Cognitive remediation as a treatment for major depression: A rationale,review of evidence and recommendations for future research

Abstract

Objective:

There is considerable literature regarding the effectiveness of cognitive remediation (CR) in schizophrenia and in conditions such as stroke and traumatic brain injury. Patients with major depressive disorder (MDD) present with significant cognitive impairment which in many cases may not resolve with treatment. Neurobiological data suggest that this may relate to underlying dysfunction of pre-frontal cortical areas of the brain and their connections with limbic structures. There has been limited research into specific CR to activate these areas and target impaired cognitive function in MDD. We therefore review current evidence, examine the theoretical basis for and present a rationale for research into CR in MDD. In addition, we will examine important methodological issues in developing such an approach.

Method:

Based on preliminary studies using CR-based techniques, data from CR in schizophrenia, data regarding baseline and residual cognitive impairment in depression, and knowledge of the neurobiology of MDD, we examine the possible utility of CR strategies in the treatment of MDD and make recommendations for research in this area.

Results:

A small number of previous studies have examined specific CR in MDD. The studies are small and inconclusive. However, data on the neuropsychological function and neurobiology of MDD suggest that this is an approach that deserves further attention and research.

Conclusions:

Further research is required in carefully selected populations, using well-defined CR techniques and some form of comparator treatment.

Keywords

Cognition depression mood disorder treatment

Introduction

Major depressive disorder (MDD) is defined by affective disturbance, physiological symptoms and maladaptive thought patterns. Standard biomedical and psychological approaches to treating mood disorders target those diagnostic symptoms. A positive treatment response is generally associated with an improved subjective sense of well-being and the restoration of functioning across most domains of life. However, treatment response is typically slow and symptom amelioration is often not associated with a full functional recovery.

Studies suggest that in some groups of patients there may be significant residual cognitive impairment despite an overall positive treatment response (Douglas et al., 2011; Reppermund et al., 2007). Many treatment-responsive patients also experience residual functional impairment in domains such as interpersonal relationships, academic or occupational achievement, and social engagement, which may be related to residual cognitive impairment.

In the past 10 years, a growing body of evidence suggests that although neurocognitive impairment is a core and stable feature of many neuropsychiatric conditions, it is a malleable treatment target. Pharmacological enhancement of neurocognition has become a clinical, research and regulatory priority (Green et al., 2004), yet to date the statistically significant effects of medications on cognition in mood disorders trials (Khan et al., 2004; Watson et al., 2012) do not necessarily translate to clinically meaningful effects.

Psychological methods have been used to attempt to treat neurocognitive impairment more directly. In schizophrenia, for example, the targeting of cognitive dysfunction has been examined extensively in studies using a range of programmes using specific cognitive training and other behavioural treatment strategies. These approaches have had beneficial effects (McGurk et al., 2007).

Similar non-pharmacological strategies, specifically aimed at remediating cognitive deficits in depression, warrant further examination. While deficits in aspects of social cognition such as emotion recognition have been described in depression (Bourke et al., 2010; Douglas and Porter, 2010), there is more direct evidence of these abnormalities being altered by conventional treatments (Harmer et al., 2009). There is also on-going research into the possibility of tailoring psychological therapies to target emotional processing more directly (see Roiser et al., 2012 for review). Therefore, in this paper, we focus on specific cognitive remediation (CR) of non-emotion-related cognition (i.e. neurocognition) and the possible effects on mood of a CR approach.

In this paper, we will consider evidence for a CR approach in other conditions, before examining the rationale for this approach in major depression, evidence from preliminary studies in depression and methodological issues that need to be addressed in future research.

Cognitive remediation in other conditions

Although there is limited research in MDD, there is considerable evidence of the effectiveness of CR from studies of older adults, in brain injury and in people with schizophrenia. In schizophrenia, for instance, there is a consensus that a range of CR-based programmes have a useful effect on cognitive function, with a weaker effect on behavioural symptoms (Medalia et al., 2009; Twamley et al., 2003). A recent meta-analysis suggests an improvement in cognitive function with a moderate to large effect size, but with a smaller effect size for psychosocial functioning (McGurk et al., 2007). With supplemental skills training, however, cognitive improvements generalise more to improvements in everyday functioning (Bowie et al., 2012). The moderate effect size changes seen in schizophrenia might translate into even more clinically significant changes in depression, given a relatively smaller deficit at baseline. In addition, there are overlapping clinical and neurocognitive boundaries and some evidence of common genetic aetiologies across the schizophrenia and mood disorder spectra (Craddock et al., 2009).

Other conditions in which forms of CR have been used include anorexia nervosa, where, as in depression, cognitive impairment is of a smaller magnitude than in schizophrenia. Tchanturia et al. (2008), in a case series in 27 patients, showed an improvement in aspects of cognitive function and depressive symptoms after 10 individual sessions of CR therapy.

A type of cognitive training, the attention training technique (ATT), is one of the key components of Adrian Wells’ metacognitive therapy (Wells, 2009). Metacognitive therapy was developed to treat excessive, inflexible, self-focused thinking processes such as rumination and worry, seen to be maintaining factors in psychological disorders, particularly anxiety and depression. ATT has been used as a stand-alone technique but so far most of the studies have been uncontrolled (Wells, 2007). There are several publications in conditions other than depression. Wells (1990) first utilised the ATT as a stand-alone technique in a single-case design with a patient with panic disorder. ATT has also been used in case studies with schizophrenia (Levaux, 2011) and auditory hallucinations (Valmaggia et al., 2007), anxiety disorders (Wells et al., 1997), obsessive-compulsive disorder (OCD) (Moritz et al., 2011), childhood social phobia (McEvoy and Perini, 2009), hypochondriasis (Papageorgiou and Wells, 1998) and pain (Sharpe et al., 2010).

Neurobiological effects of CR in other conditions

It is well demonstrated that forms of cognitive training can have effects on brain structure and function. Several non-human animal studies have demonstrated the ability of cognitive training to induce the re-organisation of cortical circuits (see Vinogradov et al., 2012 for review). In elderly human participants, functional plasticity has been demonstrated, associated with working memory improvement following visual perception training (Berry et al., 2010).

In conditions other than mood disorder, there is evidence that CR has significant effects on neurobiological systems similar to those which are abnormal in mood disorders. For example, in schizophrenia, Penades et al. (2002) demonstrated a decrease in hypo-frontality following CR and Wykes et al. (2002) demonstrated increased frontal activation during the performance of a working memory task, which was greatest in the patients who derived the greatest benefit from the therapy. Vinogradov et al. (2009) demonstrated a statistically significant change in serum brain-derived neurotrophic factor (BDNF) levels following 50 hours of CR, in first-episode psychosis, with levels returning close to those of healthy control participants. There was also an increased blood oxygen level-dependent (BOLD) response in the medial pre-frontal cortex in response to an ‘untrained’ memory task. Haut et al. (2010) found, in patients with schizophrenia, that 25 hours of CR with a particular focus on working memory, increased activation in the dorsolateral pre-frontal cortex (DLPFC), anterior cingulate and frontopolar cortex. Eack et al. (2010) demonstrated neuroprotective effects of CR in patients with schizophrenia or schizoaffective disorder, with preservation of grey matter volume in the left hippocampus, parahippocampal gyrus, and fusiform gyrus, as well as volumetric increases in the left amygdala. These changes were associated with improvements not only in behavioural measures of neurocognition, but with social cognition, supporting a neurobiological mechanism for cognitive improvement and its generalisation.

Neuropsychological impairment in depression

Neuropsychological impairment is now well established as a core feature of depression (Austin et al., 1999; Merriam et al., 1999; Porter et al., 2003, 2007) and it is increasingly recognised that at least some patients continue to have residual impairment despite response to treatment in other symptom areas (Douglas et al., 2011; Reppermund et al., 2007). Indeed, the relationship between different aspects of neuropsychological function and clinical state, as assessed using traditional mood rating scales, is unclear. In recent reviews, McDermott and Ebmeier (2009) found significant correlations between depression severity and episodic memory, executive function and processing speed, but not between depression severity and semantic memory or visuo-spatial memory. Douglas and Porter (2009) found that improvement in clinical state was related to improvement in verbal memory and verbal fluency, while measures of executive functioning and attention tended to remain impaired across treatment. In late-life major depression, improved psychomotor speed was most closely related to treatment response (Douglas and Porter, 2009). The conflicting results may relate to differences in the clinical characteristics of groups studied and difficulties in measurement – but also suggest that cognitive dysfunction is to some extent a separate phenomenon worthy of specific treatment.

Evidence regarding the functional significance of neurocognitive dysfunction in MDD is surprisingly sparse. One small study does suggest a relationship between executive function at admission, in patients hospitalised for depression, and later social and occupational outcome (Withall et al., 2009). There are also several studies in bipolar disorder suggesting a link between impairments in neurocognition (particularly verbal memory and executive functions) and poor global psychosocial adjustment (Jaeger et al., 2006; Malhi et al., 2007; Martinez-Aran et al., 2004), occupational difficulties (Martinez-Aran et al., 2007) and interpersonal relationships (Laes and Sponheim, 2006). In one study of euthymic bipolar patients, working memory, but not clinical symptoms, predicted social adjustment (Dittmann et al., 2007). In a 1-year follow-up study, Tabares-Seisdedos et al. (2008) found baseline information processing speed and global neurocognitive change scores to predict the level of disability at follow-up.

In our study of cognitive dysfunction in inpatients with MDD, we demonstrated significant impairment of processing of emotional stimuli (Douglas and Porter, 2010), as well as psychomotor, memory and executive impairments with differences from healthy controls, with an effect size in several tasks of 1.0 (Douglas et al., 2011). This finding, along with others (Reppermund et al., 2007), demonstrates a significant impairment in the sorts of cognitive functions that are likely to be required to make progress in psychotherapy. Our contention is that if these cognitive functions can be strengthened by targeted exercises, patients will be more able to benefit from cognitive, interpersonal and behavioural therapies that require verbal interactions, recalling specific therapeutic techniques, organising behavioural responses, and abstract generalisations to novel functioning environments.

Brain dysfunction in MDD

While brain imaging evidence in MDD is complex, there is a general pattern of increased activity of the limbic system with under-activity of the executive areas of the brain (Drevets et al., 2008). In particular, the amygdala consistently shows increased glucose metabolism and regional cerebral blood flow in MDD (Drevets, 2001) and increased activation during emotional processing in major MDD (see Bourke et al., 2010 for a review). Reversal of this process has been demonstrated following treatment with antidepressants (Fu et al., 2004, 2007) and with cognitive therapy (Fu et al., 2008). Studies of amygdala activation in MDD suggest not only increased activation but also reduced functional connectivity with frontal regions which increases after treatment with antidepressants (Chen et al., 2008). In addition, there is good evidence of overactivity of the subgenual cingulate both in MDD and in ‘normal sadness’ (Mayberg et al., 1999).

In contrast, pre-frontal cortical areas appear to be underactive in MDD, particularly the dorsolateral pre-frontal cortex (Harvey et al., 2005; Siegle et al., 2007b), orbito-frontal cortex and medial pre-frontal cortex (Drevets, 1999). The pre-frontal cortex appears to have an important role in inhibiting limbic regions such as the amygdala (Davidson et al., 2003), which may then relate to overactivity of limbic regions (Davidson et al., 2003; Drevets, 1999).

Rumination is a characteristic and disabling feature of MDD (Ingram et al., 1987) and there may be an association between the degree of rumination and length of episodes of MDD (Ingram, 1984; Nolen-Hoeksema et al., 1993). There is also evidence that the rumination level at end of treatment predicts relapse (Michalak et al., 2011). Rumination may also be associated with amygdala activation (Siegle et al., 2006) and the metabolic rate in the amygdala has been shown to correlate with negative affect in depressed patients (Abercrombie et al., 1998).

Previous studies of CR in MDD

To date, only a few small published studies have examined the use of CR in mood disorders. Elgamal et al. (2007), in 12 patients with MDD, found that CR, when added to treatment as usual, resulted in significantly greater improvement in cognitive function than over the same time period in a matched group of MDD comparison subjects who received only treatment as usual. Improvements were in attention, verbal learning and memory, psychomotor speed and executive function. There was no change in depressive symptom scores. However, both groups had a relatively low symptom rating score on the Hamilton Depression Rating Scale: 17-item (HDRS-17) at baseline, with the active treatment group having an average of 13 and the comparison group 7.5, making it less likely that a significant difference would be shown. Allocation of treatments was not randomised.

Siegle et al. (2007a) examined the utility of repeated practice on two tasks specifically hypothesised to activate the pre-frontal cortex: the Paced Auditory Serial Addition Test and the Attention Control Training Intervention, a computerised modification of Well’s ATT (Wells, 2000). The intervention aimed to increase ‘cognitive control’. Nineteen patients with MDD received six 35-minute ‘cognitive control’ training sessions over a 2-week period while seven received care as usual. The ‘cognitive control’ group showed a significantly decreased self-rated depression score (Beck Depression Inventory; BDI) and a decreased incidence of rumination. In addition, six patients in the ‘cognitive control’ group completed functional magnetic resonance imaging (fMRI) before and after treatment, showing evidence of a decrease in amygdala activity in an emotional sorting task and increased activity in the DLPFC in response to high working memory load. These studies provide preliminary evidence for the concept that cognitive activation targeting functions of the frontal lobes has a positive effect on the outcome of major depressive disorder and concomitant expected biological changes.

Using the ATT, Papageorgiou and Wells (2000) report a case series of four patients with recurrent MDD, who, after a 3–5-week baseline, received five to eight sessions of ATT. All improved markedly in terms of depression severity and this was maintained at the 12-month follow-up. ATT has been incorporated as one of the core components of metacognitive therapy (MCT) for depression (Wells et al., 2009). A multiple-baseline study of four outpatients with recurrent persistent MDD again found rapid and sustained improvement, although the specific effects of ATT were not differentiated from other aspects of MCT treatment. Similar findings of impressive effects sizes in reducing symptoms of depression were reported in a larger single case design study of MCT in patients with recurrent MDD (n = 12) (Wells et al., 2012). Once again though, the specific effects of ATT were not distinguished from MCT as a package.

Naismith et al. (2010) studied 16 patients with a history of MDD but who were not suffering from depression at recruitment. The patients were randomised to receive a ‘neuropsychological educational approach to remediation’ (NEAR) or to no additional treatment. The active intervention used CR based on commercially available computer games, selected according to patients’ strengths, 1 hour twice a week for 10 weeks. There was no significant effect on mood; however, the patients had only mild residual symptoms of depression. There was a significant advantage of therapy compared with waitlist in improving aspects of cognitive function, particularly verbal memory.

Naismith et al. (2011) randomised 44 older patients to receive intervention (once again based on NEAR) immediately or be waitlisted. Forty-one patients took part in follow-up testing. Patients had a lifetime history of major depression but were ‘stabilised on medication’. As in the previous study, they had very low scores at baseline (6.3 active, 9.3 waitlist). The programme consisted of weekly 1-hour sessions of computerised cognitive training. In addition, there was a weekly 1-hour group session of psychoeducation. Once again there was, perhaps unsurprisingly given the low baseline level of depression, no effect on mood. However, there were positive and significant effects on aspects of learning and memory, including verbal memory. The authors discuss the possibility that this may indicate the possibility of neurogenesis occurring in the hippocampus during CR.

In a further study using the NEAR approach, Lee et al. (2013) randomised 49 patients with a single episode of depression or psychosis to receive once weekly 2-hour sessions for 10 weeks of CR based on NEAR. Ten patients in each group had suffered a single episode of depression but details are not given regarding whether the patients were currently classified as depressed. The baseline HDRS-17 was relatively low at 13. Patients were also selected to have at least one score on neuropsychological tests of 1.5 standard deviations below their predicted scores based on the National Adult Reading Test Intelligence Quotient Score. There was a significant beneficial effect of CR on the neuropsychological component ‘immediate learning and memory’ and on the Social Functioning Scale (Birchwood et al., 1990). The lack of an effect of diagnosis in the analyses is suggested to imply that the intervention is equally effective regardless of diagnosis. An alternative interpretation is that the group sizes are too small to be adequately powered to show a differential effect of diagnosis.

Alvarez et al. (2008) examined the effects of a computerised programme which used two games involving predicting the next element of a series and mental arithmetic. Both games had multiple levels of difficulty which increased automatically as success was achieved. The patients were 31 students with major depression. There was no clinician-rated depression rating scale but BDI rating scale scores were around a mean of 24 at baseline. Patients were divided into three groups randomly assigned to receive CR (n = 10), CR plus antidepressant treatment (n = 10) or only antidepressant treatment (n = 11). On the BDI, there was a significant interaction between time and group, with post hoc comparisons revealing a significantly greater reduction in the CR group compared with the antidepressant group. However, there were no differences between the combined and the other two groups. Issues of lack of power are clearly likely to have affected the results. The study also reports results on the Wechsler Adult Intelligence Score (Wechsler, 1955). Both verbal and performance components showed an advantage for the groups receiving CR but the data are not further broken down into subtests, making interpretation of this interesting result difficult.

Bowie and colleagues (2013) reported data from a study using a form of CR in treatment-resistant MDD. Treatment involved 15 hours of group treatment plus supplemental online computerised exercises completed at home. The treatment group (n = 17) had significantly greater improvements on attention/information processing speed and verbal memory domains compared with a waitlist control group (n = 16), and a trend for larger improvements in real-world functioning. In this chronically treatment-resistant sample, there was also a 12.5% reduction in depressive symptoms in the CR group compared with no change in the waitlist group.

One very small preliminary study has used ‘memory training’ in inpatients with severe MDD. The rationale was to use cognitive training to minimise cognitive deficits in patients receiving electroconvulsive therapy (ECT), with some preliminary evidence that this was successful in an open trial in eight patients. This demonstrates the feasibility and practicality of using such an approach in severely depressed patients (Choi, 2011; Choi et al., 2011).

Features of CR in MDD

There are various potential goals of CR in MDD which can be summarised as follows:

Unlike CR in schizophrenia, there is preliminary evidence that CR may be useful, at least as a component of, treatment designed to reverse the fundamental disease process. While the evidence is limited, the idea that activation of parts of the pre-frontal cortex using cognitive exercises, thereby increasing pre-frontal control over limbic areas and reversing a fundamental abnormality of brain circuitry in MDD, is theoretically valid (Siegle et al., 2007a, 2007b). For this reason, we have devoted a significant amount of space below to discussion of tasks specifically designed to activate dorsolateral and ventromedial pre-frontal cortex.

As in schizophrenia, a potential aim is to exercise specific pathways with the goal of remediating specific areas of cognitive function. This is subtly different from the first aim since it views cognitive impairment as a related but independent part of the syndrome. There has, of course, been considerable debate in depression regarding whether neuropsychological impairment should be a separate target for treatment, with at least one treatment study in bipolar depression indicating a dissociation between effects on mood and effects on cognition (Watson et al., 2012). Therefore, specifically targeting impaired areas of cognition in depression appears to be an appropriate strategy. In the studies specifically of MDD, attention and information processing or psychomotor speed and verbal memory are the most often improved, although groups are small and methodology variable (Bowie et al., 2013; Naismith et al., 2010, 2011; Siegle et al., 2007a). In schizophrenia, analyses of much larger groups suggest that areas of cognition able to be improved, for at least 6 months after ending the programmes, are executive function, and working and verbal memory (see Medalia and Choi, 2009 for review).

As in schizophrenia, CR for depression may focus on adapting to established deficits by strategy formation rather than attempting to reverse deficits. For instance, aspects of cognition mediated by the hippocampus may be impaired after multiple episodes, which may have resulted in damage to the hippocampus (Videbech and Ravnkilde, 2004). In this case, the formation and practice of strategies to deal with memory deficits may be the best approach.

As noted, CR for schizophrenia has generally used multiple methods including both cognitive exercises and other methods such as strategy coaching, teaching, group discussion and compensation for deficits by using cognitive strengths. In MDD, there may be varying goals of CR. For instance, CR may be used as an adjunctive treatment in the acute phase to accelerate or promote recovery. Alternatively, it may be used as a treatment designed to promote further functional recovery following symptom stabilisation, possibly as an adjunct to other intensive psychosocial treatment. In the acute phase, treatment may need to be very different, especially in very severe MDD in which the ability of patients to engage in psychological therapies may be very limited. In this situation a very basic ‘drill and practice’ therapy with minimal social interactions may be more appropriate, at least until stabilisation. In less severe or residual MDD, there is likely to be more opportunity to engage in the other CR strategies. Below, we will focus mainly on the cognitive exercise part of the therapy before moving on to discuss some of the other possible aspects of therapy.

Method of delivery

The use of computer software provides precision and flexibility of treatment to meet the dynamic needs of each individual. Computers standardise the presentation of stimuli and recording of responses at the level of milliseconds. By continuously evaluating performance, the difficulty level of the exercises (e.g. duration, complexity, or integrity of stimuli) can be adjusted to match the individual’s performance initially and as it changes over time. There are several advantages to this flexibility in adjusting parameters. First, participants often become quite discouraged when initially engaged in challenging cognitive exercise, which might confirm negative attributions that often serve to maintain the clinical symptoms. Second, treatment is most likely to be effective and appealing when the tasks are diverse and sufficiently challenging to foster new learning and strengthen existing skills as patients progress. In addition, computers can usually present tasks in a more engaging and interesting way than, for instance, pen and paper tasks, and can give immediate feedback on performance.

Specific tasks

According to the biological rationale, tasks that are likely to be useful may be those that are specifically known to engage pre-frontal cortex in their performance. For example, the traditional Trail Making Test (TMT) and particularly TMT-B (alternating from letters to numbers) has been shown to activate dorsolateral and ventrolateral pre-frontal cortex areas (Nakahachi et al., 2010). There is also an increase in pre-frontal blood flow during performance of this task in healthy volunteers (Kubo et al., 2008) while the Tower of London task appears to activate pre-frontal cortex (Newman et al., 2009), particularly DLPFC (Lazeron et al., 2000). Attractive, interactive versions of these tasks are available online with the ability to manipulate difficulty and provide feedback.

Several further tasks known to activate areas of the pre-frontal cortex might also be included in activation batteries. Intensive working memory tasks such as N-back tasks appear to activate areas including DLPFC, ventrolateral pre-frontal cortex and frontal pole (Owen et al., 2005). As noted, as part of a package of CR in schizophrenia, these have been shown to activate DLPFC, anterior cingulate and frontopolar cortex. Go/no go paradigms have reliably been shown to activate orbito-frontal cortex (Horn et al., 2003).

Two tasks in particular have been used in preliminary studies. First, the PASAT (Paced Auditory Serial Addition Task), which was used in the preliminary study of Siegle et al. (2007a). This is a very widely used neuropsychological test (see Tombaugh, 2006 for review), which can be practised. The PASAT has been shown to activate DLPFC during performance and Siegle et al. (2007a) argued that because it is relatively frustrating and elicits a degree of negative emotion, it is designed to improve controlled selective attention in the face of more automatic ruminative cognitions. Its major disadvantage is that it is not interesting and many patients find it positively aversive. For example, Diehr et al. (2003) noted that ‘several participants in our longitudinal research stated that they prefer the lumbar puncture procedure to the PASAT’. This would appear to make this task an unlikely candidate for repeated practice in MDD, at least without additional strategies to address these attributions and motivate participation.

The second task is the ATT (Wells, 2000). The ATT was designed to be used as part of MCT. The ATT aims to enhance the ability to allocate attentional resources in a flexible manner, in particular away from perseverative cognitive processes such as rumination or worry. Within MCT, the ATT is routinely practised in session and is prescribed as a homework exercise twice per day, with advice that it will take at least several weeks to see benefits, perhaps as long as 8 weeks. The ATT practice lasts 10–12 minutes and is available as an audio-download, although it may be practised in vivo. One study reported that two of the four depressed patients were not completing the ATT homework as directed (as they reported feeling distracted by rumination) (Papageorgiou and Wells, 2000); however, a second multiple baseline study found no difficulties with adherence to instructions (Wells et al., 2009). Compliance with ATT homework was noted in an internet self-help application (Moritz et al., 2011).

The attention training task has three phases: focused attention, switching attention and finally broad attention to a range of simultaneous sounds. Patients monitor their own progress over time by rating their attentional focus before and after each ATT practice on a −3 (external focus) or +3 (internal focused) scale. Feedback is not possible with this task, and, in the usual clinical delivery, patient compliance with advice to practise twice daily is assessed by self-report. Computerised online delivery would allow automatic collection of homework compliance.

There are several commercially available packages that use online, computerised and therefore replicable tasks. Some of these have been used in schizophrenia and indeed in some of the preliminary trials in MDD (Bowie et al., 2013). To date, head-to-head comparisons of these tasks have not been conducted and many have not been examined thoroughly in mental disorders.

Treatment frequency/total dose

There is no useful data in mood disorders regarding frequency of sessions or total dose. In schizophrenia, groups receiving on average seven sessions appeared to have similar effects to those receiving up to 33 sessions (Krabbendam and Aleman, 2003). Dosing will probably vary in MDD depending on phase of illness and primary target.

Feedback regarding performance and level of difficulty

Two important issues are likely to be whether direct and immediate feedback is given during performance of the task and what the optimal level of difficulty should be. The two issues are clearly related since feedback may negatively influence performance, thereby making the task more difficult, and the level of difficulty influences the feedback. Feedback may be an immediate part of the task, and may be unavoidable for tasks where responses must be correct to proceed to the next response, or can be given after completion of a block of the task in the form, for example, of a score or percentage correct.

There is some evidence regarding what may be an appropriate level of difficulty for activating tasks in MDD. However, this is very preliminary. First, theoretically, a low level of frustration may be necessary to engage pre-frontal cortex in the face of amygdala activation; hence, tasks should not operate at too high a success rate. In addition, this would be practically unlikely to result in rapid improvement in cognitive performance. Second, there has been previous research on the response of depressed subjects to immediate feedback of failure on neuropsychological tests and the suggestion of a ‘catastrophic response to failure’ (Elliott et al., 1997). However, this effect has been questioned (Shah et al., 1999). In a subsequent study, while a similar effect occurred to that in the study of Elliot et al. (1997), it was in fact a failure of depressed subjects to improve performance following an error rather than a decrement in performance following an error (Douglas et al., 2009). Errors in this study were made at a percentage correct rate of approximately 80% in the depressed group. It may have been this level of errors which reduced motivation to improve following an error compared with healthy controls. In the Ontario study (Holshausen et al., 2011), participants were asked to self-adjust the exercise parameters so that their average accuracy levels approximated 80%. Some set their parameters so that their performance was substantially greater than 80% accuracy. Others had a very dynamic adjustment giving them a large range but average accuracy around 85%. Therefore, it seems that patients naturally adjust to have a success rate of slightly above 80%. An important point is that the optimal success rate may vary depending on the mood state of patients and it may be appropriate to have a lower success rate as patients become more euthymic.

Overall, maintaining the success rate at approximately 85–90% may avoid a negative response to failure, may facilitate a feeling of success and allow therapists to give relatively positive feedback. However, this may need to be altered as treatment progresses. An important question as we move forward is to determine the degree to which self-paced versus automated parameter adjustment affects progress in therapy. The advantage of the former, even if the range of difficultly is high, stems from the argument that autonomy is a critical feature for cognitive mastery in CR (Medalia et al., 2009).

Other components of CR in MDD

As noted, the severity of the illness may dictate the approach taken. In the acute phase of the illness, it may only be possible to engage in very basic ‘drill and practice’ cognitive exercises. Also, in patients in whom there has been significant symptomatic improvement but there is residual cognitive dysfunction, those functions that do not appear to change during treatment might be targeted in a pure cognitive exercise approach.

However, in many cases, other psychosocial approaches are likely to be appropriate. For instance, a psychological therapy component may be added, including strategy coaching, which can be a critical component for transferring cognitive gains to everyday functioning through the use of flexible problem-solving strategies. More novel is enhancing the development of metacognitive awareness or ‘thinking about thinking’ in a more objective context. For example, MCT provides a framework for the use of the ATT (Wells, 2009). MCT is based on a top-down information processing model, the self-regulatory executive functioning model. In this model, Wells proposes that metacognitive beliefs operate with regard to how to employ attentional resources; for example, an individual decides to employ rumination as a strategy to understand why something negative happened or why he or she might be depressed. In MCT, the function and effectiveness of those metacognitions which are driving unhelpful cognitive strategies is challenged and strategies such as detached mindfulness (decentring from cognitions and deciding to disengage from them) and the ATT are used to assist the individual in redeploying attention away from unhelpful strategies. This then allows the individual to be free to utilise any other more flexible and potentially effective problem-solving approaches as appropriate. Such an approach may add to the effectiveness of ATT and indeed other cognitive exercises as a stand-alone strategy, although whether the extra complexity in MCT is required is not yet clear.

Future research

While the use of the various cognitive and metacognitive techniques which may be part of CR for MDD is fairly well established in other disorders, this is not the case for MDD. Only the small preliminary studies described have investigated this, and there are many questions which remain to be answered. Very preliminary evidence, as discussed above, is available to guide decisions – for instance, regarding task difficulty – but for many treatment variables there is minimal evidence. Below, we therefore make a series of recommendations for further research in this area.

Methodological issues and recommendations for future research

Group selection and specification

The usefulness and nature of CR therapy will vary between different groups of depressed subjects. Studies should characterise subjects well, so that it is clear which groups of patients the research applies to. This should include some measure of baseline neuropsychological function and repeated assessment on key measures at the end of treatment. Clinical characteristics, such as duration of illness, severity and treatment responsiveness of symptoms, and the presence or history of hypomanic or manic symptoms are also important variables to control. Treatment history, particularly recent use of ECT or other somatic interventions, will be important variables to control or statistically adjust.

Task selection

Studies should use easily replicable tasks which are either very well described, easily available online or able to be supplied to other centres. Very few studies to date include actual materials, although some are available commercially. The use of standardised or replicable methods for delivering treatment is essential as CR evolves.

Degree of difficulty

We suggest that tasks be manipulated to give a success rate of 80–90%. However, this recommendation is preliminary and should be modified in the light of future research. Studies should be able to manipulate success rate and should specify this.

Treatment frequency

There are little data as yet on this variable. It is clearly very important and should be clearly specified along with actual compliance rates with that frequency. We know that spaced practice and retrieval is an effective technique for new learning, but many CR programmes offer sessions once per week. The advantage of computerised tasks, particularly online tasks, is that it allows more frequent dosing and compliance with treatment is measured.

Comparison treatment

A major issue in any research of this nature is the specificity of the treatment being used. It may simply be that increased input from therapists/clinicians and not the CR is having a beneficial effect. Therefore, selection of the comparison treatment is crucial. This will differ depending on the group being studied. Ideally the comparison therapy should be of equivalent total time and distribution of sessions with equivalent face-to-face versus homework-based practice. The latter can, however, be difficult to ensure, especially if the homework is perceived to be differentially useful by different treatment groups.

Measurements / generalisability

While numerous studies in other areas have demonstrated positive effects of various cognitive exercise paradigms on cognitive tasks directly related to the training, there is little evidence of generalisation to unrelated tasks in healthy populations (Owen et al., 2010). In schizophrenia, cognitive gains transfer more when supplemental skills training helps those with this neurodevelopmental disorder acquire the skills they might have missed (Bowie et al., 2012). In the treatment-resistant MDD study, a trend was found for improved psychosocial functioning and a strong positive correlation was observed between cognitive and functional gains (Bowie et al., 2013). In determining whether CR actually improves cognitive function in depressed patients, it is important to evaluate this on tasks that are not directly related to the training tasks. In schizophrenia, research standard batteries have been developed for measuring cognitive changes (Nuechterlein et al., 2004). In depression, these may need to be modified, but a consensus on the measurement of cognitive change would help to standardise research and make it easier to conduct pooled analyses. The measurement of clinical progress is generally done using mood rating scales. However, these may not be sensitive enough to detect subtle patient differences and are prone to inaccurate measurement (Lombardo et al., 2012). Measurements of psychosocial functioning are clearly particularly important and once again the development of consensus regarding how this should be measured would be helpful (see McKnight and Kashdan, 2009 for review).

Conclusions

MDD, whether unipolar or bipolar, involves significant cognitive impairment. Surprisingly, the relationship of this impairment to the core syndrome of depression is still unclear and there has been little attempt to target cognition in either pharmacological studies or using psychological treatments.

However, ongoing cognitive impairment may:

hamper the process of psychological therapy

give rise to significant psychosocial impairment

remain after the main syndrome is treated

be a risk factor for relapse.

Targeting cognitive impairment with CR may improve these cognitive symptoms. If targeted towards specific neurobiological abnormalities in MDD it may also prove an important adjunct to treatment of the whole syndrome. While preliminary studies are small and heterogeneous, there is sufficient evidence in MDD and to be extrapolated from CR in schizophrenia and other conditions to suggest that CR is an approach which should be further investigated in MDD, and we have suggested a number of areas that should be addressed in further research.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Declaration of interest

GM has received research support from AstraZeneca, Eli Lilly, Organon, Pfizer, Servier and Wyeth; has been a speaker for AstraZeneca, Eli Lilly, Janssen-Cilag, Lundbeck, Pfizer, Ranbaxy, Servier and Wyeth; and has been a consultant for AstraZeneca, Eli Lilly, Janssen-Cilag, Lundbeck and Servier. The other authors have no conflicts to declare.

References

Abercrombie

Schaefer

Larson

. (1998) Metabolic rate in the right amygdala predicts negative affect in depressed patients. Neuroreport 9: 3301–3307.

Alvarez

Sotres

JFC

Leon

. (2008) Computer program in the treatment for major depression and cognitive impairment in university students. Computers in Human Behavior 24: 816–826.

Austin

Mitchell

Wilhelm

. (1999) Cognitive function in depression: A distinct pattern of frontal impairment in melancholia? Psychological Medicine 29: 73–85.

Berry

Zanto

Clapp

. (2010) The influence of perceptual training on working memory in older adults. PLOS ONE 5: e11537.

Birchwood

Smith

Cochrane

. (1990) The Social Functioning Scale. The development and validation of a new scale of social adjustment for use in family intervention programmes with schizophrenic patients. British Journal of Psychiatry 157: 853–859.

Bourke

Douglas

Porter

(2010) Processing of facial emotion expression in major depression: A review. Australian and New Zealand Journal of Psychiatry 44: 681–696.

Bowie

Gupta

Holshausen

. (2013) Cognitive remediation for treatment-resistant depression: Effects on cognition and functioning and the role of online homework. Journal of Nervous and Mental Disorders 201: 680–685.

Bowie

McGurk

Mausbach

. (2012) Combined cognitive remediation and functional skills training for schizophrenia: Effects on cognition, functional competence, and real-world behavior. American Journal of Psychiatry 169: 710–718.

Chen

Suckling

Ooi

. (2008) Functional coupling of the amygdala in depressed patients treated with antidepressant medication. Neuropsychopharmacology 33: 1909–1918.

10.

Choi

(2011) Translating basic neuroscience memory paradigms to treat ECT-related memory impairment. In: Cognitive Remediation in Psychiatry, New York, NY, 10 June.

11.

Choi

Lisanby

Medalia

. (2011) A conceptual introduction to cognitive remediation for memory deficits associated with right unilateral electroconvulsive therapy. The Journal of ECT 27: 286–291.

12.

Craddock

O’Donovan

Owen

(2009) Psychosis genetics: Modeling the relationship between schizophrenia, bipolar disorder, and mixed (or ‘schizoaffective’) psychoses. Schizophrenia Bulletin 35: 482–490.

13.

Davidson

Irwin

Anderle

. (2003) The neural substrates of affective processing in depressed patients treated with venlafaxine. American Journal of Psychiatry 160: 64–75.

14.

Diehr

Cherner

Wolfson

. (2003) The 50 and 100-item short forms of the Paced Auditory Serial Addition Task (PASAT): Demographically corrected norms and comparisons with the full PASAT in normal and clinical samples. Journal of Clinical and Experimental Neuropsychology 25: 571–585.

15.

Dittmann

Seemuller

Schwarz

. (2007) Association of cognitive deficits with elevated homocysteine levels in euthymic bipolar patients and its impact on psychosocial functioning: Preliminary results. Bipolar Disorders 9: 63–70.

16.

Douglas

Porter

(2009) Longitudinal assessment of neuropsychological function in major depression. Australian and New Zealand Journal of Psychiatry 43: 1105–1117.

17.

Douglas

Porter

(2010) Recognition of disgusted facial expressions in severe depression. British Journal of Psychiatry 197: 156–157.

18.

Douglas

Porter

Frampton

. (2009) Abnormal response to failure in unmedicated major depression. Journal of Affective Disorders 119: 92–99.

19.

Douglas

Porter

Knight

. (2011) Neuropsychological changes and treatment response in severe depression. British Journal of Psychiatry 198: 115–122.

20.

Drevets

(1999) Prefrontal cortical-amygdalar metabolism in major depression. Annals of the New York Academy of Sciences 877: 614–637.

21.

Drevets

(2001) Neuroimaging and neuropathological studies of depression: Implications for the cognitive-emotional features of mood disorders. Current Opinion in Neurobiology 11: 240–249.

22.

Drevets

Price

Furey

(2008) Brain structural and functional abnormalities in mood disorders: Implications for neurocircuitry models of depression. Brain Structure and Function 213: 93–118.

23.

Eack

Hogarty

Cho

. (2010) Neuroprotective effects of cognitive enhancement therapy against gray matter loss in early schizophrenia: Results from a 2-year randomized controlled trial. Archives of General Psychiatry 67: 674–682.

24.

Elgamal

McKinnon

Ramakrishnan

. (2007) Successful computer-assisted cognitive remediation therapy in patients with unipolar depression: A proof of principle study. Psychological Medicine 37: 1229–1238.

25.

Elliott

Sahakian

Herrod

. (1997) Abnormal response to negative feedback in unipolar depression: Evidence for a diagnosis specific impairment. Journal of Neurology, Neurosurgery, and Psychiatry 63: 74–82.

26.

Williams

Brammer

. (2007) Neural responses to happy facial expressions in major depression following antidepressant treatment. American Journal of Psychiatry 164: 599–607.

27.

Williams

Cleare

. (2004) Attenuation of the neural response to sad faces in major depression by antidepressant treatment: A prospective, event-related functional magnetic resonance imaging study. Archives of General Psychiatry 61: 877–889.

28.

Williams

Cleare

. (2008) Neural responses to sad facial expressions in major depression following cognitive behavioral therapy. Biological Psychiatry 64: 505–512.

29.

Green

Nuechterlein

Gold

. (2004) Approaching a consensus cognitive battery for clinical trials in schizophrenia: The NIMH-MATRICS conference to select cognitive domains and test criteria. Biological Psychiatry 56: 301–307.

30.

Harmer

Goodwin

Cowen

(2009) Why do antidepressants take so long to work? A cognitive neuropsychological model of antidepressant drug action. British Journal of Psychiatry 195: 102–108.

31.

Harvey

Fossati

Pochon

. (2005) Cognitive control and brain resources in major depression: An fMRI study using the n-back task. Neuroimage 26: 860–869.

32.

Haut

Lim

Macdonald

3rd (2010) Prefrontal cortical changes following cognitive training in patients with chronic schizophrenia: Effects of practice, generalization, and specificity. Neuropsychopharmacology 35: 1850–1859.

33.

Holshausen

Gupta

Jokic

. (2011) The efficacy of a neurocognitive enhancement intervention in treatment-resistant major depressive disorder and the predictive role of internet-based homework. In: Cognitive Remediation in Psychiatry, New York, NY, 10 June.

34.

Horn

Dolan

Elliott

. (2003) Response inhibition and impulsivity: An fMRI study. Neuropsychologia 41: 1959–1966.

35.

Ingram

(1984) Toward an information processing analysis of depression. Cognitive Therapy and Research 8: 443–478.

36.

Ingram

Kendall

Smith

. (1987) Cognitive specificity in emotional distress. Journal of Personality and Social Psychology 53: 734–742.

37.

Jaeger

Berns

Uzelac

. (2006) Neurocognitive deficits and disability in major depressive disorder. Psychiatry Research 145: 39–48.

38.

Khan

Ginsberg

Asnis

. (2004) Effect of lamotrigine on cognitive complaints in patients with bipolar I disorder. Journal of Clinical Psychiatry 65: 1483–1490.

39.

Krabbendam

Aleman

(2003) Cognitive rehabilitation in schizophrenia: A quantitative analysis of controlled studies. Psychopharmacology 169: 376–382.

40.

Kubo

Shoshi

Kitawaki

. (2008) Increase in prefrontal cortex blood flow during the computer version trail making test. Neuropsychobiology 58: 200–210.

41.

Laes

Sponheim

(2006) Does cognition predict community function only in schizophrenia?: A study of schizophrenia patients, bipolar affective disorder patients, and community control subjects. Schizophrenia Research 84: 121–131.

42.

Lazeron

Rombouts

Machielsen

. (2000) Visualizing brain activation during planning: The Tower of London test adapted for functional MR imaging. American Journal of Neuroradiology 21: 1407–1414.

43.

Lee

Redoblado-Hodge

Naismith

. (2013) Cognitive remediation improves memory and psychosocial functioning in first-episode psychiatric out-patients. Psychological Medicine 43: 1161–1173.

44.

Levaux

Laroi

Offerlin-Meyer

. (2011) The effectiveness of the Attention Training Technique in reducing intrusive thoughts in schizophrenia: A case study. Clinical Case Studies 10: 466–484.

45.

Lombardo

Sachs

Kolluri

. (2012) Two 6-week, randomized, double-blind, placebo-controlled studies of ziprasidone in outpatients with bipolar I depression: Did baseline characteristics impact trial outcome? Journal of Clinical Psychopharmacology 32: 470–478.

46.

McDermott

Ebmeier

(2009) A meta-analysis of depression severity and cognitive function. Journal of Affective Disorders 119: 1–8.

47.

McEvoy

Perini

(2009) Cognitive behavioral group therapy for social phobia with or without attention training: A controlled trial. Journal of Anxiety Disorders 23: 519–528.

48.

McGurk

Twamley

Sitzer

. (2007) A meta-analysis of cognitive remediation in schizophrenia. American Journal of Psychiatry 164: 1791–1802.

49.

McKnight

Kashdan

(2009) The importance of functional impairment to mental health outcomes: A case for reassessing our goals in depression treatment research. Clinical Psychology Review 29: 243–259.

50.

Malhi

Ivanovski

Hadzi-Pavlovic

. (2007) Neuropsychological deficits and functional impairment in bipolar depression, hypomania and euthymia. Bipolar Disorders 9: 114–125.

51.

Martinez-Aran

Vieta

Colom

. (2004) Cognitive impairment in euthymic bipolar patients: Implications for clinical and functional outcome. Bipolar Disorders 6: 224–232.

52.

Martinez-Aran

Vieta

Torrent

. (2007) Functional outcome in bipolar disorder: The role of clinical and cognitive factors. Bipolar Disorders 9: 103–113.

53.

Mayberg

Liotti

Brannan

. (1999) Reciprocal limbic-cortical function and negative mood: Converging PET findings in depression and normal sadness. American Journal of Psychiatry 156: 675–682.

54.

Medalia

Choi

(2009) Cognitive remediation in schizophrenia. Neuropsychology Review 19: 353–364.

55.

Medalia

Revheim

Herlands

(2009) Cognitive Remediation for Psychological Disorders: Therapist Guide. New York: Oxford University Press.

56.

Merriam

Thase

Haas

. (1999) Prefrontal cortical dysfunction in depression determined by Wisconsin Card Sorting Test performance. American Journal of Psychiatry 156: 780–782.

57.

Michalak

Holz

Teismann

(2011) Rumination as a predictor of relapse in mindfulness-based cognitive therapy for depression. Psychology and Psychotherapy 84: 230–236.

58.

Moritz

Wess

Treszl

. (2011) The attention training technique as an attempt to decrease intrusive thoughts in obsessive-compulsive disorder (OCD): From cognitive theory to practice and back. Journal of Contemporary Psychotherapy 41: 135–143.

59.

Naismith

Diamond

Carter

. (2011) Enhancing memory in late-life depression: The effects of a combined psychoeducation and cognitive training program. American Journal of Geriatric Psychiatry 19: 240–248.

60.

Naismith

Redoblado-Hodge

Lewis

. (2010) Cognitive training in affective disorders improves memory: A preliminary study using the NEAR approach. Journal of Affective Disorders 121: 258–262.

61.

Nakahachi

Ishii

Iwase

. (2010) Frontal cortex activation associated with speeded processing of visuospatial working memory revealed by multichannel near-infrared spectroscopy during Advanced Trail Making Test performance. Behavioural Brain Research 215: 21–27.

62.

Newman

Greco

Lee

(2009) An fMRI study of the Tower of London: A look at problem structure differences. Brain Research 1286: 123–132.

63.

Nolen-Hoeksema

Morrow

Fredrickson

(1993) Response styles and the duration of episodes of depressed mood. Journal of Abnormal Psychology 102: 20–28.

64.

Nuechterlein

Barch

Gold

. (2004) Identification of separable cognitive factors in schizophrenia. Schizophrenia Research 72: 29–39.

65.

Owen

Hampshire

Grahn

. (2010) Putting brain training to the test. Nature 465: 775–778.

66.

Owen

Mcmillan

Laird

. (2005) N-back working memory paradigm: A meta-analysis of normative functional neuroimaging studies. Human Brain Mapping 25: 46–59.

67.

Papageorgiou

Wells

(1998) Effects of attention training on hypochondriasis: A brief case series. Psychological Medicine 28: 193–200.

68.

Papageorgiou

Wells

(2000) Treatment of recurrent major depression with attention training. Cognitive and Behavioral Practice 7: 407–413.

69.

Penades

Boget

Lomena

. (2002) Could the hypofrontality pattern in schizophrenia be modified through neuropsychological rehabilitation? Acta Psychiatrica Scandinavica 105: 202–208.

70.

Porter

Bourke

Gallagher

(2007) Neuropsychological impairment in major depression: its nature, origin and clinical significance. Australian and New Zealand Journal of Psychiatry 41: 115–128.

71.

Porter

Gallagher

Thompson

. (2003) Neurocognitive impairment in drug-free patients with major depressive disorder. British Journal of Psychiatry 182: 214–220.

72.

Reppermund

Zihl

Lucae

. (2007) Persistent cognitive impairment in depression: The role of psychopathology and altered hypothalamic-pituitary-adrenocortical (HPA) system regulation. Biological Psychiatry 62: 400–406.

73.

Roiser

Elliott

Sahakian

(2012) Cognitive mechanisms of treatment in depression. Neuropsychopharmacology 37: 117–136.

74.

Shah

O’Carroll

Rogers

. (1999) Abnormal response to negative feedback in depression. Psychological Medicine 29: 63–72.

75.

Sharpe

Nicholson Perry

Rogers

. (2010) A comparison of the effect of attention training and relaxation on responses to pain. Pain 150: 469–476.

76.

Siegle

Carter

Thase

(2006) Use of FMRI to predict recovery from unipolar depression with cognitive behavior therapy. American Journal of Psychiatry 163: 735–738.

77.

Siegle

Ghinassi

Thase

(2007a) Neurobehavioral therapies in the 21st century: Summary of an emerging field and an extended example of cognitive control training for depression. Cognitive Therapy and Research 31: 235–262.

78.

Siegle

Thompson

Carter

. (2007b) Increased amygdala and decreased dorsolateral prefrontal BOLD responses in unipolar depression: Related and independent features. Biological Psychiatry 61: 198–209.

79.

Tabares-Seisdedos

Balanza-Martinez

Sanchez-Moreno

. (2008) Neurocognitive and clinical predictors of functional outcome in patients with schizophrenia and bipolar I disorder at one-year follow-up. Journal of Affective Disorders 109: 286–299.

80.

Tchanturia

Davies

Lopez

. (2008) Neuropsychological task performance before and after cognitive remediation in anorexia nervosa: A pilot case-series. Psychological Medicine 38: 1371–1373.

81.

Tombaugh

(2006) A comprehensive review of the Paced Auditory Serial Addition Test (PASAT). Archives of Clinical Neuropsychology 21: 53–76.

82.

Twamley

Jeste

Bellack

(2003) A review of cognitive training in schizophrenia. Schizophrenia Bulletin 29: 359–382.

83.

Valmaggia

Bouman

Schuurman

(2007) Attention training with auditory hallucinations: A case study. Cognitive and Behavioral Practice 14: 127–133.

84.

Videbech

Ravnkilde

(2004) Hippocampal volume and depression: A meta-analysis of MRI studies. American Journal of Psychiatry 161: 1957–1966.

85.

Vinogradov

Fisher

De Villers-Sidani

(2012) Cognitive training for impaired neural systems in neuropsychiatric illness. Neuropsychopharmacology 37: 43–76.

86.

Vinogradov

Fisher

Holland

. (2009) Is serum brain-derived neurotrophic factor a biomarker for cognitive enhancement in schizophrenia? Biological Psychiatry 66: 549–553.

87.

Watson

Gallagher

Porter

. (2012) A randomized trial to examine the effect of mifepristone on neuropsychological performance and mood in patients with bipolar depression. Biological Psychiatry 72: 943–949.

88.

Wells

(1990) Panic disorder in association with relaxation-induced anxiety: An attention training approach to treatment. Behavior Therapy 21: 273–280.

89.

Wells

(2000) Emotional Disorders and Metacognition Innovative Cognitive Therapy. New York: Wiley.

90.

Wells

(2007) The Attention Training Technique: Theory, effects and a metacognitive hypothesis on auditory hallucinations. Cognitive and Behavioral Practice 14: 134–138.

91.

Wells

(2009) Metacognitive Therapy for Anxiety and Depression. New York, NY: Guilford Press.

92.

Wells

Fisher

Myers

. (2009) Metacognitive therapy in recurrent and persistent depression: A multiple-baseline study of a new treatment. Cognitive Therapy and Research 33: 291–300.

93.

Wells

Fisher

Myers

. (2012) Metacognitive therapy in treatment-resistant depression: A platform trial. Behaviour Research and Therapy 50: 367–373.

94.

Wells

White

Carter

(1997) Attention training: Effects on anxiety and beliefs in panic and social phobia. Clinical Psychology and Psychotherapy 4: 226–232.

95.

Wechsler

(1955) Wechsler Adult Intelligence Scale Manual. New York: The Psychological Corporation.

96.

Withall

Harris

Cumming

(2009) The relationship between cognitive function and clinical and functional outcomes in major depressive disorder. Psychological Medicine 39: 393–402.

97.

Wykes

Brammer

Mellers

. (2002) Effects on the brain of a psychological treatment: Cognitive remediation therapy: Functional magnetic resonance imaging in schizophrenia. British Journal of Psychiatry 181: 144–152.