284 vol 28 no 4 april 2015

Craig, A.D. (2009). How do you feel –now? The anterior insula and humanawareness. Nature ReviewsNeuroscience, 10(1), 59–70.

Cristea, I.A., Kok, R.N. & Cuijpers, P.(2015). Efficacy of cognitive biasmodification interventions in anxietyand depression: Meta-analysis.British Journal of Psychiatry 206, 7–16.

Cuijpers, P., Karyotaki, E., Weitz, E. et al.(2014). The effects of

psychotherapies for majordepression in adults on remission,recovery and improvement: A meta-analysis. Journal of AffectiveDisorders, 159, 118–126.

Drevets, W.C., Price, J.L., Simpson, J.R.,Jr. et al. (1997). Subgenual prefrontalcortex abnormalities in mooddisorders. Nature, 386(6627),824–827.

Fenno, L. Yizhar, O. & Deisseroth, K.

(2011). The development andapplication of optogenetics. AnnualReview of Neuroscience 34, 389–412.

Gaynes, B.N., Lloyd, S.W., Lux, L. et al.(2014). Repetitive transcranialmagnetic stimulation for treatment-resistant depression: A systematicreview and meta-analysis. Journal ofClinical Psychiatry, 75(5), 477–489.

Groenewold, N.A., Opmeer, E.M., deJonge, P. et al. (2013). Emotional

valence modulates brain functionalabnormalities in depression.Neuroscience and BiobehaviouralReviews, 37(2), 152–163.

Holmes, E.A., James, E.L., Kilford, E.J. &Deeprose, C. (2010). Key steps indeveloping a cognitive vaccineagainst traumatic flashbacks:Visuospatial Tetris versus verbal pubquiz. PLoS One, 5, e13706.

Layard, R. & Clark, D.M. (2014). Thrive.

Over the past 25 years the pace ofprogress in neuroscience researchhas been extraordinary, withadvances in both understandingand technology. We might expectthat this would stimulate improvedunderstanding and treatment ofmental health problems, yet ingeneral this has not been the case.In fact, our standard treatmentapproaches have barely changed indecades, and still fail many peoplesuffering from mental distress.

Why is there this disconnectbetween knowledge andapplication? And could we be on the brink of an exciting new era of cooperation between the twodisciplines, increasing theeffectiveness of existing treatmentsand even suggesting new ones?

Mental health problems exert anenormous social and economicburden. By way of example, the

cost of depression alone to the Britisheconomy is over £9 billion annually(Thomas & Morris, 2003), and more than twice as many people die fromsuicide each year in the UK than arekilled on its roads (Office for NationalStatistics). These stark figures aresurprising given that depression istreatable, with strong evidence supportingboth pharmacological and psychologicalinterventions.

Why have these social and economiccosts remained so high even as treatmentshave improved? Access to treatment, inparticular psychological therapy, hasincreased in recent years (Layard &Clark, 2014), but of course could bebetter. And while the stigma associatedwith mental health problems is greatlyreduced from what it was a generationago (Office for National Statistics), we are still far from the point where publicperceptions of mental health problemsequate to those of physical conditions.However, these are not the only importantfactors. As mental health practitioners arewell aware, even when a person comesforward for treatment, selecting the righttreatment pathway is a huge challenge.

The impressive statistical evidencebase from treatment trials masks greatvariability: different treatments work fordifferent people (and while this articlefocuses on depression, the same holds forall mental health problems). For example,a large community-based study ofmedication for depression revealed quite

disheartening results. Only around onethird of those in the study recovered fullyon the first antidepressant prescribed(Trivedi et al., 2006). For psychologicaltreatments the picture is only a littlebetter: Cuijpers and colleagues (2014)found that less than half recovered fullyacross a variety of commonly usedpsychotherapies (with evidence-basedtreatments such as cognitive behaviouraltherapy performing best).

In other words, while our standardtreatments work well for some, they failmany, resulting in people suffering forlonger than they should. There arecurrently no methods available instandard mental health practice that can predict, even with modest accuracy,which treatment is going to work forwhich individual. And the many peoplefor whom treatments fail serve as areminder that developing betterinterventions remains a priority.

The difficulty in applyingneuroscience research The past quarter of a century haswitnessed numerous advances inneuroscience, such as neuroimaging andnon-invasive electrical brain stimulationin humans, and optogenetics in animals(the highly selective activation of neuronsusing laser light, see Fenno et al., 2011).If we believe that ‘minds are what brainsdo’ (Minsky, 1988), we might expect thisto lead to improved treatment of mentalhealth problems. Yet arguably, the onlyappreciable impact of neuroscienceresearch on routine mental health practicehas been in the use of animal models todevelop new drugs, which has yieldedfew new treatments in this field over thepast decade.

This disconnect between modernneuroscience research and mental healthpractice partly reflects the unresolved‘hard’ problem of consciousness: Howdoes the brain generate experience?Mapping between activity in neurons orcircuits and subjective experience remains

references

resources

questions

Holmes, E.A., Craske, M.G. & Graybiel,A.M. (2014). Psychological treatments:A call for mental-health science.Nature, 511, 287–289.

Why do different people with the samesymptoms respond to differenttreatments?Why should clinical psychologists careabout understanding symptoms at thelevel of the brain?

ARTICLE

What has neuroscienceever done for us?Jonathan Roiser (winner of the Society’s Spearman Medal 2013) considers thecase of mental health

a huge conceptual, indeed philosophical,challenge. Good science (includingclinical science) requires reliablemeasurement, and neuroscience dealswith what can be measured objectively at the level of the brain. In animals,neuroscience measurements andmanipulations can be causally related to behaviour, but experience can only beinferred indirectly; human studies haveattempted to link brain function tosubjective experiencemeasured using self-report, but for ethicalreasons these studies are largely correlational.More broadly, we lack a generally acceptedneuroscientificexplanation of howbrains make minds(though there have been some attempts, e.g. Craig, 2009).

By contrast, clinicalcharacterisations ofmental health problems,whether conceptualisedas categorical disordersor lying on a spectrum,are based on symptomsthat in many cases onlyexist subjectively. Inother words, mentalhealth practice takes subjectiveexperience as its starting point – there is no objective test for low mood, worryor hallucination. Whether one adopts a categorical or spectrum view, mentalhealth practitioners rely on descriptivedefinitions, in which the symptomsspecify the spectrum or diagnosis.

Why can’t we diagnose disorders using brain scans?Mapping symptomatically andcategorically defined mental disordersonto brain circuits is arguably a hopelesstask. To start with, consider the clinicalheterogeneity. Two individuals diagnosedwith depression could have completely

different symptoms: one experiencing lowmood, poor appetite, difficulty sleeping,worthlessness and suicidal thoughts; theother anhedonia, excessive sleep, fatigue,psychomotor retardation and difficultymaking decisions. There is no obviousreason why these very different clinicalpresentations should correspond to a unitary underlying causal mechanism(meaning here simply the processes that give rise to symptoms, regardless

of whether at thepsychological orneuroscientificlevel).

However, a more fundamentalchallenge is that the even the samesymptoms could be caused bycompletely differentmechanisms. Takethe simplisticexample of a cough.Using a descriptive,approach one coulddefine a ‘coughsyndrome’ – cough(cardinal symptom),might commonly co-occur with runnynose, sore throat,fatigue, etc. A cough

has many potential causes: viral orbacterial infection in the lung; asthma; oreven cancer. But these are conditions thatare defined mechanistically, not accordingto symptoms. Most importantly theyrequire radically different treatments:antibiotics for bacterial infections;steroids for asthma; chemotherapy,radiotherapy or even surgery for lungcancer.

Although the mechanisms drivingdepressive symptoms will be far morecomplex than those of a cough, thisexample serves to highlight the inherentlimits of a purely symptom-baseddiagnostic system, especially whenconsidering treatment selection. It alsosuggests that the notion that there might

be a single mechanism for a descriptivediagnosis like depression is probablywishful thinking. For this reason, we will never be able to diagnose ‘depression’with a brain scan; because symptom-based categories probably conflateclusters of individuals in whom diversebrain processes drive symptoms.Moreover, neuroscientific explanationsdeal with continuous measurements, and therefore naturally align better withspectrum approaches than withcategorical biomedical models. This ispart of the logic that has motivated theAmerican National Institute of MentalHealth to move away from fundingstudies based on descriptive categoricaldiagnoses.

The argument here is not that mentalhealth practitioners should simply ignoresymptoms altogether. Symptoms willalways be the first pointers towardsmechanisms – just as they are in otherclinical disciplines. Someone who visitstheir GP with a cough is not immediatelysubjected to an invasive procedure. But ifthe cough persisted for several monthsthen a doctor might use a diagnostic testto help figure out its cause, beforerecommending a treatment. In mentalhealth practice the symptoms are thediagnosis (or spectrum), by definition.There are no standard diagnosticprocedures to aid GPs in deciding, forexample, whether a depressed individualmight be best treated by drugs targetingthe serotonin or noradrenaline systems,or psychological interventions targetingcognitive schemata or avoidance ofreinforcing activities, or somecombination of these. And while anexperienced psychologist might be able to tailor an intervention such as CBTaccording to her assessment of the factorsthat triggered and maintain symptoms,trial-and-error in treatment selection iscommon.

What do we mean by a ‘cause’?I propose that in order to improve ourunderstanding and treatment of mental

read discuss contribute at www.thepsychologist.org.uk 285

spearman medal

London: Allen Lane.Lozano, A.M., Mayberg, H.S., Giacobbe, P.

et al. (2008). Subcallosal cingulategyrus deep brain stimulation fortreatment-resistant depression.Biological Psychiatry, 64(6), 461–467.

Mayberg, H.S., Liotti, M., Brannan, S.K.et al. (1999). Reciprocal limbic-cortical function and negative mood:Converging PET findings indepression and normal sadness.

American Journal of Psychiatry, 156(5),675–682.

Minsky, M. (1988). The society of mind.New York: Simon and Schuster.

Roiser, J.P., Elliott, R. & Sahakian, B.J.(2012). Cognitive mechanisms oftreatment in depression.Neuropsychopharmacology, 37(1),117–136. d

Ruhe, H.G., Mason, N.S. & Schene, A.H.(2007). Mood is indirectly related to

serotonin, norepinephrine anddopamine levels in humans.Molecular Psychiatry, 12(4), 331–359.

Sroufe, L.A. & Rutter, M. (1984). Thedomain of developmentalpsychopathology. Child Development,55, 17–29.

Thomas, C.M. & Morris, S. (2003). Cost ofdepression among adults in Englandin 2000. British Journal of Psychiatry,183, 514–519.

Trivedi, M.H., Rush, A.J., Wisniewski, S.R.et al. (2006). Evaluation of outcomeswith citalopram for depression usingmeasurement-based care inSTAR*D: Implications for clinicalpractice. American Journal ofPsychiatry, 163(1), 28–40.

In mental health practice thesymptoms are the diagnosis

286 vol 28 no 4 april 2015

spearman medal

health problems we need to specify thecauses of symptoms at the level of theorgan that generates them: the brain. Itmight be claimed that such an approachis too reductionist, ignoring the complexpsychosocial context in which symptomsdevelop. In the standard biopsychosocialmodel, brain-based explanations ofmental health problems are usuallygrouped together with other ‘biological’factors, such as genetic or hormonal, andas a counterpoint to ‘psychosocial’ causes,such as interpersonal stress or early lifeexperience. But this is a strange division.Of course brain function is influenced by genetics; but it is also profoundlyinfluenced by the environment,particularly the social environment. To put it another way, brain-basedexplanations of symptoms are neither‘biological’ nor ‘psychosocial’, but insteadmust incorporate both.

If one accepts the philosophicalposition that all subjective experiencearises from activity in brain circuits thenthe brain becomes an obvious place tosearch for the mechanisms drivingsymptoms, from uncontrollable elation to outright despair. Indeed, it could beargued that integrating psychological andneuroscientific information is the onlyway in which a truly holistic explanationof mental health problems can beproposed, as the brain is the interface at which genetic and environmentalinfluences interact to produce thoughts,perceptions, beliefs and feelings. Andthere is no reason to assume that a brainmechanism can only be targeted using a ‘biological’ treatment – psychosocialinterventions change the brain too.Importantly, this is not to privilegeneuroscientific models over psychologicalmodels of symptoms. Each discipline hasmuch to learn from the other, since theyaddress the same questions but in acomplementary fashion, at different levelsof explanation. Ultimately they requireintegration: ‘mindless’ neuroscience and‘brainless’ psychology are both incompleteexplanatory frameworks.

A useful notion is to considerdifferent types of causes of symptomslying on a spectrum, the extremes ofwhich I will term ‘proximal’ and ‘distal’.Proximal causes are directly related to themechanisms driving symptoms, and areuseful targets for treatment; they are oftenidentified through basic science research.For example, lung cancer is (proximally)caused by malfunction in the machinerythat regulates cell division. Traditionallung cancer treatments tackle this causeby removing the malfunctioning cells(surgery) or killing them (standardchemotherapy and radiotherapy).

However, ‘cancer’ is not really a single entity butinstead a cluster of illnesses,and modern advances incancer biology have yieldedmore precise (andconsequently more useful)proximal causes, pinningdown the exact geneticmutations that drive differentsorts of tumours. Suchfindings have raised theprospect of an era of‘personalised medicine’ inoncology, as targeted drugs aredeveloped for specific mechanisms(Herceptin being a well-known example).For tumours driven by the relevantgenetic mechanism, treatment with thesedrugs can be highly effective. However,they are ineffective (even detrimental) forother types.

By contrast, distal causes are indirectlyrelated to the mechanisms drivingsymptoms, and are useful targets forprevention; they are often identifiedthrough epidemiology research. Again,take the example of lung cancer, which is(distally) caused by cigarette smoking inthe majority of cases, though it must becaused by other factors in people whohave never smoked. These could begenetic (lung cancer is heritable), othertypes of environmental trigger (e.g. radongas exposure) or some interactionbetween the two. Given theoverwhelming evidence that lung canceris (distally) caused by smoking, efforts atprevention rightly focus on reducing itsincidence. However, after a tumour hasdeveloped an oncologist must focus onthe proximal cause when proposing acourse of treatment. Of course, there isoften some relationship between proximaland distal causes – smokers do tend todevelop similar tumours. However, twopatients could have cancers with verysimilar proximal causes, requiring thesame treatment, even if one is a regularsmoker and the other has never smoked.

In mental health practice, this bringsto mind the (now largely historical)distinction between ‘reactive’ versus‘endogenous’ depression; originallythought to be distinct sub-types of illnessdriven by distinct environmental andgenetic (distal) causes, and suited todifferent interventions. In fact, thisclassification was unreliable andprediction of response to treatment wasnever shown to be consistent. In otherwords, similar distal causes do notnecessarily imply similar proximal causes, as has long been recognised indevelopmental psychopathology whereclose parallels can be drawn with the

concepts of equifinality and multifinality(Sroufe & Rutter, 1984). As in oncology,the hope is that better specification of theproximal causes of mental healthproblems will result in better treatment.

Causes of depressionThe majority of studies of depressionhave focused on distal causes (whichpsychologists might consider‘underlying’). These include: heritabilityand genetics; hormonal and immunefactors; upbringing and early lifeexperience; and personality. Moreproximal causes include: various forms of stress, particularly social; high-levelpsychological constructs derived fromcognitive theories (e.g. dysfunctionalnegative schemata); low-level constructssuch as negative information processingbiases (also important in anxiety); anddisrupted transmission inneurotransmitter systems such asserotonin.

However, none of these factors isspecified at the level of activity in braincircuits. Dysfunctional negative schemataare internal representations of theenvironment that are proposed to triggerand sustain symptoms – but schematamust themselves be encoded in the brain.The hypothesis of disrupted serotonintransmission is derived from the effects ofantidepressant drugs – but serotonin doesnot usually directly induce activity inneurons; instead it modulates ongoingactivity. The proposal that low levels ofthis ‘happy hormone’ somehow directlyinduce depressive symptoms is not wellsupported by available evidence (Ruhe etal., 2007), and leaves a large conceptualgap in the serotonin hypothesis.

What of attempts to specifydepression at the level of brain circuits? In the 1990s seminal studiesdemonstrated that regional brainmetabolism and blood flow (proxies for activity) were altered in depressedindividuals. These identified consistentdifferences between depressed and never-

Serotonin does not usually directly induce activity in neurons

read discuss contribute at www.thepsychologist.org.uk 287

spearman medal

depressed individuals in brain structureand function in specific parts of theprefrontal cortex. For example, depressedindividuals had hyperactivity in thesubgenual anterior cingulate cortex(sgACC: Drevets et al., 1997), whichnormalised following recovery (Mayberget al., 1999). Mayberg also showed thatinducing sad mood in non-depressedpeople had the opposite effect of recoveryfrom depression, increasing sgACCactivity. Subsequent work usingfunctional magnetic resonance imaging(fMRI) demonstrated convincingly thatthe sgACC forms part of a circuit thatsupports the brain’s processing of basicemotional information, suggesting a potential role in generating negativebiases. Dozens of further studies haveconfirmed robust abnormalities in thiscircuit in depression, especially duringemotional processing (Groenewold et al.,2013). Taken together, these studiessupport the notion that negative affectiveperception and experience in depressedpeople are caused (proximally) bydisrupted function in the brain circuitsthat support normal emotionalprocessing.

Interesting, but useful?The information provided byneuroscience may be interesting, but is it useful? It is up to neuroscientists toconvince mental health practitioners thattheir discoveries can make a difference tointervention, and explain what form suchapplication might take. Despite theconceptual problems in mappingsymptoms onto brain circuits, such effortshave already provided some surprisingprogress in the treatment of depression.

One conclusion that can be drawnfrom neuroscience studies is to confirmthat ‘depression’ is indeed mechanisticallyheterogeneous, and not a single entity atthe level of the brain. Despite thestatistical evidence for group differences,even the most robust brain-imagingabnormality in depression (reducedsgACC volume) cannot differentiatebetween depressed and never-depressedindividuals reliably. If one were tocalculate a ‘normal range’ of sgACCvolumes in a non-depressed population,the majority of depressed individualswould lie within it.

But might we be able to exploit thisvariability in brain structure and functionamong individuals with the samediagnosis, in order to improve treatmentselection? A series of studies using similardesigns (reviewed in Roiser et al., 2012)has yielded some intriguing preliminaryresults. These scanned the brains of

depressed people before a course oftreatment, while they performed simpleemotion processing tasks. Importantly,different treatments were used in differentstudies: specific medications in some, andCBT or behavioural activation therapy inothers. These treatments were thenadministered over several weeks, and a final assessment of symptom changewas taken (using standard interviewssuch as the Hamilton Rating Scale forDepression), which was related tobaseline brain activation.

Consistent findings were observedacross studies, but in different directionsfor psychological and pharmacologicaltreatments (Roiser et al., 2012).Psychological treatments worked best in individuals with relatively normalbaseline sgACC activation duringnegative emotional processing. Bycontrast, pharmacological treatmentsworked best in individuals with abnormalbaseline sgACC activation. Importantly,this was independent of baseline severity.

Identifying the regions that operateabnormally in depressed individuals has also led to the development of newapproaches that intervene directly at thelevel of brain circuits, both invasive andnon-invasive. Deep brain stimulation(DBS) is a surgical technique thatinvolves applying continuous electricalstimulation directly to a localised brainregion through an implanted electrode,like a pacemaker for the brain.Importantly, and unlike earlierpsychosurgical approaches, DBS does not involve destruction of brain tissue,and the stimulator can be removed if it is ineffective. Initially developed to treatParkinson’s disease, it was first attemptedin depression about a decade ago,targeting the sgACC. An open-label trial of DBS in chronically depressedindividuals produced encouraging results, with approximately 50 per centcategorised as responders to treatmentafter 12 months (Lozano et al., 2008).However, these impressive preliminaryresults need to be replicated inrandomised controlled trials. Similar trials have shown that DBS to other brainregions can also treat chronic obsessive-compulsive symptoms.

Clearly, surgical interventions willonly ever be used in the most extremecases. However, non-invasive methods of direct brain stimulation have also beendeveloped, in particular using repetitivetranscranial magnetic stimulation (rTMS)to target the dorsolateral part of theprefrontal cortex (DLPFC), which isextensively connected with the sgACC.rTMS is completely different toelectroconvulsive therapy – it is

Jonathan Roiseris at the Institute ofCognitive Neuroscience,University College Londonj.roiser@ucl.ac.ukwww.twitter.com/jonroiser

anatomically specific and delivered atmuch lower intensity (with only a tinyrisk of seizure). Several high-quality trialshave provided convincing evidence thatdepressive symptoms can be reducedthrough daily DLPFC rTMS (Gaynes etal., 2014). In America this technique has already received Federal DrugAdministration approval as a potentialtreatment option in depressed individualswho have not responded to medication.

Neuroscience findings might alsostimulate new avenues of research todevelop innovative psychologicaltreatments. A classic example was thedevelopment of exposure therapy to treat phobias, based on the principles of extinction derived from behaviouralneuroscience. More recently, it has beenproposed that using visuo-spatialdistraction to prevent the initialconsolidation of memories in theaftermath of traumatic exposure could act as a ‘cognitive vaccine’, stopping thedevelopment of flashbacks in the firstplace (Holmes et al., 2010). Finally,cognitive bias modification has beendeveloped with the aim of targeting low-level negative emotion processing directly,but this requires more investigation withhigh-quality trials (Cristea et al., 2015).

A new era?At first glance the question posed in mytitle, echoing the famous scene in MontyPython’s The Life of Brian, may haveprovoked a similar sentiment to that ofthe character Reg, who cannotacknowledge that the Romans havedelivered any civic improvements at all.Although modern neuroscience researchhas, as yet, had minimal impact onmental health practice, we are on thebrink of an exciting period.

In the short term the most importanteffect will be to encourage us to changethe way we think about symptoms,focusing on proximal causes at the levelof the brain and how these relate topsychological processes. Longer term, thehope is that by recognising mechanisticheterogeneity we will develop betterclassification systems, new approaches tointervention, and further tools to enablepractitioners to choose the righttreatment for the right individual.