This PhD thesis consists of four original papers, including a systematic review and meta-analysis, a pilot-study, and the to date largest nationwide randomized controlled trial examining the effects of expressive writing intervention (EWI) on psychological adjustment, somatic symptoms, and health-related quality of life among women treated for primary, loco-regional breast cancer.
The available literature, including studies originating from the Psychooncology Research Unit, Aarhus University Hospital, indicates that a significant proportion of women treated for breast cancer report increased levels of distress and reduced quality-of-life. Data from our ongoing national cohort of women treated for primary breast cancer show that only half of the women have received any medical or psychological treatment or support for their distress, indicating a need for exploring brief cost-effective interventions aimed at reducing cancer-related distress. One such intervention is expressive writing intervention. EWI is a brief psychological intervention, which instructs participants to disclose - in writing - their inner thoughts and feelings in relation to a stressful life event. Numerous controlled studies have shown that participants assigned to write about major emotional challenges in their lives show beneficial effects across a number of outcomes, including psychological adjustment, biological effects, and health outcomes. Whereas most of the early studies of EWI focused on healthy students, more recent research has demonstrated that EWI may also improve psychological and physical health outcomes in some medical populations.
The research presented in this thesis addressed six main objectives: Objective 1 – To review relevant findings from the literature regarding the effects of EWI and to conduct a systematic review and quantitative analysis of the findings on EWI with cancer patients. Objective 2 – To explore the feasibility of all study procedures by conducting a small pilot study testing all aspects of the main project protocol. Objective 3 – To identify the characteristics of the nationwide cohort at baseline with respect to the prevalence of somatic symptoms, depression, distress, and sociodemografic patterns, and to compare the present sample with a large Danish cohort of 3500 Danish women treated for breast cancer. Objective 4 – To investigate the impact of EWI on the outcomes of cancer related distress, mood and depression and to explore the influence of a number of possible moderating factors, e.g. repressive coping, rumination, social constraint, intrusive thoughts and alexithymia. Objective 5 – To investigate the impact of EWI on self-reported physical health and health care utilization, as well as the possible moderating factors of repressive coping, rumination, social constraint, intrusive thoughts and alexithymia. Objective 6 – To suggest how to design future studies of EWI and to outline a short recommendation regarding a possible implementing of EWI in health care settings.
Paper 1 is a literature review and meta-analysis of studies of expressive writing intervention (EWI) focusing on health outcomes in samples of cancer patients. A list of randomized clinical trials (RCT) of EWI with cancer patients published before December 2009 was compiled using relevant search engines and previously published articles and reviews. Meta-analyses were conducted to estimate the overall mean weighted effect sizes, both across all available studies and for studies grouped by the quality of RCTs and by outcome categories. The search identified a total of 12 publications on emotional disclosure interventions and cancer patients. EWI had been used in 10 of the studies (n=661) with number of participants ranging from 30 to 234 and an average sample size of 82.6. Across all available studies, EWI did not show the hypothesized therapeutic effect on either psychological or physical outcomes. However, when including only studies using true experimental designs with writing-control groups, a significant effect was found for self-reported physical health.
Paper 2 describes the results of a pilot-study with the aim of exploring the feasibility of a home-based expressive writing intervention (EWI) with women treated for primary breast cancer. First, the feasibility of the experimental design was assessed by testing all procedures according to protocol. Second, a manipulation check of the intervention was conducted by testing the immediate effects on participant’s mood state and by qualitatively analyzing the written material. Patients treated for primary breast cancer (N=16; mean age 51.6) were recruited from the local hospital after having completed their surgical treatment, chemo-, and radiation-therapy. They were contacted by phone and randomly assigned to write about their most traumatic experience or a neutral topic for 20 minutes once a week for three weeks. Due to the small number of participants, no statistically significant effects were expected. Standardized measures of distress were mailed to the participants at baseline and three weeks after the last writing session. The intervention was concluded to be feasible in this patient setting, and as expected no results reached statistical significance.
Papers 3 and 4 describe the results of the randomized clinical trial of the effects of expressive writing intervention (EWI) on cancer-related distress, depression, and negative mood (Paper 3), and self-reported physical health and healthcare utilization (Paper 4), in a nationwide randomized clinical trial with Danish women treated for early stage breast cancer. Another aim was to explore writing topic, alexithymia, repressive coping, rumination, and social constraints as potential moderators of such effects. Four weeks after completed treatment for primary cancer the participating women (n=507; mean age: 53.6) were randomly assigned to three 20 min. home-based writing exercises, one week apart, focusing either on emotional disclosure (EWI) or a non-disclosing topic (control). The participants completed questionnaires immediately before and after writing as a manipulation check. Standardized measures were completed at baseline and at 3-months (n=435), and 9-months (n=230) post-intervention. The results do not support EWI as a general applicable intervention to reduce either cancer-related distress, depressive symptoms, negative mood or somatic symptoms and health care utilization in the general population of women with primary breast cancer. However, moderator analyses indicated that choosing to write about non-cancer experiences was associated with the larger reductions in cancer-related distress. While no support was found for social constraints as a moderator of EWI on cancer-related distress, the results suggested one component of alexithymia, difficulties describing feelings, as a potential moderator to be further studied. In relation to health care utilization the intrinsic privacy of the written disclosure task may have a stronger impact for individuals with high social constraints, and in relation to somatic symptoms it may have stronger impact for repressive copers. It appears relevant to include different measures of emotional inhibition in future studies of expressive writing intervention to further improve our understanding of for whom this intervention is beneficial.