Recent genetic findings in schizophrenia
and their therapeutic relevance. J. Psychopharmacol. 29, 85–96 53. Terrillion, C.E. et al. (2017) Decreased nucleus accumbens
expression of psychiatric disorder risk gene Cacna1c promotes
32. Breen, G. et al. (2016) Translating genome-wide association
susceptibility to social stress. Int. J. Neuropsychopharmacol.
findings into new therapies for psychiatry. Nat. Rev. Neurosci.
20, 428–433
19, 1392–1396
54. Jiang, Y.-Q. et al. (2013) Spinal morphine but not ziconotide or
33. Psychiatric GWAS Consortium Bipolar Disorder Working Group
gabapentin analgesia is affected by alternative splicing of volt-
(2011) Large-scale genome-wide association analysis of bipolar
age-gated calcium channel Ca(V)2.2 pre-mRNA. Mol. Pain 9, 67
disorder identifies a new susceptibility locus near ODZ4. Nat.
Genet. 43, 977–983 55. Kabir, Z.D. et al. (2016) L-type Ca2+ channels in mood, cognition
and addiction: integrating human and rodent studies with a
34. Schizophrenia Working Group of the Psychiatric Genomics
focus on behavioural endophenotypes. J. Physiol. 594,
Consortium (2014) Biological insights from 108 schizophre-
5823–5837
nia-associated genetic loci. Nature 511, 421–427

28 Trends in Neurosciences, January 2018, Vol. 41, No. 1

56. Ostacher, M.J. et al. (2014) Pilot investigation of isradipine in the with major depressive disorder, anxiety disorder and schizo-
treatment of bipolar depression motivated by genome-wide phrenia. Transl. Psychiatry (in press)
association. Bipolar Disord. 16, 199–203 80. Tsanas, A. et al. (2016) Daily longitudinal self-monitoring of
57. Boal, A.H. et al. (2016) Monotherapy with major antihypertensive mood variability in bipolar disorder and borderline personality
drug classes and risk of hospital admissions for mood disorders. disorder. J. Affect. Dis. 205, 225–233
Hypertension 68, 1132–1138 81. Harmer, C.J. et al. (2009) Why do antidepressants take so long
58. Bauer, M. et al. (2004) Using technology to improve longitudinal to act? A cognitive neuropsychological model of antidepressant
studies: self-reporting with ChronoRecord in bipolar disorder. drug action. Br. J. Psychiatry 195, 102–108
Bipolar Disord. 6, 67–74 82. Streijelowitz, S.A. et al. (2013) Mood instability and functional
59. Cochran, A.L. et al. (2016) Data-driven classification of bipolar I recovery in bipolar disorders. Acta Psychiatr. Scand. 128,
disorder from longitudinal course of mood. Transl. Psychiatry 6, 194–202
e912 83. Ducasse, D. et al. (2017) Affect lability predicts occurrence of
60. Insel, T.R. (2017) Digital phenotyping. Technology for a new suicidal ideation in bipolar patients: a two-year prospective
science of behavior. JAMA 318, 1215–1216 study. Acta Psychiatr. Scand. 135, 460–469
61. Nicholas, J. et al. (2015) Mobile apps for bipolar disorder: a 84. Gershon, A. and Eidelman, P. (2015) Inter-episode affective
systematic review of features and content quality. J. Med. intensity and instability: predictors of depression and functional
Internet Res. 17, e198 impairment in bipolar disorder. J. Behav. Ther. Exp. Psychiatry
62. McKnight, R.F. et al. (2017) Longitudinal mood monitoring in 46, 14–18
bipolar disorder: course of illness as revealed through a short 85. Hafeman, D.M. et al. (2016) Toward the definition of a bipolar
messaging service. J. Affective Disord. 223, 139–145 prodrome: dimensional predictors of bipolar spectrum disorders
63. Faurholt-Jepsen, M. et al. (2015) Smartphone data as an elec- in at-risk youths. Am. J. Psychiatry 173, 695–704
tronic biomarker of illness activity in bipolar disorder. Bipolar 86. Van Meter, A.R. et al. (2016) The bipolar prodrome: meta-
Disord. 17, 715–728 analysis of symptom prevalence prior to initial or recurrent mood
64. Faurholt-Jepsen, M. et al. (2016) Voice analysis as an objective episodes. J. Am. Acad. Child Adolesc. Psychiatry 55, 543–555
marker in bipolar disorder. Transl. Psychiatry 6, e856 87. Meda, S.A. et al. (2014) Multivariate analysis reveals genetic
65. McIntyre, R.S. et al. (2014) Advancing biomarker research: associations of the resting default mode network in psychotic
utilizing ‘Big Data’ approaches for the characterization and bipolar disorder and schizophrenia. Proc. Natl. Acad. Sci. U. S.
prevention of bipolar disorder. Bipolar Disord. 16, 531–547 A. 111, E2066–E2075
66. Grunerbl, A. et al. (2015) Smartphone-based recognition of state 88. Savitz, J.B. et al. (2014) Neuropathological and neuromorpho-
and state changes in bipolar disorder patients. IEEE J. Biomed. metric abnormalities in bipolar disorder: view from the medial
Health Inform. 19, 140–147 prefrontal cortical network. Neurosci. Biobehav. Rev. 42,
132–147
67. Palmius, N. et al. (2017) Detecting bipolar depression from
geographic location data. IEEE Trans. Biomed. Eng. 64, 89. Roberts, G. et al. (2017) Functional dysconnection of the inferior
1761–1771 frontal gyrus in young people with bipolar disorder or at genetic
high risk. Biol. Psychiatry 81, 718–727
68. Shou, H. et al. (2017) Dysregulation of objectively assessed 24-
hour motor activity patterns as a potential marker for bipolar I 90. Librenza-Garcia, D. et al. (2017) The impact of machine learning
disorder: results of a community-based family study. Transl. techniques in the study of bipolar disorder: a systematic review.
Psychiatry 7, e1211 Neurosci. Biobehav. Rev. 80, 538–554

69. Mackinnon, D.F. and Pies, R. (2006) Affective instability as rapid 91. Frank, E. et al. (2015) All the world’s a (clinical) stage: rethinking
cycling: theoretical and clinical implications for borderline per- bipolar disorder from a longitudinal perspective. Mol. Psychiatry
sonality and bipolar spectrum disorders. Bipolar Disord. 8, 1–14 20, 23–31

70. Henry, C. et al. (2008) Affective lability and affect intensity as 92. Whiteford, H.A. et al. (2015) The global burden of mental,
core dimensions of bipolar disorders during euthymic period. neurological and substance use disorders: an analysis from
Psychiatr. Res. 159, 1–6 the Global Burden of Disease Study 2010. PLoS One 10,
e0116820
71. Broome, M. et al. (2015) Mood instability. Br. J. Psychiatry 207,
283–285 93. Goodwin, G.M. and Geddes, J.R. (2007) What is the heartland
of psychiatry? Br. J. Psychiatry 191, 189–191
72. Marwaha, S. et al. (2012) The prevalence and clinical associa-
tions of mood instability in adults living in England: results from 94. Bhugra, D. et al. (2017) The WPA-Lancet Psychiatry Commis-
the Adult Psychiatric Morbidity Survey 2007. Psychiatr. Res. sion on the Future of Psychiatry. Lancet Psychiatry 4, 775–818
205, 262–268 95. Harrison, P.J. et al. (2017) Bipolar disorder. In Shorter Oxford
73. Mason, L. et al. (2017) Mood instability and reward dysregulation Textbook of Psychiatry (7th edn) (Harrison, P.J., ed.), pp. 233–
– a neurocomputational model of bipolar disorder. JAMA Psy- 252, Oxford University Press
chiatry Published online October 11, 2017. http://dx.doi.org/ 96. Dolphin, A.C. (2016) Voltage-gated calcium channels and their
10.1001/jamapsychiatry.2017.3163 auxiliary subunits: physiology and pathophysiology and phar-
74. Broome, M. et al. (2015) Neurobiological and behavioural stud- macology. J. Physiol. 594, 5369–5390
ies of affective instability in clinical populations: a systematic 97. Striessnig, J. et al. (2014) L-type Ca2+ channels in heart and
review. Neurosci. Biobehav. Rev. 51, 243–254 brain. WIREs Membr. Transp. Signal. 3, 15–38
75. Baker, A.P. et al. (2014) Fast transient networks in spontaneous 98. Hofmann, F. et al. (2014) L-type CaV1.2 channels: from in vitro
human brain activity. eLife 3, e01867 findings to in vivo function. Physiol. Rev. 94, 303–326
76. Patel, R. et al. (2015) Mood instability is a common feature of 99. Harrison, P.J. (2016) Molecular neurobiological clues to the
mental health disorders and is associated with poor clinical pathogenesis of bipolar disorder. Curr. Opin. Neurobiol. 36,
outcomes. BMJ Open 5, e007504 1–6
77. Trull, Y.J. et al. (2015) Affective dynamics in psychopathology. 100. Mertens, J. et al. (2015) Differential responses to lithium in
Emotion Rev. 7, 355–361 hyperexcitable neurons from patients with bipolar disorder.
78. Insel, T.R. (2014) The NIMH Research Domain Criteria (RDoC) Nature 527, 95–99
project. Am. J. Psychiatry 177, 395–397 101. Stern, S. et al. (2017) Neurons derived from patients with bipolar
79. Ward, J. et al. Genome-wide analysis in UK Biobank identifies disorder divide into intrinsically different sub-populations of neu-
four loci associated with mood instability and genetic correlation rons, predicting the patients’ responsiveness to lithium. Mol.

Trends in Neurosciences, January 2018, Vol. 41, No. 1 29

 Psychiatry Published online February 28, 2017. http://dx.doi. 111. Holmes, E.A. et al. (2016) Applications of time-series analysis to
org/10.1038/mp.2016.260 mood fluctuations in bipolar disorder to promote treatment
102. Birey, F. et al. (2017) Assembly of functionally integrated human innovation: a case series. Transl. Psychiatry 6, e720
forebrain spheroids. Nature 545, 54–59 112. Bonsall, M.B. et al. (2015) Bipolar disorder dynamics: affective
103. Tobe, B.T.D. et al. (2017) Probing the lithium-response pathway instabilities, relaxation oscillations and noise. J. R. Soc. Interface
in hiPSCs implicates the phosphoregulatory set-point for a 12, 0670
cytoskeletal modulator in bipolar pathogenesis. Proc. Natl. 113. Hambly, B. and Lyons, T. (2010) Uniqueness for the signature of
Acad. Sci. U. S. A. 114, E4462–E4471 a path of bounded variation and the reduced path group. Ann.
104. Harrison, P.J. et al. (2016) Reprogramming psychiatry: stem Mathemat. 171, 109–167
cells and bipolar disorder. Lancet 387, 823–825 114. Arribas, I.P. et al. (2017) A signature-based machine learning
105. Monteith, S. et al. (2015) Big data are coming to psychiatry – a model for bipolar disorder and borderline personality disorder.
general introduction. Int. J. Bipolar Disord. 3, 21 arXiv 1707.07124

106. Monteith, S. et al. (2017) Big data for bipolar disorder. Int. J. 115. Bidargaddi, N. et al. (2017) Digital footprints. Mol. Psychiatry 22,
Bipolar Disord. 4, 10 164–169

107. Bauer, M. et al. (2017) Ethical perspectives on recommending 116. Dubad, M. et al. (2017) A systematic review of the psychometric
digital technology for patients with mental illness. Int. J. Bipolar properties, usability and clinical impacts of mobile mood-moni-
Disord. 5, 6 toring applications in young people. Psychol. Med. Published
online June 23, 2017. http://dx.doi.org/10.1017/
108. McIntosh, A.M. et al. (2016) Data science for mental health: a UK
S0033291717001659
perspective on a global challenge. Lancet Psychiatry 3, 993–998
117. Hou, L.P. et al. (2016) Genome-wide association study of
109. Moore, P.J. et al. (2014) Mood dynamics in bipolar disorder. Int.
40,000 individuals identifies two novel loci associated with bipo-
J. Bipolar Disord. 2, 11
lar disorder. Hum. Mol. Genet. 25, 3383–3394
110. Bonsall, M.B. et al. (2012) Nonlinear time-series approaches in
characterizing mood stability and mood instability in bipolar
disorder. Proc. R. Soc. B Biol. Sci. 279, 3632

30 Trends in Neurosciences, January 2018, Vol. 41, No. 1