The prevalence rates of body image concerns among women, including dissatisfaction with one’s appearance or body weight or shape, continue to increase (
Pruis & Jaworsky, 2010). Body image concerns remain relatively stable across the life span (
Tiggemann, 2004). Even among older women, thoughts about weight occur on a daily basis, and many women report that thoughts about weight or body shape can occasionally negatively affect their well-being (
Gagne et al., 2012). Body dissatisfaction is positively associated with a myriad of psychological and physiological health consequences such as depression and anxiety (
Jacobi et al., 2004;
Szymanski & Henning, 2007), restricted and emotional eating (
Johnson & Wardle, 2005), and internalization of problematic sociocultural values related to appearance, such as valuing and striving for thinness (
Homan, 2010).
Although body image is often conceptualized as a stable, trait-like construct that describes an individual’s typical affective valence toward their body, body image is also fluid and subject to daily fluctuations (
Melnyk et al., 2004). Research examining fluctuations in women’s body image demonstrate that negative shifts in body satisfaction (i.e., periods of lowered body satisfaction) are associated with maladaptive coping strategies that increase the importance of appearance and eating disorder attitudes and behaviors (
Melnyk et al., 2004). Considering that fluctuations in evaluation (i.e., affective evaluation of appearance) and investment (i.e., relative importance of one’s appearance in one’s life) in body image are likely to occur and have consequences on women’s eating behaviors (
Carraca et al., 2011), writing interventions aimed to help women adaptively cope with daily body image experiences were used. Our knowledge of how investment in appearance-focused goals is associated with pathological eating behaviors, such as restraint (restricting calories) and disinhibition (binge eating;
Putterman & Linden, 2004;
Thøgersen-Ntoumani et al., 2010), is growing, so interventions aimed to increase positive affective components of body image, such as body satisfaction and appreciation, and lower the saliency of appearance-focused goals are needed.
Current self-guided writing interventions demonstrate promising results, such that they are effective in reducing body dissatisfaction, eating disorder symptoms, and bolstering body acceptance. For instance, expressive writing interventions that increase self-compassion (
Leary et al., 2007;
Moffitt et al., 2018;
Seekis et al., 2017;
Stern & Engeln, 2018) or self-esteem (
Leary et al., 2007;
Seekis et al., 2017) have been shown to increase body appreciation (
Seekis et al., 2017) and body satisfaction (
Stern & Engeln, 2018) and reduce eating disorder symptoms (
Kelly & Carter, 2015). Cumulative evidence suggests that self-compassion interventions are especially useful to help women regulate their emotions after reflecting on aspects of their bodies (
Stern & Engeln, 2018), which helps mitigate negative body image and disordered eating (
Turk & Waller, 2020). However, limited research exists on comparing different types of multisession writing interventions that target different protective factors (e.g., self-compassion or self-esteem) and examining their ability to reduce investment in appearance goals and bolster positive body image. Limited research also exists on evaluating the efficacy of these interventions on reducing pathological eating behaviors and increasing adaptive healthy eating behaviors (
Biber & Ellis, 2019). This is especially important considering that the absence of pathological functioning does not infer optimal or adaptive functioning (
Westerhof & Keyes, 2010). Finally, these interventions have largely been evaluated in student samples; therefore, examining their effects in a sample that includes women who are older and have more diverse educational backgrounds in addition to college women is required.
In the current study, we sought to examine the efficacy of a multisession body-focused self-compassion and self-esteem writing interventions compared to a control writing condition on women’s valuation of health-oriented (vs. appearance-oriented) weight management goals, body appreciation, bulimic symptoms, and self-reported healthy (vs. unhealthy) eating behaviors. Single session self-compassion and self-esteem writing interventions were effective in reducing state body dissatisfaction when reflecting on a negative body image scenario compared to a control group (
Seekis et al., 2017); however, reframing these events in a self-compassionate manner more robustly affected different aspects of women’s body image immediately following the session compared to bolstering positive self-evaluations (
Moffitt et al., 2018;
Seekis et al., 2017). The potential for these trends to hold or become more robust if women received multiple exposures to these interventions and for these changes to extend to other important cognitive, affective, and behavioral indices while controlling for important covariates (i.e., age, body mass index [BMI], number of sessions) remains underexplored (
Seekis et al., 2017). There is evidence to suggest that multisession self-compassion interventions may outperform those employing a single session on improving health behaviors and overall physical health (
Phillips & Hine, 2019); therefore, it is imperative to examine whether multisession body-focused self-compassion interventions yield more benefits on body image and eating behavior. This would help inform the optimal duration of body-focused writing prevention or intervention programs. Additionally, comparing body-focused self-compassion and self-esteem writing interventions can derive fruitful information regarding the effects of self-compassion relative to self-esteem when coping with daily negative experiences that may affect body image.
Self-Compassion as a Protective Factor
Self-compassion has been identified as a protective factor in body image and eating disorder literature (
Turk & Waller, 2020), as it is a method of compassionately relating to the self that focuses on experiencing and approaching failures openly, with kindness or acceptance, and with a sense of relatedness to others (
Neff, 2003b). Individuals who embody self-compassion exhibit capabilities to overcome feelings of failure or inadequacy by relating to the self in the following ways: (a) enacting care toward oneself with unconditional acceptance in order to overcome critical and judgmental evaluations of the self (self-kindness vs. self-judgment), (b) enacting acceptance of personal flaws or difficult life circumstances as part of being human in order to overcome feelings of isolation or helplessness (community humanity vs. isolation), and (c) enacting awareness and perspective taking when evaluating negative experiences in order to prevent avoidance or rumination (
Neff, 2003b).
In particular, self-compassion fosters resilience when facing distressing situations involving failure, rejection, embarrassment, or other negative life events (
Leary et al., 2007). In response to these situations, individuals high in self-compassion are less likely to ruminate or experience negative affect, more likely to accept undesirable personal traits and behaviors, and more likely to take responsibility for negative events without becoming defensive or feeling badly about themselves (
Leary et al., 2007). This type of perspective taking may protect women from experiencing daily fluctuations in physique anxiety, drive for thinness, and body dissatisfaction (
Thøgersen-Ntoumani et al., 2017). Furthermore, self-compassion mitigates the negative effects of engaging in social comparisons, particularly those related to eating behavior, on body appreciation (
Siegel et al., 2020) and moderates the relationship between weight concerns and eating pathology (
Stutts & Blomquist, 2018). Additionally, common humanity, a component of self-compassion that involves acknowledging that imperfections and feelings of inadequacy are shared with others, is especially important for women’s body image, most notably their ability to accept and appreciate their bodies despite its imperfections (
Seekis et al., 2020). Taken together, these findings suggest that self-compassion functions as an adaptive coping response to deal with appearance distress, including appearance distress that may arise from engaging in social comparisons (i.e., perceiving a disparity between oneself and others). Additionally, self-compassion offers resilience by reducing both risk factors (i.e., body shame) and experiencing those risk factors as distressing (i.e., weight concerns;
Breines et al., 2014).
Taking a self-compassionate perspective is also beneficial when individuals are dealing with setbacks related to their engagement in health behaviors (e.g., ingesting foods that are less nutritious, engaging in suboptimal levels of exercise). A recent meta-analysis that comprised 15 samples (
Sirois et al., 2015) suggested that self-compassion replenishes resources for self-regulation by enhancing emotion regulation, such as lowering negative affect and increasing positive affect, which partly explains the positive relationship between self-compassion and health promotion (
Sirois et al., 2015). These findings suggest that even when individuals fail to adhere to a specific health behavior plan (e.g., eat less deep-fried foods), being compassionate and understanding can help replenish cognitive resources to self-regulate a particular behavior, leading to higher behavioral engagement and successful execution of the desired behavior. Self-compassion is also positively associated with the satisfaction of key psychological needs, such as autonomy, competence, and relatedness, which are essential for self-regulation and psychological well-being (
Gunnell et al., 2017). Daily fluctuations in the satisfaction and frustration of these needs also play a role in binge-eating in women, whereby higher need frustration is associated with more episodes in a given day (
Verstuyf et al., 2011). The positive influence self-compassion has on basic psychological needs is another potential mechanism that explains its buffering role on eating disorder symptoms. Furthermore, self-compassion is positively associated with other key motivational constructs that support health behavior engagement. For instance, self-compassion is positively associated with self-improvement motivation (
Breines & Chen, 2012), which may explain why self-compassionate individuals are shown to engage in weight-managing behaviors to promote health and are more likely to self-regulate their eating behaviors for more self-determined reasons (i.e., because it is aligned with other important beliefs), which leads to higher engagement in healthy eating behavior (
Guertin et al., 2018).
Reviews of the literature demonstrate the robustness of self-compassion interventions on reducing negative body image and eating pathology (
Biber & Ellis, 2019;
Turk & Waller, 2020) and improving the self-regulation of health behaviors (
Biber & Ellis, 2019). The efficacy of self-compassion interventions on these outcomes is mainly due to promoting adaptive affect regulation (
Turk & Waller, 2020) by lowering defensiveness via treating oneself with compassion and unconditional acceptance and, in turn, reducing negative emotional states and self-blame, which can interfere with self-regulation and self-care (
Terry & Leary, 2011). Brief writing interventions in samples of women increase state positive affect (
Stern & Engeln, 2018) and reduce state rumination, state self-criticism (
Mosewich et al., 2013), and restrained eating (
C. A. Adams & Leary, 2007). In particular, self-compassionate writing that focuses on coping with distressing body-related scenarios reduces drive for thinness (
Seekis et al., 2020), increases body appreciation (
Seekis et al., 2017,
2020), and reduces disinhibiting eating (e.g., binge eating;
Kelly & Carter, 2015). Compared to a wait-list control group, changes in drive for thinness and body appreciation held for up to 3 months in a 2-week Facebook-enhanced mindful self-compassion intervention (
Seekis et al., 2020). Additionally, changes in state body appreciation in a single writing session self-compassion intervention held for up to 2 weeks when compared to a control writing group (
Seekis et al., 2017). These findings suggest that when women are guided to become more self-compassionate toward themselves, they experience more balanced emotions, are less likely to overidentify with their emotions linked to the event and ruminate (i.e., mindfulness), and are less likely to criticize themselves or their actions (i.e., self-kindness). These adaptive coping strategies improve women’s emotion regulation when facing appearance distress and protect them from developing negative appearance perceptions and, in turn, can help reduce eating disorder behavior as a result.
Self-Esteem as a Protective Factor
In women diagnosed with an eating disorder, self-esteem was negatively associated with eating disorder symptoms and eating and body shape concerns (
Kelly et al., 2014). Although there is limited evidence that self-esteem and body image are causally related (
Wichstrom & Von Soest, 2016), self-esteem is regarded as a protective factor, which is often targeted in eating disorder prevention programs to lower the prevalence of body image concerns and engagement in maladaptive weight-controlling behaviors among young women (
O’Dea & Abraham, 2000).
Eating disorders are considered a disorder rooted from longstanding issues with the self, identity, and individuation (
Amianto et al., 2016), so interventions focused on self-development and self-evaluation are tailored toward bolstering multiple components of self-esteem, such as self-worth (i.e., personal value), self-perception (i.e., satisfaction with the self), and perceived abilities or competence (
O’Dea & Abraham, 2000). Most notably, a multisession educational program called “Everybody’s Different” (
O’Dea, 1995) delivered to secondary students focused on ways to create a positive sense of self by embracing individuality and learning how to receive positive feedback from others (
O’Dea & Abraham, 2000). This intervention was effective in reducing appearance-contingent self-esteem, drive for thinness, and body dissatisfaction among adolescent men and women, which persisted for up to 1 year. Findings from brief writing interventions that increase self-esteem by encouraging positive self-evaluation (i.e., positive skills or traits) also show favorable results. For example, a single session writing intervention that asked women to reflect on a distressing body image scenario in a controlled setting increased women’s body satisfaction immediately after the writing session and at a 2-week follow-up compared to a control writing condition (
Seekis et al., 2017).
Although some interventions tailored toward increasing self-esteem demonstrated positive effects on body image (
O’Dea & Abraham, 2000;
Seekis et al., 2017), others show no benefits when compared to a control condition (
Ghaderi et al., 2005;
McCabe et al., 2006). One reason for these conflicting results may be due to the pitfalls of trying to maintain high self-esteem, such as engaging in social comparisons and placing contingencies on one’s self-worth (
Leary & Baumeister, 2000). Self-esteem, by nature, is partly contingent upon daily appraisals of the self in comparison to others, so it is subject to fluctuation (
Leary & Baumeister, 2000). Due to these fluctuations, self-esteem may not be consistently associated with positive psychological outcomes (
Leary & Baumeister, 2000;
Neff, 2009). Those with more contingent forms of self-esteem (i.e., contingent upon positive appraisals from others) may engage in maladaptive behaviors that impede personal growth processes that are essential for coping with painful personal challenges, such as attributing failures to external causes, trivializing failures, or dismissing information about the self (
Neff, 2009). Additionally, contingent self-esteem has negative consequences on body image (
Grossbard et al., 2009) and self-regulation (
Crocker et al., 2006). For example, contingent self-esteem is positively associated with behavioral disengagement (
Lane et al., 2002), self-deprecation (
Brown & Mankowski, 1993), cognitive rigidity (
Taris, 2000), and reduced self-efficacy (
Lane et al., 2002), which can undermine motivation to change and successful progression toward important life goals. Furthermore, it has been suggested that employing self-esteem writing interventions that do not bolster contingent self-esteem may be challenging (
Leary et al., 2007). Popularly used prompts in writing interventions often focus on affirming participants’ self-images rather than helping them take a more objective, balanced perspective about themselves (
Leary et al., 2007).
Current Study
The literature demonstrates that brief writing interventions, particularly those that are tailored toward promoting protective factors for women’s body image, can create meaningful change in women’s body image (
Seekis et al., 2017,
2020). However, few studies have sought to compare differently tailored writing interventions, such as those rooted in self-compassion or self-esteem (
Moffitt et al., 2018;
Seekis et al., 2017). Furthermore, our knowledge of the superiority of self-compassion writing interventions over self-esteem writing interventions when coping with body-related distress is limited to single exposure interventions (
Moffitt et al., 2018;
Seekis et al., 2017). A recent study suggests that this limitation impedes our knowledge on the true robustness of these writing interventions on producing favorable changes in body image and eating behaviors and the number of sessions needed to reach maximum results (
Seekis et al., 2017). In the current study, we fill these knowledge gaps by employing multisession interventions and comparing their efficacy on cognitive, affective, and behavioral indices related to body image, while examining the number of writing sessions as one of the covariates. Additionally, our knowledge is currently limited on the efficacy of these interventions on improving eating behaviors more broadly; therefore, we examined their ability to reduce pathological eating behavior, while increasing healthy eating behavior. Finally, those who have employed self-compassion or self-esteem writing interventions have infrequently examined if these interventions yielded clinically significant changes in eating disorder symptoms; therefore, we also examined the ability of body-focused self-compassion and self-esteem writing interventions to promote clinically meaningful change in those who meet clinical criteria on the Eating Disorder Inventory-1 Bulimia subscale (EDI-1-BS).
Considering that self-compassion is positively associated with striving toward health weight management goals and that promoting self-compassion can increase body appreciation (
Seekis et al., 2017,
2020), lower eating disorder symptoms (
Turk & Waller, 2020), and increase engagement in health promoting behaviors (
Biber & Ellis, 2019), we hypothesized that women in the body-focused self-compassion writing condition would demonstrate increased valuation of health-oriented (vs. appearance-oriented) goals, body appreciation, and healthy eating behaviors and reduced bulimic symptoms from pre- to post-treatment compared to the body-focused self-esteem and control writing conditions. Self-esteem was shown to increase body satisfaction (
Seekis et al., 2017) and reduce eating disorder symptoms (
O’Dea & Abraham, 2000); therefore, we also hypothesized that women in the body-focused self-esteem writing would demonstrate increased body appreciation and reduced bulimic symptoms compared to the control writing condition, but to a lesser extent than the body-focused self-compassion condition. Finally, it was hypothesized that clinically significant changes in bulimic symptoms would be associated with the type of writing condition. We hypothesized that women allocated to the body-focused self-compassion condition would be more likely to demonstrate clinically meaningful change compared to those in the body-focused self-esteem or control writing conditions due to the robustness of self-compassion interventions on promoting clinically significant change on eating disorder symptoms compared to other interventions (
Kelly & Carter, 2015) and wait-list control (
Kelly & Carter, 2015;
Seekis et al., 2020).
Results
Preliminary Analyses
The research coordinator recorded dropouts and non-compliance after making inferences by examining contents, not quality, of participants’ digital entries. Those who consented and engaged in at least one writing session, but abruptly stopped completing entries, or did not adhere to the minimum requirement of sessions as mentioned in the consent form (i.e., at least half of the sessions) were considered non-compliant and dropouts. Our initial sample comprised 174 participants; however, 48 women (28%) dropped out of the study and were not included in the analyses or sample characteristics. Due to dropout and non-compliance rates, the repeated measure ANCOVAs were slightly underpowered compared to our estimated sample size for achieving 80% power; however, G*Power estimates that the analyses did reach 70% power with our sample size of 126 participants (
Faul et al., 2007). A one-way multivariate analysis of variance (MANOVA) determined that there were no differences between women who dropped out or completed the intervention across all pre-treatment measures,
F(7, 157) = 0.84,
p = .557,
= .036. A χ
2 test of independence also revealed no association between dropout rate and meeting clinical criteria on the EDI-1-BS, χ
2(1) = .34,
p = .561, or dropout rate and condition, χ
2(2) = .20,
p = .904.
Less than 1% of data were missing on items for each variable for pre-treatment measures and no data were missing on post-treatment measures. Missing data in pre-treatment outcomes were considered missing completely at random. The Expectation-Maximization method was used to replace missing values. All outcome variables were considered normally distributed, and there were no multivariate outliers.
Table 1 contains means and standard deviations for pre-treatment and post-treatment scores for each outcome while controlling for covariates.
Assumptions for the 3 (condition) × 2 (time) repeated measures ANCOVAs for the main analyses for linearity and homogeneity of variance were met and examined via scatterplots and tests of homogeneity of variance. For covariates, three assumptions (i.e., linearity, homogeneity of regression slopes, and independence from treatment effect) had to be met for a variable to qualify as a covariate (
Field, 2013). Age, BMI, and the number of journal entries completed met the three assumptions and were used as covariates in the ANCOVAs. Pre-treatment self-esteem scores also met the three assumptions and were used when examining group differences in self-compassion, whereas pre-treatment self-compassion scores met the three assumptions and were used when examining group differences in self-esteem. Adjusting for covariates in randomized controlled trials can influence results (
Kraemer, 2015), so covariates were chosen a priori based on existing literature to reduce this risk.
A one-way MANOVA was conducted to determine pre-treatment differences in outcome measures across conditions. Baseline differences were found, F(16, 218), = 1.74, p = .042,
= .113, but only in variables used as covariates, such as BMI and age. Women in the control writing condition were older than women in the body-focused self-compassion (mean difference = 9.74, p = .009) and body-focused self-esteem (mean difference = 8.83, p = .019) conditions and had higher BMI than those in the body-focused self-compassion condition (mean difference = 5.27, p = .014). A χ2 test of independence determined that there was no association between meeting clinical criteria on the EDI-1-BS and allocation to writing condition, χ2(2) = .92, p = .631.
Table 2 contains bivariate correlations between pre-treatment outcomes. Age was significantly positively associated with BMI, self-compassion, and lower valuation of appearance goals. No significant relationship was found between age and body appreciation and bulimic symptoms. Additionally, self-compassion and self-esteem were significantly associated with body appreciation and bulimic symptoms in the anticipated directions, such that they were both positively associated with body appreciation and negatively associated with bulimic symptoms. However, BMI was not significantly associated with bulimic symptoms, although the direction of the relationship is congruent with previous literature. Finally, self-esteem and self-compassion were significantly positively associated with each other.
Manipulation efficacy
Two 3 (condition) × 2 (time) ANCOVAs were conducted to determine mean differences in self-compassion and self-esteem at each time point between participants across the three conditions. For pre–post differences in self-compassion, there was a non-significant main effect of condition, F(2, 120) = 0.47, p = .629,
= .008, a significant main effect of time, F(1, 120) = 4.86, p = .029,
= .038, and a significant interaction between time and condition, F(2, 120) = 4.62, p = .012,
= .071. Simple effects demonstrated that women in the body-focused self-compassion writing condition increased in self-compassion over time (p < .001,
= .129), whereas women in the body-focused self-esteem (p = .942,
= .000) and control (p = .116,
= .020) conditions did not. For pre–post differences in self-esteem, there was a significant main effect of condition, F(2, 120) = 4.02, p = .021,
= .063, a significant main effect of time, F(1, 120) = 4.46, p = .037,
= .036, and a significant time by condition interaction, F(2, 120) = 3.37, p = .038,
= .053. Simple effect analyses demonstrated that women in the body-focused self-compassion (p = .006,
= .060) and body-focused self-esteem (p = <.001,
= .193) conditions increased in self-esteem over time, whereas women in the control condition did not (p = .150,
= .017). Based on the effect sizes, the body-focused self-esteem condition increased the most in self-esteem. Overall, the writing interventions were considered successful based on these findings.
Inter-rater reliability
There was strong agreement between both raters when rating how much each participant’s entry exhibited resemblance to the prompts used in each condition (k = .89, p < .001). Raters also had high accuracy when asked to indicate which condition each participant had been randomized to (self-compassion = 80%, self-esteem = 90%, and control = 90%). When discrepancies occurred, raters came to a consensus by reviewing the prompts for each condition.
Main Analyses
Group differences in health-oriented weight management goals
A 3 (condition) × 2 (time) ANCOVA was conducted to determine mean differences in importance of health-oriented versus appearance-oriented weight management goals from pre- to post-treatment between participants in the body-focused self-compassion, body-focused self-esteem, and control conditions. There were non-significant main effects of group, F(2, 114) = 1.05, p = .352,
= .018, and time, F(1, 114) = 0.00, p = .962,
= .000, and a non-significant time by condition interaction, F(2, 114) = 3.55, p = .032,
= .059. As there were no significant main or interaction effects, post hoc analyses were not conducted.
Group differences in body appreciation
A 3 (condition) × 2 (time) ANCOVA was conducted to determine mean differences in body appreciation at each time point between participants in the body-focused self-compassion, body-focused self-esteem, and control conditions. There were non-significant main effects of condition, F(2, 121) = 2.25, p = .110,
= .036, and time, F(1, 121) = 0.01, p = .918,
= .000, and a non-significant time by condition interaction, F(2, 121) = 0.50, p = .606,
= .008. As there were no significant main or interaction effects, post hoc analyses were not conducted.
Group differences in healthy eating behaviors
A 3 (condition) × 2 (time) ANCOVA was conducted to determine mean differences in healthy eating behaviors at each time point between participants in the self-compassion, self-esteem, and control conditions. There were non-significant main effects of condition, F(2, 113) = 0.61, p = .547,
= .011, and time, F(1, 113) = 0.83, p = .365,
= .007, and a non-significant time by condition interaction, F(2, 113) = 4.13, p = .019,
= .068. As there were no significant main or interaction effects, post hoc analyses were not conducted.
Group differences in bulimic symptoms
A 3 (condition) × 2 (time) ANCOVA was conducted to determine mean differences in bulimic symptoms at each time point between participants in the body-focused self-compassion, body-focused self-esteem, and control conditions.
Figure 1 shows women’s bulimic symptoms from pre- to post-treatment.
Figure 1 shows that there were non-significant main effects of condition,
F(2, 119) = 3.68,
p = .028,
= .058, and time,
F(1, 119) = 4.96,
p = .028,
= .040, and a significant time by condition interaction,
F(2, 119) = 24.43,
p < .001,
= .291. Simple effect analyses demonstrated that women in the body-focused self-compassion condition significantly decreased in bulimic symptoms over time (
p < .001,
= .331), but women in the body-focused self-esteem condition (
p = .093,
= .024) and control condition showed no changes (
p = .876,
= .000).
Significant clinical changes
Of the 32 women who met clinical cut-offs on the EDI-1-BS at pre-treatment, 16 (50%) demonstrated a clinically significant change from pre- to post-treatment. A Fisher–Freeman–Halton exact test revealed that clinically meaningful change was associated with writing condition,
p = .019, Cramer’s
V = .520. Post hoc analyses revealed that clinically meaningful change was associated with being in the body-focused self-compassion condition (
p = .003,
n = 10; 83% changed), but not the body-focused self-esteem (
p = .144,
n = 4; 33% changed) or control conditions (
p = .100,
n = 2; 25% changed). See the number of participants who did and did not exhibit a clinically significant change in bulimic symptoms at post-test for each condition in
Figure 2.
Discussion
In the current study, we examined the efficacy of a brief body-focused self-compassion intervention compared to a body-focused self-esteem and control writing interventions on adult women’s body appreciation, valuation of health-oriented (vs. appearance) weight management goals, healthy (vs. unhealthy) eating behaviors, and bulimic symptoms. Our findings suggest that brief daily writing exercises tailored toward increasing self-compassion can be useful in reducing bulimic symptoms and increasing self-esteem and self-compassion in women in the short term. In particular, reductions in bulimic symptoms were found to be clinically significant, suggesting that increasing self-compassion may be more beneficial than increasing self-esteem because it reduces bulimic symptoms. Finally, this study demonstrated that for all outcomes of interest, the number of journal entries completed had no effects on observed changes over time or on the efficacy of conditions. This suggests that completing at least four exercises over a period of 7 days may be sufficient to provide favorable results for body-focused self-compassion writing interventions.
Congruent with hypotheses, women in the body-focused self-compassion condition increased in self-compassion (large effect) and decreased in bulimic symptoms over time (large effect) compared to the body-focused self-esteem and control conditions. Reductions in bulimic symptoms were clinically meaningful with the majority of women meeting clinical criteria allocated to this condition improving from pre- to post-treatment (83%; large association). Our findings are in line with other studies that show that brief self-compassion interventions can increase levels of self-compassion (
Leary et al., 2007;
Mosewich et al., 2013;
Seekis et al., 2017) and help reduce eating disorder symptoms (
Biber & Ellis, 2019;
Turk & Waller, 2020), including weekly binge eating (
Kelly & Carter, 2015). Clinically meaningful changes in other eating disorder indices, such as global symptom scores (
Kelly & Carter, 2015), body dissatisfaction, and drive for thinness (
Seekis et al., 2020), have also been demonstrated in other self-compassion interventions. Results from the manipulation check suggest that women in the body-focused self-compassion condition also increased in self-esteem (small effect) in addition to self-compassion; however, women in the body-focused self-esteem condition did not increase in self-compassion. Although this finding is exploratory and was not of primary interest to the study, these findings may demonstrate that being self-compassionate toward oneself when coping with momentary bouts of self-consciousness regarding one’s body, eating, or exercise behaviors may protect feelings of self-worth and competence (
Albertson et al., 2014) in addition to helping women regulate their emotions. Affect-regulation theories of binging and purging suggest that symptoms are used to suppress or distract from negative emotions (
Haynos & Fruzzetti, 2011). It is plausible that body-focused self-compassion interventions may reduce eating disorder symptoms by facilitating adaptive affect regulation (
Turk & Waller, 2020). Being self-compassionate may reduce the likelihood of perceiving negative experiences as overwhelming (i.e., mindfulness), in turn, reducing the use of symptoms (i.e., binge or purging) as coping behaviors to nullify or avoid unwanted feelings while bolstering a sense of competence to overcome and control unwanted behaviors.
Contrary to hypotheses, women in the body-focused self-compassion condition did not demonstrate changes in body appreciation, valuation of weight management goals, and healthy eating behaviors. Studies that have found moderate effects of self-compassion interventions on state body appreciation were longer in duration, compared the intervention to a waitlist control condition, and had a mindfulness component (
Albertson et al., 2014) or reflected on a hypothetical scenario (
Seekis et al., 2017). Self-compassion interventions that have integrated a mindfulness component affect body appreciation more robustly than waitlist (
Albertson et al., 2014) and active control groups (
Seekis et al., 2020). Furthermore, reflecting on a real compared to a hypothetical scenario may arouse more negative affect and body consciousness; therefore, a mindfulness component that focuses on mind-body connection may be essential to facilitate emotion regulation and positive body image. Women exposed to mindfulness audiotapes while undergoing a potential body self-conscious experience (i.e., trying on a bathing suit) demonstrated less negative affect and body dissatisfaction (
C. E. Adams et al., 2013). Taken together, our null findings could be due to these methodological differences and low statistical power (i.e., 70%). Furthermore, the magnitude of the relationship between self-compassion and health-promoting behaviors is small (
Sirois et al., 2015); therefore, higher statistical power may be required to detect significant changes in health-related outcomes as a result of engaging in self-compassionate writing.
Regarding the body-focused self-esteem writing condition, our hypotheses were partially supported. Congruent with hypotheses, women in the body-focused self-esteem condition increased in self-esteem, which is consistent with other studies employing similar writing interventions (
Leary et al., 2007;
Seekis et al., 2017); however, this is the first study, to our knowledge, to explore the effects of a body-focused self-esteem intervention on self-compassion. Another study employing a self-esteem writing intervention demonstrated that individuals in this condition exhibited fewer characteristics that are essential for self-compassion (
Leary et al., 2007), which may explain why increasing women’s self-esteem did not lead to increased self-compassion. For example,
Leary and colleagues (2007) found that those in the self-esteem condition were less likely to attribute failure as caused by their own actions, were more likely to believe that their experiences were unique, and were more likely to experience negative affect as a result of self-reflection. Our findings may suggest that high self-esteem is not a prerequisite for engagement in body-focused self-compassion and that the moderate correlation between these concepts may exist because self-compassion increases feelings of self-worth by treating oneself with kindness and acceptance when experiencing bouts of self-consciousness.
Contrary to our hypotheses, women in the body-focused self-esteem condition did not increase in body appreciation. It is plausible that self-esteem interventions may be more effective in improving affective components of body image that are more specific to appearance satisfaction. Self-esteem writing may encourage positive self-evaluations, which lead to more favorable social comparisons, reducing the discrepancy between actual and ideal body shape or size, thereby increasing appearance satisfaction (
Seekis et al., 2017). Body appreciation, however, is a form of positive body image that moves beyond one’s physical appearance and focuses on the acceptance of one’s imperfections in addition to caring and respecting one’s body (
Avalos et al., 2005).
Contrary to our hypotheses, women in the body-focused self-esteem condition also did not decrease in bulimic symptoms over time. Although previous studies found that self-esteem interventions are effective in reducing dieting behaviors and attitudes (
McVey & Davis, 2002;
O’Dea & Abraham, 2000), these changes may be limited to specific weight-controlling methods (e.g., dieting) rather than eating disorder behavior specifically (
McVey & Davis, 2002). Additionally, considering that women in the body-focused self-esteem condition did not increase in self-compassion, it is plausible that self-compassionate coping is essential to reduce binging and purging behaviors because self-compassion promotes better emotion regulation and, in turn, fosters self-regulatory resources (
Sirois et al., 2015).
Strengths, Limitations, and Future Directions
The current study has several strengths. This is the only study to our knowledge to examine the effectiveness of body-focused self-compassion and body-focused self-esteem writing interventions on multiple affective (e.g., self-compassion, self-esteem, body appreciation), cognitive (e.g., weight management goals), and behavioral (e.g., bulimic symptoms, healthy and unhealthy eating behaviors) outcomes that affect women’s body image, while controlling for important covariates identified in the literature (
Turk & Waller, 2020). There are also several methodological strengths. The recruitment strategy (i.e., community members as well as undergraduates) and diversity of the sample in terms of education and age enhances the generalizability of the findings, provides insight on the usefulness and suitability (i.e., interest, practicality, comprehensibility) of such interventions in the community, and examines the influence of such interventions in currently underrepresented samples (i.e., Canadian women of all ages). Additionally, the current study asked participants to reflect on real life situations that occurred earlier in the day, which is more ecologically valid than asking participants to reflect on a hypothetical scenario. Manipulation checks and evaluation of written responses to assess the validity of the interventions are additional methodological strengths. Other strengths include randomization, an active control, and examining clinically significant change.
Although our study has many strengths, it also has limitations. The sample was ethnically homogenous, such that the sample mostly comprised White women. Future studies should examine the effects of these interventions in a more diverse sample because body dissatisfaction varies by ethnic background (
Grabe & Hyde, 2006), and research examining the efficacy of self-compassion interventions in samples of ethnic minorities is lacking (
Turk & Waller, 2020). Furthermore, future research should examine the efficacy of these interventions in men because body dissatisfaction among men is prevalent (
Mellor et al., 2010). Additionally, the buffering effects of self-compassion have been shown to be gender dependent with research suggesting that self-compassion may have a larger effect on men’s mental health compared to women’s (
Bluth et al., 2017). Also, we used self-report measures, which may have resulted in socially desirable responding; however, research has found that some of the measures used in the current study are uncorrelated with social desirability scores (
Avalos et al., 2005;
Garner et al., 1983;
Neff, 2003a). Additionally, more objective measures of dietary intake should be used (e.g., Food Frequency Questionnaire;
Shim et al., 2014) as they are more accurate and appropriate to examine short-term change. Although participants could not presume our hypotheses from the description of the study, it is possible that participants interested in the study were women who frequently engage in self-reflective activities; therefore, it is plausible that women were using other techniques in addition to the prompts, leading to potential treatment contamination. Even though this study used an active control, women in this condition were not required to reflect on their bodies, eating, or physical activity habits, which is inconsistent with the body-focused self-compassion and body-focused self-esteem conditions, thereby functioning as a less effective control. Also, because the post-treatment survey was conducted 24 hours after the intervention, our findings must be interpreted with a degree of caution and better represent changes observed while women were engaging in tailored self-reflections rather than their true long-standing influences. Other limitations include low statistical power (i.e., ANCOVAs, 70%), that the non-disclosed entries were counted as incomplete, which introduces additional error into the data, and controlling for BMI and age. Controlling for BMI and age may have led to some scores, most notably healthy eating scores, to regress toward the mean.
Future research is required to examine the long-term effects of self-compassion interventions and the mechanisms by which these interventions promote positive body image and reduce eating disorder symptoms, including which components of self-compassion are most useful to promote these positive changes. For instance, a brief mindful self-compassion intervention with online Facebook discussions yielded changes in women’s body appreciation, drive for thinness, upward social comparisons, and social appearance anxiety for up to 3 months.
Seekis and colleagues (2020) speculated that some of these results could be due to the nature of their intervention, such that engaging in self-compassionate discussions about one’s appearance online with other women could have fostered higher levels of common humanity. This may have lowered their propensity to engage in social comparisons and reduced fear of negative evaluation, which facilitated unconditional acceptance and appreciation toward their own bodies (
Seekis et al., 2020). Future studies should strive to further examine which components of self-compassion are related to changes in specific outcomes to establish consistency and which modality of intervention is best suited to help facilitate improvements in each aspect. Furthermore, future research should strive to examine whether favorable changes in weight management goals and healthy eating behavior can be obtained by self-compassionate writing. The low statistical power in the current study may have affected our ability to detect significant changes in some outcome measures. Further research is required to determine whether, with sufficient power, changes in other outcome measures would be found.
Practice Implications
Our study suggests that body-focused self-compassion writing interventions can be useful in increasing protective factors for adult women’s body image and disordered eating, while reducing risk factors in the short term. As a preventive measure, children and youth should learn how to engage in self-compassion at an early age to develop a more stable positive self-attitude that is not contingent upon achievement, appearance, or social status. Policy makers, teachers, and parents can play a formative role in ensuring children and youth focus on harnessing self-compassion, which will protect them from fluctuations in their self-worth as they face many developmental (e.g., pubertal weight changes) and social (e.g., emphasis on popularity) changes. Our results also demonstrate that individuals have the capacity to follow self-guided writing activities to improve their own resilience when facing difficult or negative situations related to their bodies, eating behaviors, or physical activity. As these resources are widely accessible, cost effective, and can generate clinically meaningful change in eating disorder symptoms, clinicians and health care providers are encouraged to disseminate or facilitate these activities to improve well-being, psychological adjustment, and self-regulatory capacity in their clients. In addition, clinicians can provide brief psychoeducation on the components and utility of self-compassion and alter the focus of the writing activities to bolster self-compassion in different areas of their clients’ lives.