Enhancing self-compassion in individuals with visible skin conditions: randomised pilot of the ‘My Changed Body’ self-compassion writing intervention

ABSTRACT Background: Abnormalities in the appearance of skin are commonly associated with compromised self-body perceptions, arising from physical manifestations of the skin condition that deviate from the individual's idealised body image. These body image concerns are associated with a range of psychological issues including anxiety, depression, fear of negative evaluation, and suicidal ideation. Unfortunately, stigma and embarrassment associated with these body image concerns mean that these issues are rarely discussed in clinical medical consultations. There is thus a need for highly accessible and acceptable interventions to address skin-related body image concerns. We have previously demonstrated that a web-based self-compassion focused therapeutic writing approach, the ‘My Changed Body’ intervention, is efficacious in addressing body image concerns of women in the breast cancer context. The aim of this experimental pilot study was to investigate the feasibility of applying the My Changed Body intervention to address visible skin-related body image concerns. Methods: Participants (N = 50) with a range of visible skin conditions provided online informed consent, then completed measures of demographic and medical history, body image disturbance, self-compassion and positive and negative affect. They were then randomly allocated either to an active control expressive writing condition (n = 25) or to the My Changed Body writing condition (n = 25). Participants were blind to their condition allocation. Immediately after completing their allocated writing exercise, participants completed self-compassion and affect measures. Results: Controlling for pre-writing body image disturbance, repeated measures ANCOVAs with fixed effects revealed that self-compassion and negative affect significantly improved after the My Changed Body writing exercise, compared to the control condition. There was no between groups difference at follow-up in positive affect. Conclusions: This study suggests that the My Changed Body writing intervention may provide benefit to individuals with visible skin conditions. A randomised controlled trial is needed to further confirm these results.


Introduction
Skin conditions (e.g. eczema and psoriasis) represent the fourth-most common cause of non-fatal disease burden globally, particularly in Western countries such as Australia, the United States of America and Northern Europe (Baker, Foley, & Braue, 2013). In the United Kingdom, almost one in two (54%) will experience some kind of skin condition each year, with up to 25% of primary care consultations focussing on skin conditions (Bundy et al., 2013). The skin, as the largest and most visible organ of the body, is strongly implicated in an individual's sense of wellbeing and attractiveness (Rumsey, 2018). Stigmatisation based on physical appearance is commonly experienced by individuals with visible skin conditions and can lead to social rejection (Bundy et al., 2013;Ginsburg & Link, 1993), disparaging labels from others (Kellett, 2014), and impaired intimate relationships (Baker et al., 2013;Gupta & Gupta, 1998). Such stigmatisation has been linked to depression, anxiety and low self-esteem (Hong, Koo, & Koo, 2008;Murray & Rhodes, 2005;Picardi, Mazzotti, & Pasquini, 2006). Notably, degree of psychosocial impairment is not directly related to the severity of the skin condition symptoms (Gupta & Gupta, 1998;Law, Chuh, Lee, & Molinari, 2010;Picardi et al., 2006).
Given the highly visible nature of skin disorders, body image is frequently adversely impacted. Body image is defined as an individual's perceptions, thoughts, feelings, and behaviours associated with their physical appearance (Sarwer & Spitzer, 2015). In the case of individuals with a dermatological condition, the physical manifestations of the skin condition that deviate from the individual's idealised body image will compromise their sense of self, leading to feelings of body image disturbance (Cash, 2011;Cash & Szymanski, 1995;White, 2000). This skin condition-related disturbance of body image has been associated with increased insomnia (Gupta, Gupta, & Knapp, 2015), anxiety, depression (Hinkley, 2015), fear of negative evaluation, suicidal ideation (Gupta & Gupta, 2013) and diminished self-esteem (Kouris, Platsidaki, Kouskoukis, & Christodoulou, 2017). Reluctance to seek professional help with managing these body image concerns is associated with feelings of embarrassment and shame (Ahmed et al., 2018;Kouris et al., 2017), consistent with research reporting that bodily embarrassment is associated with less frequent medical contact, generally (Consedine, Krivoshekova, & Harris, 2007). Stress, anxiety, negative affect (emotion), and behavioural habits (e.g. scratching) can exacerbate skin conditions leading to flare ups or poor control (Shenefelt, 2010).
A number of psychosocial interventions have been developed to manage skin conditions, sometimes in conjunction with frequently-prescribed pharmacological therapies (Rumsey, 2018). These psychotherapeutic approaches include interventions such as cognitive behavioural treatment (CBT), hypnosis, and biofeedback, all of which tend to be time-and labour-intensive, requiring in-person specialist consultation (Shenefelt, 2010). To date, only CBT has demonstrated promising results (Rumsey, 2018). Online CBTfocused interventions, such as eTIPs (for psoriasis; Bundy et al., 2013) and FaceIT/YPFa-ceIT (for visible skin conditions in adults and young people respectively; Bessell, 2012;Williamson, Griffiths, & Harcourt, 2015) provide an efficacious and alternative means of accessing psychological support whilst overcoming geographical barriers to accessing in-person trained psychotherapy providers. However, these online interventions have experienced high attrition rates (Rumsey, 2018). Dropout from online interventions, generally, is a common limitation (Karyotaki et al., 2017;Köhle et al., 2015;Richards & Richardson, 2012;Sundström, Blankers, & Khadjesari, 2017) and may be related to intervention users feeling overwhelmed by the time commitment and intensity of material to cover (Beatty et al., 2017;Finlay-Jones, Kane, & Rees, 2017;Kuijpers, Groen, Aaronson, & van Harten, 2013;Postel, de Haan, Ter Huurne, Becker, & de Jong, 2010). Briefer, less intensive online interventions potentially provide a useful alternative approach to reach those individuals lacking in the time or motivation to engage with more complex online interventions. There is, therefore, a need for a brief and readily accessible online intervention to address body image concerns associated with skin conditions that are easy to use and acceptable to this population, as left untreated these difficulties may contribute to further psychological distress and compromised quality of life (Ahmed et al., 2018).
In an illness context, higher self-compassion has been associated with lower negative affect, anxiety and depression (i.e. influenza, asthma, cancer), lower stress (i.e. HIV+, chronic illness, cancer), and greater quality of life (Pinto-Gouveia et al., 2014). In the cancer context, there is evidence that individuals with low self-compassion are more likely to experience difficulties in romantic interpersonal relationships (Shaw, Sherman, Fitness, Elder, & Breast Cancer Network Australia, 2018) which are related to ongoing body image disturbance (Przezdziecki et al., 2013). There is also evidence from clinical populations of individuals with body image concerns arising from eating disorders and following physical illness, including cancer treatment, that greater self-compassion is associated with less psychological pathology and body image distress (Braun, Park, & Gorin, 2016;Duarte, Ferreira, Trindade, & Pinto-Gouveia, 2015;Przezdziecki et al., 2013;Van Dam et al., 2011). Taken together, this evidence indicates that self-compassion is a key factor in adjustment to challenging experiences, such as chronic and serious health conditions (Pinto-Gouveia et al., 2014;Sherman et al., 2018), and may be a helpful emotion regulation strategy in which to address body image difficulties associated with skin conditions by encouraging willingness to experience negative emotions, and accepting them as part of human experience (Neff, 2003;Odou & Brinker, 2015;Rodgers et al., 2018).
Self-compassion has predominantly been conceptualised as a trait, however, targeted interventions in healthy (Arch et al., 2014;Finlay-Jones et al., 2017;Neff & Germer, 2013) and physical illness (Braun et al., 2016;Sherman et al., 2018) populations have demonstrated its modifiable and trainable nature. It has been proposed that self-compassion training acts by promoting emotion regulation (Arch et al., 2014) and coping in the face of difficult experiences (Allen & Leary, 2010). In physically healthy adolescent populations with body image concerns, self-compassion focused interventions entailing meditations (Albertson, Neff, & Dill-Shackleford, 2015) and weekly training on being non-judgemental (Atkinson & Wade, 2015, demonstrated reductions in body image disturbance and body shame, and improvements in body appreciation. Given the reluctance for individuals with skin condition-related body image concerns to seek psychotherapy, and the advantages in accessibility and dissemination, internet-based interventions provide a viable means through which to deliver therapeutic interventions for this populations. Internet-based interventions, generally, have demonstrated efficacy in addressing body image difficulties. A review of 20 digital health interventions addressing eating disorder symptoms in adolescents and young adults, reported that compared with control conditions, internet-based interventions led to significant improvements in body image and reductions in negative affect (Melioli et al., 2016). Further, another digital health intervention addressing body image concerns in adolescents, BodiMojo, specifically applied a self-compassion focused therapeutic approach through a mobile application (Rodgers et al., 2018). Evidence from a randomised controlled trial of BodiMojo provides preliminary evidence that this intervention led to improvements in self-compassion and appearance-related esteem (an indicator of body image). Thus, there is a growing body of evidence supporting both the use of self-compassion focused interventions for promoting body image (Braun et al., 2016) and for digital interventions (Melioli et al., 2016;Rodgers et al., 2018) to provide benefit in this context.
One promising intervention approach designed to enhance self-compassion, and thereby potentially address psychological concerns is through self-compassion focused therapeutic expressive writing (Leary, Tate, Adams, Batts Allen, & Hancock, 2007;Sherman et al., 2018). Expressive writing is a form of emotional expression in which a person writes about a major life event or distressing concern, without regard for audience, form or other writing conventions (Merz, Fox & Malcarne, 2014). This type of expressive writing has been demonstrated through meta-analysis to improve functioning in a range of domains including self-reported physiological health, psychological health and general functioning (Smyth, 1998). Integrating the advantages of a digital intervention approach for addressing body image concerns (Melioli et al., 2016;Rodgers et al., 2018), we have developed a web-based therapeutic writing intervention known as 'My Changed Body' (MyCB), that modifies the concept of therapeutic expressive writing about one's deepest thoughts and emotions (Pennebaker, 1997;Stanton et al., 2002), by providing guided prompts with a focus on self-compassion (Przezdziecki, Alcorso, & Sherman, 2016), similar to that developed by Leary et al. (2007). The MyCB intervention aims to encourage the writer to perceive negative and distressing experiences in a supportive and caring manner . This brief, one-session intervention was originally developed to address body image concerns and associated psychological distress of women recovering from breast cancer surgery and treatment . Evidence from a large longitudinal randomised controlled trial demonstrated that, compared with expressive writing controls, breast cancer survivors undertaking the self-compassion focused MyCB writing reported significantly less body image distress and enhanced perceptions of their body image up to three months following the writing intervention . Moreover, these improvements were demonstrated to be mediated by enhancements in self-compassion at one week following exposure to the MyCB writing intervention , which in turn impacted the body image perceptions. Notably, the subset of participants in this study who also had a diagnosis of lymphoedema (a disabling and highly visible sideeffect of cancer treatment) and who experienced the greatest body image-related distress at study entry, experienced significant reductions in both depression and anxiety up to three months following completion of the MyCB self-compassion writing, compared with controls .
As a low-cost alternative to group-administered programmes, the MyCB web-based intervention provides a time-efficient, low-burden  therapeutic approach for addressing body image concerns. Since MyCB is based on enhancement of self-compassion through structured writing concerning negative experiences related to bodily disfigurement or negative changes, this therapeutic methodology is eminently translatable to other physical illness contexts. The benefit of writing about personal distress appears to be greatest in those who belong to a stigmatised group (Richards, Beal, Seagal, & Pennebaker, 2000), such as individuals with a visible skin condition. Moreover, a more time-intensive self-compassion-based training intervention has demonstrated reductions in shame and improved psychosocial and emotional functioning among people with acne (Kelly, Zuroff, & Shapira, 2009). Therefore, in light of the ever-increasing access of the general Australian population to the internet (Australian Bureau of Statistics, 2018), which is estimated to be more than 85%, the web-based MyCB intervention may provide a viable way to support individuals with visible skin conditions experiencing body image concerns. The aim of this study was to conduct a pilot trial of the MyCB intervention to assess feasibility of this therapeutic approach being translated to the visible skin condition context. We therefore hypothesised that the MyCB intervention would enhance self-compassion and modify emotions such that there would be a significant decrease in negative affect and increase in positive affect, compared to the active control condition consisting of an expressive writing exercise without self-compassion based writing prompts. By demonstrating that the MyCB intervention can produce immediate shifts in self-compassion and affect, mimicking prior work using this intervention , then this would justify a larger RCT to further assess the utility of the MyCB intervention in this population.

Participants
Individuals over 18 years who were experiencing a currently symptomatic and visible skin condition (e.g. eczema, psoriasis, acne) and who had experienced at least one negative event related to their skin condition (entailing feelings of failure, humiliation or rejection) were eligible for this study. All participants would need to complete an online Englishlanguage writing activity; hence, internet access was a requirement. Participants were recruited from a range of sources including: first year undergraduate psychology students (n = 26) from Macquarie University, Sydney, who self-enrolled into this study for course credit; members of a skin condition-related Facebook group (n = 15) responding to a Facebook posting about the study, and patients at a Sydney-based dermatology clinic (n = 9) responding to invitational flyers posted in the clinic waiting room. Facebook and clinic participants received no compensation for study participation. A total of 50 eligible individuals provided informed consent and participated in the study (see Figure 1). An additional 22 individuals provided informed consent, but did not proceed with completing the study questionnaires, and were therefore excluded from analyses.

Procedure
Following online informed consent, all participants completed the pre-writing questionnaire and were randomly and equally assigned using the Qualtrics randomiser function to either control or experimental writing conditions (both completed online) such that there were 25 participants in each condition; participants were blind to their condition allocation. Participants then immediately completed their assigned writing following specified instructions, and immediately thereafter completed a follow-up questionnaire (see Online Supplemental File for study protocol).
Study Conditions: Experimental condition was adapted from the MyCB self-compassion focused writing intervention Sherman et al., 2018) developed for use with breast cancer survivors. In this instance, participants were first instructed to briefly introduce and describe in writing their deepest thoughts and emotions regarding a negative event they had experienced in relation to their visible skin condition. After this brief written introduction, MyCB participants were then provided with five specific self-compassion focused prompts to structure their writing based on the concept of self-compassion (Neff & Dahm, 2015). The writing was undertaken in the context of having a visible skin condition, and the prompts addressed treating one's body with kindness, giving kind advice to the self, having connection with others who also experience body image difficulties, awareness of one's circumstances and reactions in a broader context, and writing a self-compassionate letter to oneself. The Control condition entailed participants writing in an unstructured expressive manner about their deepest thoughts and emotions of a negative event they had experienced in relation to their skin condition (Pennebaker & Beall, 1986), commencing with a general description of topic (as per the Experimental condition) followed by general instructions to continue writing ('Please describe the event further'). The control accounted for the potential effect of time taken to complete the writing task. Both writing conditions were allocated a maximum of 30 minutes writing time (see Online Supplemental File for writing prompts).

Measures
Self-compassion before and after the writing tasks was measured with the 12-item Self-Compassion Scale Short Form (SCS-SF) (Raes, Pommier, Neff, & Van Gucht, 2011) assessing components of self-kindness, self-judgment, common humanity, isolation, mindfulness, and over-identification. Items such as 'I try to be loving towards myself when I'm feeling emotional pain' were rated on a 5-point scale (1 'almost never' to 5 'almost always') and a mean was calculated (range 1-5) with higher scores indicating greater self-compassion (Cronbach's α = .90).
Affect before and after the writing tasks was measured with the 20-item Positive and Negative Affect Schedule (PANAS) consisting of the 10-item positive (Cronbach's α = .84) and negative (Cronbach's α = 82) affect subscales (Crawford & Henry, 2004;Watson, Clark, & Tellegen, 1988). Participants rated on a 5-point scale (1 'very slightly or not at all' to 5 'extremely') the extent to which they felt each of the 20 emotions in that present moment, such as 'distressed' or 'proud'. Summed scores were obtained for each subscale (range 10-50).
Body Image Disturbance at study entry was treated as a covariate and measured with the 7-item Body Image Disturbance Questionnaire (Cash, Phillips, Santos, & Hrabosky, 2004). This scale was chosen as it assesses body-image specific distress rather than general psychological distress. The scale shares a positive relationship with depression, anxiety, and stress scores as measured by the Depression Anxiety and Stress Scale (Collison & Mahlberg, 2018). Items such as 'do you ever avoid things because of your physical defect?' were rated (1 'never' to 5 'very often') and higher mean scores indicated greater body image disturbance (α = .82; range 1-5).
Perceived skin condition severity participants' subjective level of perceived severity of their skin condition was assessed with a single Likert-type scale ranging from 1 'Low Severity' to 5 'High Severity'.
Demographic and Medical Characteristics documented pre-allocation included age, gender, education level, skin condition type, time since skin condition onset, whether treatment was received for the skin condition (yes/no). All demographic and medical characteristics and perceived condition severity were treated as potential covariates.

Data analysis
All analyses were performed in SPSS version 25.0 (IBM Corp., 2017) unless otherwise specified. Significance was set at 0.05 for all tests. Descriptive analyses outlined sample characteristics. Chi-square and t-test analyses were used to identify covariates (i.e. testing for significant differences across pre-allocation outcome measures and sample characteristics between condition allocation and recruitment sourcestudents vs. community). To quantify the extent of body image disturbance found in the present study, a two-sample t-test was conducted in the R statistics software (R Development Core Team, 2008) to compare the mean and standard deviation of body image disturbance found in our study to that of the sample on which the scale was originally validated. The R statistics software was required for this analysis as SPSS does not have the functionality for t-test comparisons based on summary data (i.e. means and standard deviations in absence of the raw data). Repeated measures ANCOVAs with mixed effects were then conducted for each follow-up outcome by study condition, controlling for the relevant identified covariates. Sample size was determined by the number of participants who selfenrolled into the study between March and September 2016.

Ethics Statement
Institutional ethics approval was granted (Macquarie University HREC #5201600318; Trial Registration: ACTRN12618002025213). Table 1 outlines pre-allocation sample characteristics and comparisons across condition allocation. The sample sources differed by gender, age, education, and perceived symptom severity and were therefore included as covariates in subsequent analyses. Mean body image distress scores in the skin condition sample were significantly higher in the current study (M = 2.40, SD = 0.98) than for the non-clinical sample on which the measure was originally validated (M = 1.70, SD = 0.96). This confirms the inclusion criterion that participants in this study with skin conditions were experiencing elevated body image distress (t = 4.52, p < .0005; Cash et al., 2004).

Discussion
The aim of this study was to determine the feasibility of translating the evidence-based MyCB self-compassionate writing intervention to a health context other than breast cancer. Specifically, the efficacy of the MyCB intervention  for addressing body image concerns of breast cancer survivors has previously been demonstrated in a large randomised control . In this feasibility study we utilised a pre-post experimental design and applied the MyCB writing intervention approach to individuals experiencing body image concerns arising from symptoms of a visible skin condition. This was compared to an active control condition who undertook an expressive writing activity regarding their body image concerns, without receiving the self-compassion prompts. This active control condition was a strength of the design as it accounts for time and attention in the writing task, ensuring that ensured that the study results reflected the self-compassion elements of writing, rather than general expressive writing. Study participants reported greater levels of body image disturbance than those in the original validation study of the body image disturbance questionnaire, confirming that the sample investigated in this study were generally experiencing elevated body image disturbance ( Figure 4).
Following randomisation, participants in the MyCB writing condition reported significant improvements in self-compassion and negative affect compared with those in the  expressive writing active control condition. These findings are consistent with that of the initial MyCB intervention feasibility study conducted within the breast cancer context  and Leary et al. (Leary et al., 2007) who both experimentally manipulated self-compassion through writing. This reinforces the utility of a brief single dose intervention, such as the MyCB intervention, to improve negative affect and self-compassion in a different health context in which it was originally developed and trialled. Despite self-compassion theoretically having the capacity to enhance positive affect (Neff et al., 2018;, there was no significant difference in positive affect between the experimental and control conditions in this pilot study, mirroring the findings of the prior MyCB intervention feasibility study undertaken in the breast cancer context . Given that the large RCT demonstrated reductions in anxiety and depression that were maintained up to three months post-intervention , it is possible that the impact of the MyCB intervention may be predominantly on ameliorating negative affect, rather than enhancing positive affect. Furthermore, it is possible that improvements in positive affect may take time to emerge and may not have been immediately detectable, given that this pilot study only measured outcomes immediately after participants engaged in the writing  exercises. From a theoretical perspective, this pilot study adds to the growing body of literature (Braun et al., 2016;Sherman et al., 2018) supporting the use of self-compassion as a therapeutic intervention to address body image concerns.
The findings of this study need to be considered in light of the fact that this was a feasibility study. The sample size was relatively small, and there was only an immediate followup. Larger sample sizes would increase generalisability while longer follow-ups would ensure that the reported effects maintain over time, and may reveal delayed benefits of the intervention, such as possibly positive affect. A larger RCT of MyCB in this population could address these outcomes. It is also important to acknowledge that past research employing online interventions have had high drop-out rates, potentially leaving only highly motivated individuals in the study (Bessell, 2012;Bundy et al., 2013;Radtke, Ostergaard, Cooke, & Scholz, 2017;Richards & Richardson, 2012;Sundström et al., 2017;Williamson et al., 2015). This could potentially be biasing the results of online interventions. While we did not have a low drop out in this study, it is possible that only motivated individuals signed up. As such, future work should look at the efficacy of this intervention in a naturalistic setting.
Nevertheless, these findings have implications for the clinical management of body image concerns in individuals with visible skin conditions. MyCB is a private, single dose, low cost writing intervention that can be self-administered anywhere via the internet independent of clinical support. As such, patients with visible skin conditions would benefit from being directed towards self-compassionate writing exercises, such as MyCB, in conjunction with standard care. This would not only increase the time available for clinicians to work on treating the skin condition, but also help to alleviate some of the negative affect that one might experience as a result of body image concerns related to the skin condition.

Disclosure statement
No potential conflict of interest was reported by the authors.