Need for an integrative model
Given evidence for the effectiveness of psychological approaches to BD, how might outcomes be further improved? Researchers need to continue to build therapeutic interventions around theoretical models that attempt to account for the fluctuating course of BD. Such models need to evolve with new research findings, to avoid therapeutic development being based on pragmatic combinations of differing approaches, which would serve to mask both key therapeutic ingredients and mechanisms of recovery.
One influential model for the development of treatment approaches has been the instability hypothesis which suggested that vulnerability to disruption of circadian rhythms is fundamental to BD. Circadian rhythms are patterns of physiological activity that cycle over an approximately 24-h periods including sleep, body temperature, cortisol and melatonin secretion.
Circadian rhythms are entrained by both physical (e.g. light/dark) and social/behavioural zeitgebers. Numerous studies have reported disrupted circadian rhythms in acutely ill BD patients. More recently, studies have found sleep and activity disruption in remitted BD patients suggesting that it is a trait factor. Monk et al. social rhythm metric (SRM) assesses the extent to which individuals have a stable activity pattern, as a proxy measure of circadian functioning. Psychological treatment can have a significant effect on SRM, whilst instability on the SRM is associated with elevated risk of mood episodes in individuals with cyclothymia or BD II.
This proposed vulnerability and its interaction with psychosocial events, stimulated the development of Interpersonal and Social Rhythm Therapy (IPSRT) which combines standard IPT with development of stable patterns of regular activity. IPSRT is associated with stabilization of social rhythms and with longer euthymic periods than clinical management; but to date there is no evidence of a differential impact on relapse rates or mood symptoms.
A multilevel model approach to integration Jones proposed a model of BD integrating circadian instability with psychological functioning which builds on the work of Healy and Williams. The model uses a multilevel approach to cognition and emotion to explore how circadian instability might lead to the observed clinical symptoms of BD. Single-level models of emotions, such as those of Beck et al., have limitations with respect to BD. The illness course of many BD patients is characterized by extended periods in mixed emotional states, where for instance elevated mood and irritability/anger co-occur. Any model therefore needs to accommodate the generation of parallel and even potentially conflicting emotional states. Jones_ model is therefore based on the Schematic Propositional Analogical and Associative Representation Systems (SPAARS) approach of Power and Dalgleish. In SPAARS external events are first processed analogically; a rapid and modality specific level.
Further processing at a propositional level employs abstract language-based models [similar to automatic thoughts within Beck’s model of depression]. Direct emotional outputs occur via schematic and associative levels. Schematic processes integrate information from all other processing systems into knowledge at a level _beyond verbally expressible propositional concepts. The contrast here being between _prose_ description at a propositional level and _poetic_ description at the schematic level. Schemas are dynamic and flexible with respect to changing inputs from other levels and guide behaviour. The associative level permits events, either through repetition or through salience, to trigger emotional output. It also allows for different emotional outputs at different levels of the system (such as the phobic patient who _knows_ that the feared object is harmless, but _feels_ they must continue to avoid it).
In applying the SPAARS model to BD, Jones argued that circadian disruption should be integrated into the analogical stage. For example, in mania events associated with substantial environmental changes and significant circadian disruption result in analogical effects of increased energy, stamina and alertness. When such changes are interpreted through an internal attribution bias it would logically lead to positive propositional thoughts (_I am full of energy and ready to take on the world_). The schematic level, integrating information from all other levels would be associated with beliefs concerning the self as powerful, others as inferior and the world as full of opportunities. This schematic appraisal is then associated both with elevated mood and with (prodromal) behaviours which exacerbate initial changes in mood. The associative link to emotion indicates that once circadian–emotion links are established more detailed schematic appraisal may not be necessary, consistent with findings that later bipolar episodes may be triggered by more minor life events. Also, mixed states can occur when the outputs from different levels conflict, thus an associative circadian–emotion link might output elation, whereas conscious appraisal of the previous consequences of prodromal behaviour might be associated with irritability or agitation.
This model suggests circadian rhythm disruption is an important factor in triggering episodes. However, many other groups experience circadian disruption without experiencing bipolar episodes. What makes BD patients especially vulnerable? Jones draws on evidence from the psychological literature indicating that individuals with BD make internal attributions for both positive and negative events which suggests that when experiencing the effects of circadian disruption they will also make internal attributions. Additionally, Robbins and Kirmayer found individuals with a chronic psychiatric history (not restricted to panic disorder patients) have a greater tendency to make internal psychological attributions for common somatic experiences, whilst individuals at risk of BD also tend to make more positive internal attributions for physiological changes associated with hypomania. These factors suggest, not only that individuals are more prone to circadian instability, but also that the interpretations they make may lead to behaviour which exacerbates these fluctuations.
Figure 2 illustrates how this development of the instability hypothesis might integrate a range of psychological findings with respect to BD, referred to previously in this paper. In particular it is proposed that the avoidant or ruminative coping styles in response to initial mood changes can place the individual at risk of developing prodromal mood states. Whilst distraction has been reported to have benefits in unipolar depression, high-risk studies have reported associations between hypomania and both distraction and engagement in dangerous activities. It has been shown that poor prodromal coping styles for depression (withdrawal and rumination) and mania (sensation seeking/risk taking, increased goal directed activity) are associated with poorer social functioning and higher risk of relapse. Figure 2 indicates that avoidant coping styles (sensation seeking or risk taking) might relate to the _manic defense_ hypothesis. This suggests mania is triggered by efforts to avoid low mood. Although not conclusive a number of studies indicate BD and unipolar participants exhibit similar levels of pessimistic attributional styles on implicit tests, whereas BD participants show an opposite pattern on explicit attributional measures. Furthermore, both behavioural high-risk and depressed and manic BD participants exhibit attentional biases to depression but not mania-related words in an emotional Stroop paradigm.
As the individual struggles to manage an unstable internal environment, social and information processing become more difficult rendering interpretation of the external environment harder and leaving the individual more vulnerable to a high EE family atmosphere. This suggests the importance of a combined approach, addressing both social/family factors and individual needs. In this model a vulnerability to circadian fluctuation is proposed, which translates into symptoms of mania and depression via internal attributions for physiological changes. The model indicates targets for psychological treatment which include the following.
1. Coping styles – Consideration of current coping styles through the use of Wong and Lam’s prodromes interview or by using the Coping Styles Questionnaire of Nolen-Hoeksema would help assess the importance of this area for patients. The structured use of mood and activity diaries is important here to help the patient to learn to use this information to understand the relationship between events, behaviour and mood fluctuations in their illness course. These diaries also provide _baseline_ data against which to assess experiments in which patients try out more adaptive coping responses to early signs of mood change.
2. Stability of social rhythms – Both IPSRT and CBT acknowledge the importance of stabilizing sleep and activity patterns. Many patients acknowledge the extent to which chaos in their personal history is associated with illness episodes, but are less aware of more subtle disruption happening currently. The collaborative development of activity schedules to enhance the regularity of key routines and, where appropriate, sleep protocols are the main aspects of this approach.
3. Social cognition – Patterns of social cognition can be assessed both by the use of established measures and by patient’s recording of thinking patterns using thought-record sheets. Cognitive therapy techniques can help the patient to challenge unhelpful thoughts and subsequently the assumptions and beliefs (schema) that underpin them. Change can be strengthened by the use of behavioural experiments in which patients test out the consequences of acting in accordance with more adaptive thoughts and assumptions. The application and testing of such changes is important in facilitating changes at schematic and associative levels, which SPAARS argues are the direct routes to emotion.
4. Internal attributions for fluctuations in mood and physiology – As patients become clearer about the nature of their individual mood fluctuations and the effects of external events, their tendency to make internal attributions should weaken. However, this process is likely to be enhanced by including _in vivo_ tests within therapy in which physiological fluctuations occur in the absence of mood change. Possibilities here include the use of vigorous exercise or caffeine to mimic some of the arousal symptoms in prodromes of mania. Probing for mood and beliefs in response to such exercise would provide an indication of the extent to which internal attributions had been weakened in therapy.
5. Neurocognitive factors – Given evidence that BD patients experience particular difficulties with processing emotionally salient material, therapy sessions should be structured to ensure that discussion of such material does not undermine engagement with the practical tasks of therapy. Increasing adaptive cognitive and behavioural functioning leads to greater mood stability and should therefore be beneficial in terms of attention and concentration processes. However, issues such as duration of therapy sessions and provision of information in different modalities (including written session summaries) are important to consider, given findings of verbal memory deficits even in remitted patients. Evidence for planning deficits also needs to be considered and suggests that patient’s work outside sessions should be supported by clear planning and in session problem solving.
6. Family factors – High expressed emotion can be addressed by joint work to enhance communication and problem-solving skills with the patient and their carer, following an appropriate needs assessment. Psychoeducational information to carers is likely to form the basis for psychological work to help carers to address unhelpful attributions for patients_ behaviour. Comorbidity needs to be considered when planning BD treatment. Up to 93% of BD patients have a lifetime history of anxiety disorder, whilst 21% have panic disorder history. Substance abuse history has been reported in around 60% of BD patients. Additionally several studies have reported approximately 50% comorbidity for personality disorder. These cormobid disorders are associated with greater complexity in treatment and poorer therapeutic outcomes. There is evidence for the efficacy of CBT for anxiety disorder, drug/alcohol abuse and some forms of personality disorder. In assessing a BD patient for treatment these areas need to be considered and, where likely to impact on the effectiveness of therapy for BD, addressed appropriately.
It is not possible for all BD patients to receive the same intensity of psychological input and this may not be necessary. Individuals should be assessed to determine their therapy needs. From the perspective of this model (in which associative links to mood disruption grow with each episode) and from others such as the neurodegenerative hypothesis, appropriate early intervention to prevent relapses is crucial. Martinez-Aran et al., noting the relationships between cognitive impairment and psychosocial functioning, argued that reduction in cognitive impairment might result from effective relapse prevention. Furthermore, Lam et al._s reports indicate improvements in functional (social functioning) outcomes for patients with a history of relapse receiving CBT.
Where assessment indicates good levels of insight and engagement in a relatively uncomplicated BD presentation the first line of psychological treatment is likely to be psycheducational. Psychoeducational materials can potentially be integrated into routine clinical practice and therefore delivered to a large number of patients. This might include both bibliotherapy and discussions with the clinician charged with their care. Where engagement and insight are less clear or where there is evidence of more risk factors for future relapse (including medication adherence issues) then more structured psycheducational or group CBT approaches might be considered. This can be delivered in a structured protocol-driven format which can be provided by a broad range of clinicians without extensive psychotherapy training.
With individuals with high levels of complexity or a significant relapse history then more intensive psychological therapy may be the best option. This will require a clinician trained in CBT approaches delivering therapy in either an individual or group format. The group format has benefits in terms of providing treatment to more patients at one time and also allowing patients to benefit from the nonspecific effects of a shared group experience. Individual therapy although more expensive in terms of therapist time, has the important benefits of being more clearly tailored to individual need and, if formulation driven, allowing the therapist to address both core BD and comorbid difficulties depending on their impact on the illness course of the particular patient. In those individuals where family atmosphere is assessed to be a significant risk factor integration of family and individual therapy, as here, would be most appropriate.
The range of treatment targets above indicates potential benefits for both family and individual components for therapy. A recent pilot study by Miklowitz et al. reported combining FFT and IPSRT was more effective, in terms of depressive symptoms and time to relapse, than two sessions of family psycho-education along with clinical management. This suggests enhancement of the impact of IPSRT by the FFT component. As the currently structured IPSRT does not directly address all the therapy targets noted above we propose an adapted individual CBT and family approach where this is necessary. Given that BD not only has adverse effects on patients but on their families as well and the fact that familial reactions to the illness can exacerbate symptoms, it seems clear that further evaluation of family-based interventions for this patient group are required. Current evidence shows that FFT has the potential to make a significant impact on patients_ illnesses. This intervention relied heavily on increasing illness knowledge, improving problem-solving and communication skills, and training in prodromal relapse signs and plans. Whilst this was effective in helping reduce the risk of depressive symptoms it did not help reduce the risk of mania. Recent outcome research in schizophrenia suggests attending to needs of carers_ also leads to positive outcomes for both patients and carers. This development might also improve outcomes for patients suffering from BD. In particular, if addressing carer needs has an additional impact on family environment, this would be expected to confer additional benefits for patients attempting to cope with an environment which is partially unpredictable as a function of their own internal instability.
Research into the psychological aspects of BD is relatively recent. However, the developing literature indicates a number of trait differences between BD and controls in information processing, social cognition, coping and attributional styles. Studies of social and family factors indicate important effects of life events and family environment on course and outcome. Early research indicates promising results for psychological and family therapy approaches. Further progress will depend on therapy being developed from models of BD which attempt to encapsulate its complexity. One model is used to indicate the potential for theory driven developments in therapy, in particular through integrating individual CBT and needs based family approaches. Further research to refine the basic biopsychosocial models of BD is likely to provide additional indicators for the future direction of psychological interventions.
- Jones SH, Bipolar Disorders, 2005 Feb; Vol. 7
The article above contains foundational information. Articles below contain optional updates.
Reflection Exercise #10
The preceding section contained information about psychological therapies for bipolar disorder. Write three case study examples
regarding how you might use the content of this section in your practice.
What percentage of BD patients have been reported to have a history of substance abuse? Record the letter of the correct answer the