EMDR vs Supportive Therapy in Relapse Prevention in Traumatized Bipolar Patients

Background: The intervention strategies available for bipolar disorder are essentially pharmacotherapy and psychosocial interventions such as cognitive behavioral therapy, psychoeducation, and interpersonal and family therapy. However, despite the everyday use in clinical practice of both types of intervention, almost 70% of patients with bipolar disorder suffer an affective relapse within two years. Although the origin of this is multi-causal, an emerging research topic is the association of bipolar disorder with posttraumatic stress disorder (PTSD). Data from the STEP-BD study showed a prevalence of 20% of PTSD in a sample of 3158 bipolar patients, a rate almost three times higher than the prevalence of PTSD in the general population. This comorbidity has important clinical implications as the traumatized bipolar patients suffer from more rapid cycling, more suicide attempts, more substance abuse, have lower quality of life and more (hypo) manic and depressive symptoms than bipolar patients without PTSD. The results are similar in populations with severe mental illness and a history of trauma (not necessarily diagnosis of PTSD). These populations have more affective episodes, more psychiatric symptoms, increased risk of suicide, more frequency of risk sexual behaviors, more admissions to psychiatric hospitals and in overall, a greater risk of being re-traumatized. The strength of the evidence supporting the clinically relevant effects of PTSD and/or history of trauma in bipolar disorder contrasts with a surprising lack of trials aimed at treating patients who have experienced traumatic events. One form of treatment that is being increasingly used in PTSD therapy is Eye Movement Desensitization and Reprocessing (EMDR). This integrative psychotherapeutical approach uses standardized protocols and elements of cognitive-behavioral, interpersonal, and body-centered therapies in conjunction with dual stimulation (e.g. horizontal eye movements from side to side). The results of two independent meta-analyzes have shown that the EMDR therapy is as effective in the treatment of PTSD symptoms as cognitive behavioral therapy. The treatment with EMDR has also been tested successfully versus exposition therapy and waiting list in a large randomized controlled trial in patients with psychosis and PTSD. Their results showed a significant reduction of trauma scores in both intervention in comparison to the waiting list and both intervention were regarded as safe with respect to exacerbation of psychotic symptoms. Our group carried out the first randomized controlled pilot study of EMDR in bipolar traumatized patients with subsyndromal symptomatology. Our results showed that the EMDR intervention not only reduced the symptoms associated with trauma in the patients, but also had beneficial effects on the symptoms of subsyndromal mood. Following the results this study, our research group has developed a specific and comprehensive EMDR protocol for bipolar patients with a history of trauma. This protocol consists of a comprehensive survey of traumatic events, the intervention and processing of these events according to the Shapiro standard protocol and 5 sub-protocols directed to (a) enhance treatment adherence, (b) increase insight, (c) treat prodromal symptoms, (d) work on the de-idealization of manic symptoms and (e) provide mental stabilization. Aims and hypotheses: The main objective of this study is to examine if EMDR therapy with protocols specific to bipolar patients with a history of traumatic events can act as a mood stabilizer. This would result in less affective relapses and better overall and cognitive functioning after 6 months of therapy in the group that received EMDR submit compared with the ST group. Other related aims of this project are to expand the available options for psychosocial intervention in bipolar disorder, to demonstrate that the EMDR therapy is a safe and effective tool in traumatized bipolar patients and that treatment with EMDR lead to an improvement in the course and prognosis of the disease. Hypotheses: 1. The EMDR group will show less affective relapses at 12 and 24 months follow-up compared to the ST group. 2. Patients in the EMDR group will show less affective symptoms compared to the ST group at evaluations at 6, 12 and 24 months. 3. Patients in the EMDR group will show better cognition and functioning compared to the ST group at evaluations at 6, 12 and 24 months. Design: Single-blind randomized clinical trial (1:1) with two parallel branches, stratified by center for age, sex, number of previous affective episodes and cognitive state. The preventive effect of two psychological interventions, individual EMDR therapy or ST plus pharmacological treatment in patients diagnosed with bipolar I and II disorder with history of traumatic events, in a current affective phase of euthymia or subsyndromal symptoms will be evaluated. The comparative clinical effect in both branches of intervention will be assessed at five time points. There will be a pre-intervention baseline assessment, a post-intervention assessment at 6 months and follow-up evaluations at 12, 18 and 24 months. Clinical and diagnostic variables: 1. Clinical diagnosis of the participants will employ DSM-IV-TR criteria. 2. Clinical Severity of the participants will be assessed by different instruments: 1. Bipolar Depression Rating Scale (BDRS): The BDRS is the most up-to-date and appropriate tool to assess depressive and mixed symptoms in bipolar patients. Our group has been involved in the translation and validation of the scale into Spanish. 2. Young Mania Rating Scale (YMRS) which is validated in Spanish: YMRS is a hetero-administered scale of composed of 11 items aimed at quantifying the severity of manic and hypo manic episodes. 3. Clinical Global Impression Scale modified for bipolar disorder (CGI-BP-M), Spanish validation: The CGI-BP-M scale consists of 3 subscales; the first two assess the severity of acute symptoms of mania and depression, and the third evaluates the longitudinal severity of the disease. Each subscale has seven sub-categories with scores from 1 to 7 rating the severity of the disorder as normal, low, mild, moderate, marked, severe or very severe. Other clinical variables will be collected via the medical history of the patients and using a specific CRF for the study such as age of onset of the disorder, number of relapses, number of previous episodes, history and number of suicide attempts and pharmacological variables such as the number, type and dose of drugs. 3. Trauma symptoms, cognitive profiles and overall functioning will be evaluated by the following instruments: 1. Clinician-Administered PTSD Scale (CAPS) which is validated in Spanish. 2. Impact of Event Scale Revised (IES-R). 3. Distressing Event Questionnaire. 4. Traumatic Life Events Questionnaire. c. Functioning Assessment Short Test (FAST). d. Screen for Cognitive Impairment in Psychiatry (SCIP) which is validated in Spanish. e. Screen for social cognition, emotional intelligence: Mayer-Salovery-Caruso Emotional Intelligence Test. Statistical analysis: 1. Computation of sample size The study will assess the relative efficacy of a specific EMDR intervention protocol versus Supportive Therapy (ST) in the clinical stabilization (decrease in the number of relapses) of traumatized bipolar patients. For this reason the number of affective episodes after treatment with a follow up of 24 months will be used as the dependent variable of the study. The sample size computation is based on a recent study comparing CBT versus ST with 24 months follow up on the relapse in bipolar patients. The calculation of the sample size was based on a survival analysis using the statistical package "powerSurvEpi" for R (http://www.r-project.org/) using an alpha = 0.005 instead of 0.05 for allow correction for multiple comparisons. The number of patients required to detect a hazard ratio = 2 in a cox regression with a statistical power of 80% and alpha = 0.005 is n = 36 per intervention group (two groups, EMDR and ST, equals to a total n = 72). A sample of this size should show clinically relevant differences. Assuming a percentage of drop-outs of about 10-15% of the participants it would be necessary to recruit approximately 82 patients, 41 allocated in each intervention arm. 2. Analysis of the main variables The baseline distribution of socio-demographic and clinical characteristics between the groups will be analyzed using descriptive statistics. Continuous variables with a normal distribution will be analyzed with Multivariate Analysis of Variance (MANOVA). The main hypothesis of survival time (in weeks) will be analyzed using Kaplan-Meier curves with tests of significance tests based on log-rank test. The change in baseline clinical variables compared to baseline at strategic time points of the trial will be analyzed by repeated measures ANOVA factor in for time, treatment conditions and their interaction. The Wilcoxon test will be used in cases that do not meet the assumptions of normality. Between groups differences in main categorical and clinical variables will be analyzed using the Chi-squared test. Those variables that are statistically significant will be used as covariates in a linear logistic regression of factors associated with the magnitude of the effect in order to determine which factors are better predictors of performance. For each of the analyses the effect size index (Hedges g index or Pearson r index) will be estimated. The latest version of the SPSS software (v. 18, 20) will be used to carry out all the statistical analysis. 3. Analysis of clinical efficacy The mains statistical analysis will employ an intention to treat (ITT) approach to provide unbiased comparisons among the treatment groups. The "Last Observation Carried Forward" approach (LOCF) will be employed as a measure of minimizing losses.