Assessing Bipolar Lifestyles Experiences (ABLE)

Bipolar disorder is characterised by high rates of substance misuse, with studies in psychiatric and community settings reporting current rates of 48-70% (Miller et al, 1989, Bernadt and Murray, 1986), lifetime prevalence 3-6 times more likely than the general population (Reigier et al, 1990), and a greater proportion of substance users than any other axis 1 psychiatric disorder. Such substance misuse can have profound effects on mood and behaviour which worsen the course, outcome and morbidity of BD (Salloum and Thase, 2000). For example, drugs and alcohol may disrupt sleep and circadian rhythms (Brower, 2001, Rosenwasser, 2001), which can worsen prodromal symptoms and hence increase risk of relapse (Wehr et al, 1987, Frank et al, 2000). Furthermore, impulsiveness, which has been shown to be particularly elevated in bipolar patients with comorbid substance abuse (Swann et al, 2004), has been proposed as a possible factor in the significantly poorer treatment outcome and increased risks of suicide and violence observed in individuals with co-morbid substance misuse compared to those bipolar disorder alone (Salloum, 2000, Strakowski et al, 1988, Bowden et al, 2000).

Importance of the proposed research and relevance to priorities and needs of the NHS

NICE guidelines for bipolar disorder indicate the key role of psychological interventions whilst NICE guideline for substance misuse also highlights the role of psychosocial interventions in both changing patterns of substance use and addressing comorbid affective symptoms (NICE 2006, 2007). However, in spite of its obvious importance, there has been very little research into people with such comorbidity. Two pilot studies of group CBT for substance misuse and bipolar disorder (Weiss et al, 2000, Schmitz et al, 2002) indicated limited patient gains and no reported improvement in relapse (Goldberg, 2001). Weiss reported recently on a definitive trial of group CBT which indicated some improvements in substance use but some deterioration in bipolar symptoms (Weiss et al., 2007). Recent data indicate that long term disabled dual diagnosis bipolar patients who received generic dual disorders treatment (combining assertive outreach, motivational interventions, counselling and social support) exhibited increased rates of remission from substance misuse compared with baseline, although there was no control group and no assessment was made of the effects of this intervention on bipolar relapse (Drake et al, 2004).

Need for research in this area

Research has yet to clarify the precise nature of the relationship between features of the disorder, elevated levels of substance misuse, and the linkage with poor patient outcomes (Strakowski and DelBello, 2000). Recently work has explored the ‘self medication’ hypothesis indicating that an important feature of substance misuse in bipolar disorder is its role in ameliorating distressing BPD symptoms, in particular those relevant to depression (Weiss, 2004). However, Healey et al., (2009) found patients’ experiences with alcohol and drugs were idiosyncratic and complex. Some patients reported using drugs to enhance or prolong periods of mania, whilst some used drugs to ‘dampen’ down the effects of mania. Others used cannabis to control anxiety, relax and manage stress.

Clearly, developing an effective treatment protocol would be greatly helped by an understanding of the reasons for use/change in substance use and the association between psychosis, substance use and adverse outcomes. The limited research available suggests that this understanding may need to take account of possible fluctuations associated with different phases of the illness in order to, for example, identify critical time periods when people may be more motivated to make changes in substance use; or to provide information on high risk times for relapse.

This randomised control trial (RCT) aimed to investigate the benefits of a therapy (Integrated Motivational Interviewing and Cognitive Behavioural Therapy (MiCBT)) for people with Bipolar Disorder. We aimed to help people with Bipolar Disorder to consider lifestyle choices, including how much alcohol they drink. This was a service user defined research project, which means that service users had been involved in choosing the research topic, designing the study and carrying out the research.

Main research questions

• Phase 1: to explore experiences of substance use for individuals with bipolar disorder
• Phase 2: to develop a time limited intervention for substance use in bipolar disorder based on the information gathered in phase 1 and work carried out previously by the team (Barrowclough 2005)
• Phase 3: to test the feasibility of implementing this intervention and its effectiveness for individuals with comorbid substance misuse and bipolar disorder.

Phase 1

The overall aim of this phase was to examine the experiences of substance use in individuals with Bipolar Disorder. The Study consisted of 2 parts:

• Part 1 used Q methodology (Stephenson 1953) to examine individuals’ immediate and delayed experiences of alcohol and cannabis use in order to identify factors associated with use and the consequences of use. The results were used to inform the development of a scale.
• Part 2 used the ‘Experience Sampling Procedure’ (ESM; de Vries 1992; Delespaul, 1992) to examine the contextual factors of cannabis use and the effects of cannabis use on bipolar symptoms and mood in people with BPD. This would allow close investigation of the relationships between participants’ mood (affect), bipolar symptoms, stress-related activity and cannabis use.

Phase 2

Used data from phase 1 to adapt the Mi-CBT therapy manual from the Manchester Midas study (Barrowclough 2005) for use with individuals with Bipolar Disorder. Focus groups were convened to discuss the planned intervention with service users, carers and relatives and relevant multidisciplinary clinicians. Service users included were purposively selected to represent a range of demographic characteristics, substance use patterns and different phases of Bipolar Disorder. Comments informed the final version of the intervention.

Phase 3

Was a treatment delivery and feasibility study. A randomised controlled trial would compare the feasibility and effectiveness of delivering the intervention developed in phase 2 versus treatment as usual. Consecutively eligible patients would be individually randomised to either intervention.

The primary outcomes to be assessed in phase 3 were: feasibility of patient recruitment and consent to participate, adherence to protocol, retention within both arms across assessment, intervention and 6 months follow-up period.

Secondary outcomes would include: Number of bipolar relapses, time to relapse and frequency and severity of substance misuse; measures of bipolar symptomatology (depression, mania, suicidality and irritability) at end of therapy and follow-up, in line with the key targets of the planned intervention; stabilisation of social rhythms and reduction in impulsiveness as a result of treatment and Optimal therapy delivery would be assessed through qualitative interviews which would indicate issues relevant to the process of therapy implementation, problems and possible solutions for optimising therapy delivery within clinical settings.

Setting

The study was run in Manchester.

Target population

Patients with bipolar 1 or 2 affective disorder who regularly use substances participated in the study.

The trial recruited in the following areas:

• Manchester Mental Health and Social Care NHS Trust
• Greater Manchester West Trust