From competition to compassion: A caregiving approach to intervening with appearance comparisons

Body Image 25 (2018) 148–162

Contents lists available at ScienceDirect

Body Image journal homepage: www.elsevier.com/locate/bodyimage

From competition to compassion: A caregiving approach to intervening with appearance comparisons Kiruthiha Vimalakanthan ∗ , Allison C. Kelly, Sarina Trac Department of Psychology, University of Waterloo, 200 University Avenue West, Waterloo, Ontario, N2L 3G1, Canada

a r t i c l e

i n f o

Article history: Received 16 November 2017 Received in revised form 17 March 2018 Accepted 19 March 2018 Keywords: Appearance comparisons Compassion Body dissatisfaction Social mentalities theory Social comparisons

a b s t r a c t This study used a novel intervention grounded in social mentalities theory to compare the effects of cultivating a caregiving versus competitive mentality when intervening with appearance comparisons. For 48 hours, 120 female undergraduates were randomly assigned to use one of three strategies whenever they made unfavorable appearance comparisons: cultivating compassion and loving-kindness toward the comparison target (Caregiving); comparing themselves favorably to the target in non-appearance domains of superiority (Competition); or distracting themselves (Control). Although there was no main effect of condition, trait social comparison orientation interacted with condition to predict outcomes. Among women engaging more frequently in social comparison, the Caregiving condition was more effective than the Competition condition at reducing body dissatisfaction, restrained eating, and body, eating, and exercise-related comparisons. Findings suggest that cultivating a compassion-focused, caregiving mentality when threatened by appearance comparisons could be beneficial to women who engage more frequently in social comparison. © 2018 Elsevier Ltd. All rights reserved.

1. Introduction To compare is human: indeed, “keeping up with the Joneses” has never been easier, with the advent of modern technology and the concurrent development of mass and social media. However, while comparing ourselves to others can sometimes be beneficial in non-appearance domains (Buunk, Collins, Taylor, Van Yperen, & Dakof, 1990; Collins, 1996; Helgeson & Taylor, 1993; Wood, Taylor, & Lichtman, 1985), little evidence suggests this is true for the domain of physical appearance. Current research suggests that a higher frequency of appearance comparisons is strongly linked with higher levels of body dissatisfaction and disordered eating (Murnen & Smolak, 2015; Myers & Crowther, 2009; O’Brien et al., 2009). Women in particular are much more likely than men to make unfavorable appearance comparisons (Morrison, Kalin, & Morrison, 2004) and to feel distressed and body-dissatisfied afterwards (Strahan, Wilson, Cressman, & Buote, 2006). Furthermore, women frequently continue to make appearance comparisons even when the practice becomes harmful to their body image (Strahan et al., 2006). Perhaps most worryingly, unfavorable appearance comparisons are more frequent and emotionally damaging to those

∗ Corresponding author. E-mail addresses: [email protected] (K. Vimalakanthan), [email protected] (A.C. Kelly), [email protected] (S. Trac). https://doi.org/10.1016/j.bodyim.2018.03.003 1740-1445/© 2018 Elsevier Ltd. All rights reserved.

who are already high in body dissatisfaction and/or eating pathology (Groesz, Levine, & Murnen, 2002; Leahey, Crowther, & Ciesla, 2011; Leahey, Crowther, & Mickelson, 2007) and have been implicated in the maintenance of eating disorders (Blechert, Nickert, Caffier, & Tuschen-Caffier, 2009). In sum, unfavorable appearance comparisons seem to drive a cycle wherein women who are already unhappy with their bodies are made unhappier. How, then, do we intervene with these comparisons and alleviate the distress they cause? Social comparison theory might provide some ideas. This theory postulates that individuals frequently compare themselves to others to determine their standing in various life domains (Festinger, 1954). In downward or favorable comparisons, the target is perceived to be inferior in the comparison domain, whereas in upward or unfavorable comparisons, the target is perceived to be superior. Downward comparisons tend to have positive effects on affect and self-esteem (Gibbons, 1986; Hakmiller, 1966; Lemyre & Smith, 1985; Morse & Gergen, 1970), and can make one feel better when threatened by others’ perceived superiority (Wills, 1981) or when coping with distressing life events (Wood et al., 1985). Upward comparisons can provide useful information to guide self-evaluation in a given domain, and can serve as a source of motivation for self-improvement (Buunk et al., 1990; Collins, 1996; Helgeson & Taylor, 1993), but tend to have negative effects on affect and self-esteem (Morse & Gergen, 1970; Salovey & Rodin, 1984).

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

Research on appearance comparisons, however, is largely at odds with the wider literature on social comparison theory. For example, although Festinger (1954) argued that individuals will stop making upward comparisons if they become generally unfavorable or detrimental to their self-image, Strahan et al. (2006) found that women continue to make upward appearance comparisons even when their self-views are threatened. Their findings also suggest that women tend to make more comparisons to irrelevant targets (e.g., models). This is again out of step with the broader social comparison literature, where it is suggested that people avoid making comparisons with irrelevant others due to the dearth of diagnostic information (Wood, 1989). Taken together, current research suggests that the framework of social comparison theory falls short in explaining the phenomenology of appearance comparisons, highlighting a need for new perspectives. Empirical research on interventions for appearance comparisons further suggests a need for new theoretical perspectives. Social comparison theory might suggest that individuals who make frequent upward appearance comparisons should counteract the negative consequences of these comparisons by making more downward appearance comparisons. For example, women who typically compare themselves to more attractive others could focus on finding less attractive women with whom to compare themselves. They could also find other domains (e.g., intelligence, social skills) in which they are superior to their attractive comparison target. Indeed, these downward comparison strategies appear to be the most studied approach to date (e.g., Bailey & Ricciardelli, 2010; Lew, Mann, Myers, Taylor, & Bower, 2007). The research surrounding this downward comparison strategy is nevertheless mixed. Consistent with what social comparison theory might predict, Lew et al. (2007) found that highly body-dissatisfied female undergraduates who, after viewing pictures of fashion models, made downward comparisons with them on non-appearance domains, experienced improved body and weight satisfaction, decreased appearance anxiety, and decreased desires to lose weight. Another study by van den Berg and Thompson (2007) further suggested that participants who viewed images of downward comparison targets in the appearance domain showed increases in body satisfaction and positive mood. Other studies, however, challenge the idea that downward appearance comparisons are universally protective. Rancourt, Schaefer, Bosson, and Thompson (2016) found that although these comparisons were positively associated with body satisfaction for Asian and Caucasian women, for Hispanic/Latina women, downward comparisons were as detrimental to eating behaviors and attitudes as upward comparisons. Similarly, in a “dating game” experiment, women presented with a thin peer “competitor” experienced a reduction in body satisfaction and confidence, while those presented with an oversized peer experienced no compensatory effects on body satisfaction and confidence (Lin & Kulik, 2002). Further, Lin and Soby (2016) found that women who frequently made downward appearance comparisons tended to show an increased drive for thinness and greater dietary restraint. In general, the social comparison theory-inspired strategy of encouraging downward appearance comparisons appears to have mixed success in lessening body dissatisfaction and eating pathology (Fitzsimmons-Craft, 2017; Leahey et al., 2011). We further propose that even though downward comparisons may sometimes offer temporary emotional benefits, this practice is a competitively-motivated one that carries important costs for the comparer. Indeed, when making downward comparisons focused on one’s relative superiority in a given domain, whether appearance or other, individuals implicitly adopt a competitive orientation with those in their social environment. Such an orientation, when routinely adopted, can be harmful on physical, psychological, and social fronts. Consistent with this theory, Lin and Soby (2016) found

149

that those women who engaged in both upward and downward appearance comparisons endorsed more body image concerns and maladaptive behaviors relative to women who only made unidirectional comparisons (i.e., only upward or only downward). That is, the tendency to compare, no matter the direction, may be what is most toxic. This theory, together with current findings in the field point to a need for new theoretical viewpoints from which to approach the problem of damaging appearance comparisons. 1.1. Social mentalities theory: from competition to compassion Social mentalities theory (Gilbert, 1989, 2000) may offer a promising perspective through which to understand and intervene with the maladaptive tendency to engage in unfavorable appearance comparisons. This evolutionary theory postulates that individuals can adopt various mentalities, or mindsets, in their interactions with others, each of which drives the formation of relationships that indirectly promote survival and reproduction. Each of these mindsets serves to organize our minds such that certain patterns of attention, thinking, feeling, and behaving are triggered. In turn, these patterns allow for the enactment of various evolutionarily important relationships, including relationships based on care-giving/care-seeking, co-operation and reciprocity, dominance/submissiveness, and sex (Gilbert, 2000, 2005). From a particular social mentality, “the self” is construed in one way (e.g., caregiving, dominant) while “the other” is construed in another (e.g., care-seeking, subordinate). As a result, one’s approach and reaction to interactions with others depend upon the social mentality adopted at the time. For example, seeing another person suffer might be a pleasurable experience if one’s current mindset is oriented towards competing with them, but a distressing one if one’s current mindset is oriented towards caring for them (Gilbert, 2014). The mentality that is most relevant to social comparisons is the social ranking or competitive mentality (Gilbert, 2000). This mentality orients individuals towards appraising their rank or status relative to others, and to behaving accordingly. For example, an individual might dominate versus submit if s/he believes s/he is higher versus lower rank than a peer (Gilbert, 2000). When this competitive mindset is active, we see others as competitors for desired, limited resources (e.g., food, mates), and our primary cognitive concern is to compare ourselves to them to determine who is superior or inferior. This concern with social rank is elevated when individuals feel insecure in their social environments (Gilbert, McEwan, Bellew, Mills, & Gale, 2009)—as many bodydissatisfied women do (Pinto-Gouveia, Ferreira, & Duarte, 2014). Therefore, rather than intervening with upward comparisons by encouraging downward comparisons and propagating a competitive orientation, there may be people for whom it is more beneficial to shift to a different mindset altogether. Specifically, we expect that individuals who are highly prone to comparing themselves with others and likely to be over-reliant upon the competitive mentality may need to practice adopting a different mindset in their interactions with others – one that can reduce a sense of competitiveness and increase a sense of social connectedness. To that end, Gilbert’s (2005) conceptualization of a caregiving, compassionbased mindset may be a promising mentality for such individuals to work on adopting. In line with work by Bowlby, 1982, Gilbert (2005) suggests that humans are evolutionarily hard-wired to be caring and compassionate, arising from our motivation to take care of our offspring. When we adopt a caregiving mentality, we are oriented towards supporting and connecting with others rather than trying to establish our relative rank. We see others as fellow human beings who, like ourselves, experience suffering and desires to be happy (Gilbert, 2005, 2010). A large body of research supports the benefits of adopting a caregiving mentality. The Buddhist tradition of loving-kindness

150

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

meditation (Salzberg, 1995)—in which practitioners are guided to direct well-wishes towards a range of individuals—is rooted in loving-kindness, which Salzberg (2011, p. 178) describes as “a quality of the heart that realizes how connected we all are. . .a form of inclusiveness of caring.” Evidence suggests that this approach, which is likely made possible by the caregiving mentality, has yielded a plethora of physical and psychological benefits: reducing pain, negative affect, depressive symptoms, and psychological distress, while promoting physical health, positive affect, mindfulness, life satisfaction, resilience, and feelings of social connection, even towards strangers (Carson et al., 2005; Fredrickson, Cohn, Coffey, Pek, & Finkel, 2008; Hutcherson, Seppala, & Gross, 2008; Seppala, Hutcherson, Nguyen, Doty, & Gross, 2014). crucial output of the caregiving Another mindset—compassion—has also been shown to be beneficial to physical and mental well-being. Compassion may be defined as “an orientation of mind that recognizes pain and the universality of pain in human experience and the capacity to meet that pain with kindness, empathy, equanimity, and patience” (Feldman & Kuyken, 2011, p. 143). The benefits of cultivating compassion and/or providing compassion-motivated help for others include reduced stress and mortality risk, and increased psychological resilience and social connection with others (Brown, Nesse, Vinokur, & Smith, 2003; Cosley, McCoy, Saslow, & Epel, 2010; Hutcherson et al., 2008; Konrath, Fuhrel-Forbis, & Lou, 2012; Poulin, Brown, Dillard, & Smith, 2013; Seppala et al., 2014). Compassion is also potent due to its ability to foster self-compassion (Breines & Chen, 2013; Hermanto & Zuroff, 2016; Neff & Pommier, 2013)—the tendency to respond to personal distress with care and support for oneself (Gilbert, 2005; Neff, 2003), which is a key predictor of well-being and positive body image (e.g., Wasylkiw, MacKinnon, & MacLellan, 2012; Zessin, Dickhäuser, & Garbade, 2015). 1.2. Study objectives Although the advantages of cultivating a caregiving mentality toward others are well-documented, this approach has yet to be extended to the domain of physical appearance, body image, and eating. Furthermore, no studies to our knowledge have manipulated compassion with the aim of studying its effects on body image. The current study sought to address this gap and to investigate whether deliberate activation of the caregiving mentality would help to alleviate appearance comparisons and their negative correlates, especially among women who are higher in social comparison orientation. Social comparison orientation refers to the propensity to compare oneself to others (Gibbons & Buunk, 1999). Individuals high in social comparison orientation display a more interpersonal rather than introspective orientation; are more sensitive to the behavior of others; possess a degree of uncertainty about the self, with an interest in self-improvement by reducing this uncertainty; and engage in more social comparisons (Buunk & Gibbons, 2006; Gibbons & Buunk, 1999). Because these individuals are arguably accustomed to approaching their social interactions from a competitive mentality, we thought they would be most likely to benefit from cultivating care and compassion for others. Specifically, we thought that shifting to a caregiving mindset when making appearance comparisons – by evoking compassion and kind feelings towards comparison targets – would reduce the salience of the comparisons, inhibit competitiveness, and ultimately decrease the tendency to make appearance comparisons and the body dissatisfaction and disordered eating with which these comparisons are linked. To this end, we decided to study three brief interventions that entailed the practice of a particular strategy whenever women made an unfavorable appearance comparison. Downward comparisons are currently the most studied strategy for counteracting

unfavorable appearance comparisons and this approach maps on well to the competitive social mentality. We therefore decided to compare a caregiving intervention, in which participants were taught to cultivate feelings of compassion and loving-kindness towards comparison targets, to a “competition” intervention, in which women were taught to make downward comparisons to the target on non-appearance domains. This type of downward comparison strategy has had previous success in a study by Lew et al. (2007). They found that downward comparisons made on non-appearance domains to images of fashion models buffered the adverse effects of exposure to thin-ideal media among female undergraduates with high body dissatisfaction; this strategy also had a protective effect on these women’s body and weight satisfaction, appearance anxiety, and their desire to lose weight. We also included a control condition, in which women were taught to distract themselves with a counting task, allowing us to disentangle the influence of general cognitive engagement on outcomes. Given that the caregiving intervention was a novel intervention, our first aim was to assess its credibility and feasibility and to compare it to the extant downward comparison intervention as well as our distraction-based control condition. Our second aim was to investigate how women in the Caregiving condition would fare relative to those in the Competition and Control conditions in terms of their post-intervention: (a) levels of body dissatisfaction; (b) levels of restrained eating; and (c) frequency of social comparisons related to body appearance, eating, and exercise behavior. This latter variable would allow us to examine the impact of our interventions across the full range of comparisons associated with body and eating disturbances. Importantly, we reasoned that participants’ body mass index (BMI) might influence their post-intervention scores as well as their response to a particular intervention. This is because BMI is consistently identified in the literature as a predictor of body dissatisfaction and other eating disorder symptomatology (e.g., Paxton, Eisenberg, & Neumark-Sztainer, 2006; Snoek, van Strien, Janssens, & Engels, 2008). We therefore sought to control for BMI when examining the effects of condition on outcomes, and also sought to examine its interaction with condition on an exploratory basis. Our third and primary aim was to investigate whether women’s social comparison orientation would moderate the effects of our interventions on outcomes. We hypothesized that it would be those higher in social comparison orientation, and arguably most entrenched in the competitive mentality, who would benefit most from the novel approach of cultivating a caregiving mentality toward comparison targets.

2. Method 2.1. Overview of the procedure This multi-part study included two online sessions, an in-lab session, and a 48-h intervention (the “contractual period”). In the first session (15 minutes), which took place online, participants completed self-report trait-level measures. In the in-lab session (1 h), which took place at least 24 h after the first session, participants completed pre-intervention questionnaires and were introduced to their randomly assigned intervention (the “self-help strategy”). The key intervention period of the study was the 48h period following the in-lab session, during which participants practiced the intervention they had learned. While this portion of the study was inspired by a 24-h contract protocol that has been used successfully in previous experimental research (e.g., Boone, Soenens, Vansteenkiste, & Braet, 2012), we lengthened the period to 48 h in order to allow our participants enough time to practice and experience gains from their assigned intervention. While

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

relatively short in duration, we were concerned that extending the time period beyond 48 h would jeopardize compliance. Finally, participants completed post-intervention measures in a final online session (15 min) at the end of the 48-h intervention period. 2.2. Participants A sample of women were recruited via an online university research participant pool (44%) and campus- and communitywide advertisements (56%). While both men and women make appearance-related comparisons, we focused on women because they make more upward comparisons in this domain; these upward comparisons are also associated with greater body image dysfunction among women than men (Strahan et al., 2006; Thompson & Heinberg, 1993); and an all-female sample would make it easier to make links to extant research in the field. In the interest of achieving external validity, recruitment materials advertised the study as one for “Self-Help Strategies for Body Dissatisfaction” and were explicit that the research would be testing the effects of interventions designed to counteract appearance comparisons and their harms. To that end, the description also stated that potential participants might not find themselves suitable for the study if they did not regularly make physical appearance comparisons (i.e., at least a few times a day). This caveat was not a formal exclusion criterion, given that some women may not entirely be aware of the frequency of their comparisons. Participants were compensated with either a combination of 1.5 bonus participation marks allocated towards psychology courses plus $5, or else a total of $20. Of the 187 participants who signed up for the study, 136 completed the required pre-lab measures, and scheduled and attended the lab session. Of these 136 participants, seven failed to complete post-intervention measures: this included 3 out of 46 participants from the Caregiving condition, 2 out of 47 from the Competition condition, and 2 out of 43 from the Control condition. Attrition rates were not significantly different between conditions, X2 = 0.28, p = .87. These dropouts were excluded from analyses. We also excluded four participants who failed more than one of the attention-check questions that had been inserted in our questionnaires: this included one participant from the Caregiving condition, two from the Competition condition, and one participant from the Control condition. Finally, five more participants were excluded due to outlying values as modelled by Tukey boxplots on the variables of social comparison orientation and body mass index: this included four participants from the Competition condition, and one participant from the Control condition. The final sample consisted of 120 females (42 participants in the Caregiving condition, and 39 participants in each of the Competition and Control conditions) with a mean age of 20.7 (SD = 2.24), and mean body mass index (BMI; kg/m2 ) of 21.7 (SD = 3.10). Ethnic composition was as follows: 28.4% East Asian, 27.5% White/Caucasian, 16.7% West Indian/Caribbean, 10.8% South Asian, 6.9% Other, 4.9% Southeast Asian, and 2.9% Black/African; 2.0% of our participants declined to answer this question. 2.3. Measures 2.3.1. Moderator variable Participants completed the following measure as part of a battery of trait questionnaires provided in the first on-line session. 2.3.1.1. Social comparison orientation. This construct was assessed online prior to the lab session using the Scale for Social Comparison Orientation (INCOM; Gibbons & Buunk, 1999). This 11-item self-report questionnaire measures how often one compares oneself with others. Sample items include “I often compare how I

151

am doing socially (e.g., social skills, popularity) with other people” and “If I want to find out how well I have done something, I compare what I have done with how others have done.” Items were rated on a 5-point Likert scale ranging from 1 (I disagree strongly) to 5 (I agree strongly). After two of the items were reversecoded, the item responses were averaged to yield a composite score of trait social comparison orientation. Scores on the INCOM have acceptable test-retest reliability, demonstrating reasonable temporal stability ranging from .60 (for 1 year in American samples) to .72 (for 7.5 months in Dutch samples) in its initial validation study (Gibbons & Buunk, 1999). Cronbach’s alpha in the original sample was .83 (Gibbons & Buunk, 1999); in our own sample, the alpha was .79, indicating acceptable internal consistency. The measure also demonstrates good construct validity and discriminant validity, and a number of studies have evidenced that individuals higher in social comparison orientation as measured by this scale engaged in more social comparison than those lower on the scale (Gibbons & Buunk, 1999). The mean social comparison orientation score in our sample was 3.90 (SD = 0.46), which falls within 1 SD of the mean of 3.60 (SD = 0.58) reported in a validation study of the INCOM using a large sample of Americans (Gibbons & Buunk, 1999). 2.3.2. Dependent variables Participants completed the following measures on two different occasions: (a) as part of pre-intervention measures during the in-lab session, immediately before being taught their randomly assigned intervention; and (b) as part of post-intervention measures, during the online session after the 48-h intervention period had ended. In both instances, instructions on these measures were amended such that participants were asked to report only on the preceding 48-h period. 2.3.2.1. Body dissatisfaction. This construct was assessed using the Body Shape Questionnaire – 16-item B version (BSQ-16B; Evans & Dolan, 1993). Sample items from this self-report questionnaire include “Have you felt excessively large and rounded?” and “Have you been afraid that you might become fat (or fatter)?” Items were rated on a 6-point Likert-type scale ranging from 1 (Never) to 6 (Always) and responses were summed to yield a composite score of body dissatisfaction. Cronbach’s alpha in the original validation sample was .95 (Evans & Dolan, 1993), and in our own sample was .94, indicating excellent internal consistency. Scores on the BSQ–16 B also demonstrate good convergent and discriminant validity, as reported in a validation study of the BSQ–16 B (Evans & Dolan, 1993) using a sample of British women attending a family planning and well woman clinic. The measure is equivalent to the full-length BSQ in showing convergent relationships with measures of eating disorder symptomatology, BMI, and self-reported weight category; and divergent relationships with measures of anxiety, depression, age, and parity (Evans & Dolan, 1993). Our sample’s mean body dissatisfaction score pre-intervention was 57.21 (SD = 17.92) out of a maximum of 96, which is just over 1 SD of the mean of 40.0 (SD = 16.35) reported in the aforementioned validation study. Our sample’s higher mean may be attributed to the fact that the study was advertised as investigating self-help strategies for body dissatisfaction, which likely attracted women with higher body dissatisfaction. 2.3.2.2. Restrained eating. This construct was assessed using the Dietary Restraint subscale of the Eating Disorder Examination Questionnaire (EDE-Q; Fairburn, 2008), a 28-item self-report measure of eating disorder symptomatology. The five items in this subscale (e.g., “Have you had a definite desire to have an empty stomach with the aim of influencing your shape or weight?”) were rated on a 7-point Likert-type scale ranging from 0 (Not at all) to 6 (Markedly) and averaged. Scores on this subscale have excellent

152

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

2-week test-retest reliability (Luce & Crowther, 1999), and Cronbach’s alpha in the original validation sample of American college women ranged from .84 to .85; in our own sample, the alpha was .84, indicating good internal consistency. Scores on the subscale also demonstrate acceptable criterion validity and good concurrent validity when administered to a sample of Australian women (Mond, Hay, Rodgers, Owen, & Beumont, 2004)—that is, agreement with scores on its corresponding subscale on the interview-based Eating Disorders Examination (r = .71, p < .001). The mean dietary restraint score at baseline was 2.62 (SD = 1.69) out of a maximum of 6, which falls just over 1 SD of the mean of 1.25 (SD = 1.32) reported in a community sample of women (Fairburn & Beglin, 1994), and is also within 1 SD of the mean of 3.61 (SD = 1.82) reported in a clinical sample of women with eating disorders (Smith et al., 2017). 2.3.2.3. Eating disorder-related comparison orientation. This construct was assessed using the Body, Eating, and Exercise Comparison Orientation Measure (BEECOM; Fitzsimmons-Craft, Bardone-Cone, & Harney, 2012). This 18-item self-report questionnaire measures the extent to which individuals compare their body, their eating habits, and their exercise habits to others on three respective subscales. Items were rated on a 7-point Likerttype scale ranging from 1 (Never) to 7 (Always). Item responses were summed to yield each of the subscale scores, which were then summed to yield a composite score of eating disorder-related comparison orientation, which was the focus of the current study. Scores on the BEECOM have high 2-week test-retest reliability (Fitzsimmons-Craft et al., 2012), and Cronbach’s alpha in the original validation sample of American college women for the total scale score was .97; in our own sample, the alpha was .95, indicating excellent internal consistency. Scores on this measure also demonstrate good construct validity, as evidenced by the scale’s significant positive correlations with measures of social comparison orientation, eating disorder symptomatology, and body dissatisfaction in its validation sample (Fitzsimmons-Craft et al., 2012). Pre-intervention, the mean eating disorder-related comparison orientation score in our sample was 82.13 (SD = 23.16) out of a maximum of 126, which falls within 1 SD of the mean of 67.68 (SD = 23.84) obtained in the validation sample of young college women (Fitzsimmons-Craft et al., 2012). Nevertheless, the BEECOM mean in the present sample was significantly higher, t(558) = 5.80, p < .001, effect size r = .25. 2.3.3. Intervention credibility This construct was assessed using a modified version of the Credibility/Expectancy Questionnaire (CEQ; Devilly & Borkovec, 2000). This 6-item self-report questionnaire measures individuals’ expectancies about a given intervention and their beliefs in its credibility. Participants completed this measure immediately after learning their assigned intervention. For the purposes of this study, the wording of items was altered in two cases: instead of “therapy” or “treatment,” the word “strategy” was used; and instead of “trauma symptoms,” the phrase “body dissatisfaction” was used. Four items were rated on a 9-point Likert-type scale ranging from 1 (Not at all logical/Not at all useful/Not at all confident/Not at all) to 9 (Very logical/Very useful/Very confident/Very much), while two items were rated on a 0-100% scale in 10% increments. To facilitate scoring, all items initially rated on a 9-point scale were converted to percentages. Credibility was then calculated by taking the mean of the first three items (e.g., “At this point, how logical does the strategy offered to you seem?”). Expectancy was calculated by taking the mean of the last three items (e.g., If you were to practice this strategy for the next month, how much improvement in your body dissatisfaction do you think will occur?”). The maximum possible rating on each of these variables was therefore 100%. Scores on the CEQ have demonstrated good test-retest reliability when admin-

istered to an Australian sample of patients in a veteran residential program (Devilly & Borkovec, 2000), and Cronbach’s alpha in our sample for each of these subscales was .85 and .89 respectively, indicating excellent internal consistency. 2.3.4. Intervention compliance This construct was assessed with two different questions while collecting data on participants’ experiences with their assigned intervention. These questions were included in the postintervention battery of questionnaires completed by participants online after the 48-hour contract period. For the first item (“On what percentage of the comparisons that you made did you implement this strategy as instructed?”), participants provided a rating from 0–100%. For the second item, (“When you implemented this strategy as instructed, how much effort did you put into implementing this strategy?”) participants provided a rating on a 5-point Likert-type scale ranging from 0 (Not at all) to 4 (Extremely), which was then converted to a percentage to maintain consistency. 2.4. Procedure Upon recruitment, participants were directed to a link to a Qualtrics survey. The Informed-Consent Letter was presented and consent obtained at this time. Next, participants provided demographic information and their weight and height, and completed the INCOM. Participants were then scheduled for an in-lab session with a researcher, which generally took place anywhere from 1 to 8 days after completion of online questionnaires. All in-lab procedures were completed via Qualtrics on a desktop computer in a private room. First, participants completed baseline measures of the dependent variables described above. Then, they listened to a series of audio clips with text accompaniment during which they were introduced to the concepts of body dissatisfaction and appearance comparisons. Participants learned about the prevalence of body dissatisfaction among young women in modern society and the frequency of media messages relaying appearance ideals. They also learned that making comparisons with other women was a common occurrence, especially when feeling preoccupied or dissatisfied with one’s body. A broad range of potential targets (e.g., family members, friends, acquaintances, strangers) and settings (e.g., on social media, during lectures, on public transit, while spending time with friends) for comparison were mentioned, and an example of an upward appearance comparison was provided: “One particular example might be that you checked Facebook last night and saw your high school classmate’s vacation pictures and noticed how much thinner she is now than you are.” Next, participants learned that the goal of the in-lab session was to help them rehearse the “self-help strategy” that they would ultimately practice during their 48-h “contractual period.” First, in order to simulate the experience of making a comparison in the real world, participants were asked to vividly recall a recent distressing appearance comparison they had made to another woman: “Now please think back to a recent time in which you started to compare yourself to another woman in terms of appearance – a real-life comparison that made you really feel inadequate and dissatisfied with your appearance or body. Really bring this comparison to mind focusing on the various aspects of the other person’s appearance that you thought were superior to yours.” 2.4.1. Experimental manipulation Following this recall, participants were introduced to and led through their randomly assigned “self-help strategy.” They learned

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

that they would be asked to employ this particular strategy in response to any appearance comparisons made over the subsequent 48-h “contract period.” The overall structure for strategy rehearsal across conditions was the same, involving the following components: an introduction discussing the strategy’s benefits, an interactive portion leading participants through the process of applying the strategy in response to their recently-recalled appearance comparison, and a brief conclusion addressing anticipated reticence to the specific strategy and encouraging participants to practice the strategy. 2.4.1.1. Caregiving condition. The rationale presented to participants assigned to practice this self-help strategy normalized the competitive mindset that individuals adopt when they are focused on making comparisons, and then discussed its disadvantages, such as feeling self-focused and disconnected from others. Participants were then introduced to the idea that they also have a compassionfocused mindset at their disposal, which is what is active when they care for others. Deliberately shifting to this compassionate mindset in response to appearance comparisons was presented as a self-help strategy that could foster feelings of happiness and social connectedness. The presentation and wording of this selfhelp strategy was adapted from Gilbert’s (2010) previous work in compassion and social mentalities, and from loving-kindness meditation (Salzberg, 1995). In this condition, participants were asked to shift away from seeing their appearance comparison target as a competitor and instead towards seeing her as a fellow human being, and then to generate caring thoughts and feelings towards her: “In this compassionate mindset, shift away from seeing this person as a competitor, or someone who looks better than you, but instead focus on the fact that you are both human beings, and try to generate caring thoughts and feelings towards them. Really get in touch with the part of yourself that wants other people to be free from suffering and happy, and send these well-wishes to this person.”

2.4.1.2. Competition condition. The rationale presented to participants assigned to practice this self-help strategy suggested that they could minimize any sense of inadequacy from their recalled appearance comparison by thinking of the various ways in which they might be superior to their comparison target. Participants were presented and familiarized with the self-help strategy of generating qualities, skills, or accomplishments that they possess to a greater degree than their comparison target. This strategy’s presentation and wording were based on previous research in which downward comparisons in non-appearance domains were employed to intervene with the adverse effects of upward appearance comparisons to thin-ideal representations in the media (Lew et al., 2007). In this condition, participants were asked to focus on domains outside of appearance in which they feel particularly superior to their target: “. . .we want you to identify various other domains outside of appearance (e.g., intelligence, work ethic, athletic accomplishments, academic or career accomplishments, quality of life, social relationships, etc.) in which you are better than this person. . . . Really get in touch with the part of yourself that knows you are better than other people in certain ways – the part of you that is proud of your talents and successes.” In order to focus on getting into this mindset, participants were asked to recall aspects of themselves, their life, and their achievements of which they feel proud, or which they value: “To help you get into this mindset, you might try to recall a time when you were more successful than others; for example, a time when you got a highly desired co-op job or got a higher mark on a test than your friends. Or simply think of things about yourself and your life that you value and pride yourself on, for example your ability to form deep friendships.” They were led to bring these intentions and feelings of competition to mind and then redirect them to the target of their recalled comparison: “Now tell yourself the various ways in which you might be better than this person. For example, ‘I think I’m better than her at forming lasting friendships’ or “I have gotten better co-op jobs than her” or “People think I’m more genuine than her.” Try to think of personally-relevant comparisons you can make with this person where you believe you are better.”

To facilitate the adoption of this mindset, participants were asked to recall a time when they felt compassionate towards another person or animal. They were led to bring these intentions and feelings of compassion to mind and then to redirect them to the target of their recalled appearance comparison: “Imagine yourself expanding as if you are becoming calmer, wiser, stronger, and more mature. . .really able to care for or help that person. Pay attention to your body as you remember your feelings of kindness. Create a compassionate facial expression. Spend a moment with any expansion and warmth in your body. Note a real genuine desire for this other person to be free of suffering and to flourish. Now bring to mind this person you were recently comparing yourself to, while staying in touch with your compassionate feelings and intentions. Keep these alive and direct these feelings of compassion toward this person. With this person in mind, and these compassionate intentions within you, imagine saying: ‘May you be well,’ “May you be happy,” and “May you be free from suffering.”

153

2.4.1.3. Control condition. The rationale presented to participants assigned to practice this self-help strategy introduced them to the approach of using mental strategies to distract oneself from continuing to make comparisons, and the benefits of distraction. Participants were presented and familiarized with the self-help strategy of using mental distractions as “short-term time-outs” to interrupt negative and/or stressful states of mind. In this condition, participants were asked to count backwards in threes from 50, prioritizing accuracy and even pace:

Table 1 Correlations (and 95% Confidence Intervals) between Moderator and Dependent Variables at Baseline. Variables

Social comparison orientation

Body dissatisfaction

Eating disorder-related comparison orientation

Social comparison orientation Body dissatisfaction Eating disorder-related comparison orientation Restrained eating Body mass index

– .11 (−.07–.28) .31** (.14–.46) .08 (−.10–.26) .02 (−.16–.20)

– .66** (.54–.75) .61** (.48–.71) .22* (.04–.39)

– .53** (.39–.65) .09 (−.09–.27)

* **

p < .05. p < .001.

Restrained eating

Body mass index

– .01 (−.17–.19)

–

54.5 (18.0) 2.1 (1.5) 78.9 (26.3) 49.9 (18.6) 2.3 (1.9) 69.7 (24.1) 60.5 (17.2) 3.0 (1.8) 85.0 (19.2) Note. Sample sizes per condition: Caregiving = 42, Competition = 39, Control = 39.

49.9 (18.9) 2.1 (1.8) 68.9 (25.5)

Pre-Intervention Means (SDs) Post-Intervention Means (SDs)

Competition Caregiving

Conditions

56.7 (18.4) 2.7 (1.7) 82.4 (23.5) 16–96 0–6 18–126

Table 1 presents Pearson zero-order correlation coefficients between all study variables at Time 1. Pearson correlations indicated that body dissatisfaction had strong positive correlations with both eating disorder-related (i.e., body, eating, and exercise) comparison orientation and restrained eating, as well as a modest positive correlation with BMI. Eating disorder-related comparison orientation and restrained eating shared a strong positive correlation. Finally, social comparison orientation and eating disorder-related comparison orientation shared a modest though significant positive correlation. However, social comparison orientation was not significantly correlated with our other variables. Table 2 presents means and standard deviations for all study variables, by condition, both pre- and post-intervention. We first conducted t-tests to ensure that our final sample did not differ from participants who had dropped out on baseline levels of our dependent variables or credibility and expectancy; no significant differences were found between the two groups, ps = .15–.95. We conducted a MANOVA to examine whether our random assignment

Scale Ranges

3.1. Preliminary analyses

Variables

3. Results

Table 2 Dependent Variable Means and Standard Deviations across Conditions Pre- and Post-Intervention.

In order to further facilitate study compliance, participants were also given a few minutes after signing the contract to generate examples of opportunities (in the next 48 hours) for them to employ the strategy, along with the specifics of how they might do so (Gollwitzer, 1999). A few examples were presented on the contract as prompts. This written activity was completed privately on the participant’s copy of the contract and was not seen by the researcher. At the end of each participant’s 48-hour “contractual period,” they received a link to a battery of post-intervention questionnaires consisting of the dependent variables listed above.

Post-Intervention Means (SDs)

“This is a contract made between (participant) and (researcher) on (date). I, (participant) confirm that in the next 48 hours, as soon as I find myself comparing my body to another woman’s body, I will commit to [immediately shifting my mindset to a compassionate one/immediately shifting my mindset to a competitive one that focuses on my superior qualities/performing my counting distraction task]. I understand what I am required to do and agree to do this consistently throughout the next 48 hours. For example, if I encounter a person whom I believe to be thinner or more attractive than I am, I will shift from trying to figure out ways in which they look better than me, and instead focus on [developing caring, compassionate feelings toward them, and wish them happiness and strength in whatever struggles they are going through/how I am more intelligent, athletic, or hard-working/counting backwards in threes from 50, prioritizing my accuracy and even pace].”

Body dissatisfaction Restrained eating Eating disorder-related comparison orientation

Control

2.4.2. Contract signing As the final task of the in-lab session, participants were asked to sign a written contract with the researcher committing them to employ their assigned self-help strategy for the next 48 h whenever they made an upward appearance comparison to another woman. This contract protocol has been used successfully in previous experimental research (e.g., Boone et al., 2012). Two copies of the contract were signed to increase credibility and compliance: one was retained by the researcher and one was retained by the participant. An excerpt is presented below:

Pre-Intervention Means (SDs)

Post-Intervention Means (SDs)

“In this distraction task, we would like you to focus on counting backwards in threes from 50 and then continuing on with your day.. . .The goal here is to prioritize getting the numbers right and keeping an even pace.”

47.1 (20.5) 1.9 (1.5) 67.2 (27.3)

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

Pre-Intervention Means (SDs)

154

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

155

Table 3 Credibility, Feasibility, and Compliance across Conditions. Condition Means (and Standard Deviations)

Significant Contrasts

Caregiving

Competition

Control

Credibility

63.2% (17.6%)

60.0% (15.3%)

51.4% (18.4%)

Expectancy Proportion of Strategy Use Effort

42.8% (16.7%) 79.5% (25.4%) 72.8% (16.4%)

48.3% (16.3%) 69.7% (31.6%) 75.4% (12.6%)

39.18.5% (18.0%) 78.8% (26.3%) 75.4% (13.4%)

Caregiving > Control Competition > Control Competition > Control N/A N/A

Note. Sample sizes per condition were: Caregiving = 42, Competition = 39, Control = 39. Scores for Credibility and Effort were converted to percentages for ease of comparison with the other measures. The overall effect of condition was significant in predicting credibility (p < .05), a trend in predicting expectancy, and not significant in predicting proportion of strategy use or effort.

was successful in preventing baseline differences across conditions and found an effect of condition on baseline levels of our three dependent variables, F(6, 232) = 2.20, p = .04, effect size r = .09. Three separate ANOVAs were run as follow-up to probe this finding. Condition did not predict baseline levels of body dissatisfaction or eating disorder-related comparison orientation, ps = .33–.51, but predicted baseline levels of restrained eating at a trend-level, F(2,117) = 2.90, p = .06. One significant contrast emerged: participants in the Control condition tended to have lower restrained eating scores at baseline than participants in the Competition condition, F(1, 117) = 5.74, p = .02. Nevertheless, the general absence of statistically significant relationships between condition and baseline scores (i.e., all ps > .05) supports the ANCOVA assumption of independence between the independent variable and covariate. Of note, we found support for all other ANCOVA assumptions, including the tests of homogeneity of regression. That is, the strength of the positive relationship between pre- and post-measures did not differ significantly by condition, ps = .07–.55.

approximately 40% if they committed to their assigned intervention for over a month (see Table 3). In addition, condition predicted expectancy at a trend level, F(2, 117) = 2.87, p = .06. Participants had higher expectations for the Competition condition than the Control condition, F(1, 117) = 5.66, p = .02. There were no differences in perceived expectancy between the Caregiving and Control conditions, F(1, 117) = 0.94, p = .33, or the Caregiving and Competition conditions, F(1, 117) = 2.11, p = .15.

3.2.3. Compliance and effort As seen in Table 3, overall levels of self-reported compliance were high, with participants across conditions reporting that they performed their learned strategy on average 69.7% to 79.5% of the time. There were no significant differences between the three conditions in compliance, F(2, 116) = 1.53, p = .22. Similarly, average levels of self-reported effort put into the strategy were high, ranging from 72.8% to 75.4%, and mean levels did not differ between the three conditions, F(2, 117) = 0.43, p = .65.

3.2. Credibility and feasibility of interventions Our first aim was to determine the acceptability and feasibility of our interventions. Specifically, we wished to determine whether the novel Caregiving intervention was as acceptable and convincing to participants as the downward-comparison Competition intervention. Means and standard deviations were calculated and examined for perceived credibility, expectancy, effort, and compliance. Between-condition differences were also examined through ANOVAs in which condition was the sole predictor of each variable. For analyses that yielded a significant effect of condition, 1-df contrasts were used to test three planned comparisons: (1) Caregiving vs. Competition condition; (2) Competition vs. Control condition; and (3) Caregiving vs. Control condition. 3.2.1. Credibility As reported in Table 3, mean credibility values, which ranged from 51.4% to 63.2%, suggested that participants across conditions seemed to find their intervention somewhat credible—that is, upon learning about the intervention, they found it relatively logical, believed to a moderate degree that it would be successful, and were somewhat confident in recommending the intervention to friends. There was nevertheless an effect of condition on Credibility, F(2, 117) = 5.08, p < .01. Contrasts revealed that participants found the control condition significantly less credible than both the Caregiving, F(1, 117) = 9.61, p < .01, and Competition, F(1, 117) = 4.93, p = .03, conditions, respectively. There was no difference in perceived credibility between the Caregiving and Competition conditions, F(1, 117) = 0.70, p = .40. 3.2.2. Expectancy Mean values generally suggested that participants across conditions believed their body dissatisfaction would improve by

3.3. Changes in outcome variables over time The second aim of the study was to determine the relative impact of the Caregiving condition on outcome variables. To do this, we ran ANCOVAs in which Time 2 levels of body dissatisfaction, restrained eating, and eating disorder-related comparison orientation were our dependent variables, and respective Time 1 levels of these variables were controlled. Of note, raw Time 1 scores were used to allow us to interpret the intercept as an indicator of mean change from pre to post which, although not of primary interest, would still allow us to determine whether participants’ scores on the dependent variables changed over time across conditions. For each dependent variable, three ANCOVAs were run. In the first ANCOVA, intervention condition (Caregiving, Competition, Control) was entered as the predictor of interest and BMI (standardized) was a covariate; here, we could also examine the test of the intercept to examine whether there were changes over time. In the second ANCOVA, we also examined the interaction between BMI and condition to explore the possibility that the efficacy of particular interventions might depend on participants’ BMI. Finally, the third ANCOVA additionally included trait social comparison orientation and its interaction with condition as predictors. This interaction term would allow us to test our primary hypothesis that trait social comparison orientation would moderate the impact of condition on outcomes. When there were significant interactions, we replaced raw social comparison orientation scores with standardized scores in the model and probed the interactions by graphing mean estimates of the relevant dependent variable at higher (1.5 SD above the mean) and lower (1.5 SD below the mean) levels of social comparison orientation within each condition. These estimates were then compared to one another using pairwise contrasts.

156

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

Table 4 F and R2 Values from Three ANCOVA Models Predicting Residual Changes in Body Dissatisfaction, Restrained Eating, and Eating Disorder-Related Comparison Orientation. ANCOVA 1

ANCOVA 2

ANCOVA 3

DV = Time 2 Body Dissatisfaction

Intercept Time 1 Body Dissatisfaction BMI Condition BMI × Condition SCO SCO × Condition Explained Variance R2

df

F

p

df

F

p

df

F

p

111

2.31 34.19 0.01 0.21 – – –

.13 <.001 .93 .81 – – –

109

0.88 34.16 0.01 0.21 0.96 – –

.35 <.001 .93 .81 .39 – –

106

0.02 37.74 0.01 0.40 0.89 0.80 6.81

.88 <.001 .92 .67 .37 .36 .002

– – –

– –

.24

.25

ANCOVA 1

.34

ANCOVA 2

ANCOVA 3

DV = Time 2 Restrained Eating

Intercept Time 1 Restrained Eating BMI Condition BMI × Condition SCO SCO × Condition Explained Variance R2

df

F

p

df

F

p

df

F

p

111

0.04 48.32 0.27 0.10 – – –

.84 <.001 .61 .91 – – –

109

0.37 49.15 0.27 0.10 1.95 – –

.54 <.001 .61 .91 .15 – –

106

0.67 50.58 0.28 0.10 1.81 0.00 3.29

.41 <.001 .60 .91 .17 .95 .04

– – –

– –

.31

.33

.37

ANCOVA 1

ANCOVA 2

ANCOVA 3

DV = Time 2 Eating Disorder-Related Comparison Orientation

Intercept Time 1 Eating Disorder-Related Comparison Orientation BMI Condition BMI × Condition SCO SCO × Condition Explained Variance R2

df

F

p

df

F

p

df

F

p

111

1.51 67.99 0.05 0.68 – – –

.22 <.001 .82 .51 – – –

109

1.28 67.60 0.05 0.68 0.68 – –

.26 <.001 .82 .51 .51 – –

106

0.28 71.50 0.06 0.72 0.72 1.42 3.93

.60 <.001 .81 .49 .49 .24 .02

– – – .38

– – .39

.44

Note. BMI = body mass index; SCO = social comparison orientation.

3.3.1. Body dissatisfaction As seen in Table 4, the test of the intercept in the first ANCOVA revealed that holding other variables constant, there was no significant change in body dissatisfaction across conditions from pre to post (Intercept = 18.22, SE = 11.97), t(111) = 1.52, p = .13. This ANCOVA additionally revealed that, controlling for BMI and Time 1 levels of body dissatisfaction, there was no effect of condition on Time 2 levels of body dissatisfaction. In other words, levels of body dissatisfaction did not change differentially over time as a function of condition. The second ANCOVA revealed that there was no effect of Condition × BMI, suggesting that the interventions did not perform differently based on participants’ BMI. Finally, the third ANCOVA revealed that trait social comparison orientation interacted with condition to predict body dissatisfaction at Time 2, F(2, 106) = 6.81, p = .002, effect size r = .25. In order to determine the nature of this interaction, Time 2 body dissatisfaction scores, controlling for Time 1 scores, were estimated and graphed within each condition for participants with lower (1.5 SD below the mean) and higher (1.5 SD above the mean) levels of trait social comparison orientation (see Fig. 1). Two significant contrasts emerged when comparing within conditions. In the Caregiving condition, individuals with higher trait social comparison orientation had significantly lower body dissatisfaction scores after the 48-hour contract period compared to those with lower trait social comparison orientation, t(106) = 2.39, p = .02,

effect size r = .15, whereas the reverse pattern emerged in the Competition condition, t(106) = −2.59, p = .01, effect size r = .15. When comparing across conditions, individuals with higher trait social comparison orientation in the Caregiving condition reported significantly lower body dissatisfaction after the contract period than their counterparts in the Competition condition, t(106) = −2.55, p = .01, effect size r = .15, and, at a trend-level, than those in the Control condition, t(106) = −1.75, p = .08, effect size r = .13. In contrast, individuals with lower trait social comparison orientation had significantly higher post-intervention body dissatisfaction in the Caregiving condition compared to those in the Competition, t(106) = 3.47, p < .001, effect size r = .18, and Control conditions, t(106) = 2.76, p < .01, effect size r = .16. 3.3.2. Restrained eating As seen in Table 4, the test of the intercept in the first ANCOVA revealed no significant change in restrained eating from pre to post while holding other variables constant (Intercept = 0.21, SE = 1.02), t(111) = 0.20, p = .84. In this model, which controlled for BMI and Time 1 levels of restrained eating, there was also no effect of condition on Time 2 levels of restrained eating. The second ANCOVA additionally revealed that there was no effect of Condition × BMI. Finally, the third ANCOVA revealed that trait social comparison orientation interacted with condition to predict restrained eating at Time 2, F(2, 106) = 3.29, p = .04, effect size r = .17.

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

157

Fig. 1. Trait social comparison orientation moderates the effect of condition on body dissatisfaction. Controlling for pre-intervention body dissatisfaction and body mass index, in the Caregiving condition, those with higher trait social comparison orientation reported lower post-intervention body dissatisfaction than those with lower trait social comparison orientation. Those with higher trait social comparison orientation in the Caregiving condition also reported lower post-intervention body dissatisfaction than their counterparts in the Competition condition. In contrast, those with lower trait social comparison orientation in the Caregiving condition had significantly higher body dissatisfaction scores than their counterparts in each of the other two conditions. Lower/higher social comparison orientation represents −1.5/ + 1.5 standard deviations from the mean. Note. SCO = social comparison orientation.

Fig. 2. Trait social comparison orientation moderates the effect of condition on restrained eating. Controlling for pre-intervention restrained eating and body mass index, in the Caregiving condition, those with higher trait social comparison orientation reported lower post-intervention restrained eating than those with lower trait social comparison orientation. Those with higher trait social comparison orientation in the Caregiving condition also reported lower post-intervention restrained eating than their counterparts in the Competition condition. In contrast, those with lower trait social comparison orientation in the Caregiving condition had significantly higher restrained eating scores than their counterparts in the Competition condition. Lower/higher social comparison orientation represents −1.5/ + 1.5 standard deviations from the mean. Note. SCO = social comparison orientation.

Fig. 2 depicts Time 2 restrained eating estimates within each condition for participants with lower and higher levels of trait social comparison orientation. The only significant contrast within conditions was in the Caregiving condition, where individuals with higher trait social comparison orientation had significantly lower restrained eating scores after the contract period compared to those with lower trait social comparison orientation, t(106) = 2.08, p = .04, effect size r = .14. When comparing across conditions, individuals with higher trait social comparison orientation in the Caregiving condition had significantly lower restrained eating scores after the contract period compared to their counterparts in the Competition condition, t(106) = −2.06, p = .04, effect size r = .14. In contrast, individuals with lower trait social comparison orientation in the Caregiving condition tended to have higher restrained eating scores

after the contract period than their counterparts in the Competition condition, t(106) = 1.97, p = .05, effect size r = .14. 3.3.3. Eating disorder-related comparison orientation1 As seen in Table 4, the test of the intercept in the first ANCOVA revealed that holding other variables constant, eating disorder-

1 Findings for this dependent variable are presented using the full scale; the same analyses conducted on individual comparison orientation subscales—that is, Body, Eating, and Exercise—yielded the same pattern of results. In the interest of brevity, and bolstered by indications of high internal consistency both in the original validation sample and our own sample, we decided to present our findings here using the full scale of the measure.

158

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

Fig. 3. Trait social comparison orientation moderates the effect of condition on eating disorder-related comparison orientation. Controlling for pre-intervention eating disorder-related comparison orientation and body mass index, those with higher trait social comparison orientation in the Caregiving condition reported lower postintervention eating disorder-related comparison orientation than their counterparts in each of the other two conditions. Lower/higher social comparison orientation represents −1.5/ + 1.5 standard deviations from the mean. Note. SCO = social comparison orientation.

related comparison orientation did not change significantly from pre to post (Intercept = 17.94, SE = 14.58), t(111) = 1.23, p = .22. In addition, there was no effect of condition on Time 2 levels of eating disorder-related comparison orientation controlling for BMI and Time 1 levels. There was also no effect of Condition × BMI in the second ANCOVA. Finally, the third ANCOVA revealed that trait social comparison orientation interacted with condition to predict eating disorder-related comparison orientation at Time 2, F(2, 106) = 3.93, p = .02, effect size r = .19. Fig. 3 shows estimated Time 2 levels of eating disorder-related comparison orientation for participants with lower and higher levels of trait social comparison orientation. There were no significant contrasts when comparing within conditions; however, there was a trend for those with lower social comparison orientation in the Competition condition to have lower post-intervention body dissatisfaction than those with higher social comparison orientation, t(106) = −1.91, p = .06, effect size r = .13. Contrasts across conditions indicated that individuals with higher trait social comparison orientation in the Caregiving condition were significantly less oriented towards making eating disorder-related comparisons after the contract period compared to their counterparts in the Competition, t(106) = −2.56, p = .01, effect size r = .15, and Control conditions, t(106) = −2.27, p = .03, effect size r = .14. 3.4. Summary of results Our first aim in this study was to assess the credibility and feasibility of the Caregiving intervention. Participants reported moderate credibility and expectancy towards this intervention as well as high compliance; furthermore, these ratings did not differ significantly from the downward comparison intervention, supporting the relative acceptability and feasibility of the Caregiving intervention. Our second aim was to investigate the effect of condition on body dissatisfaction; restrained eating; and frequency of eating disorder-related comparisons. Controlling for Time 1 scores and BMI, there was no main effect of condition on Time 2 scores, nor was there a significant BMI by condition interaction. Our third and primary aim was to investigate whether social comparison orientation would moderate the effects of our interventions on outcomes. As hypothesized, social comparison orientation inter-

acted with condition to predict Time 2 levels of body dissatisfaction, restrained eating, and frequency of eating disorder-related comparisons. Specifically, the Caregiving intervention was most beneficial for women with higher levels of social comparison orientation, while the Competition intervention was most beneficial for women with lower levels of social comparison orientation.

4. Discussion Inspired by social mentalities theory, this study assessed the acceptability, feasibility, and impact of a novel compassion-based intervention designed to reduce appearance-based social comparisons and their associated harms. To explore these aims, we compared the effects of three different strategies that women were taught to use after each unfavorable appearance comparison they made during a 48-h period: shifting to a caregiving mentality centered on engendering feelings of compassion and well-wishes towards the comparison target; making downward comparisons to the same target in non-appearance domains thought to evoke a competitive mentality; and as a control condition, performing a distraction task. Findings supported the relative acceptability and feasibility of the caregiving intervention. Although there was no effect of condition on comparisons, body dissatisfaction, and restrained eating, the Caregiving intervention was more beneficial for those higher in social comparison orientation—consistent with our hypothesis—whereas the Competition intervention was more effective for those lower in social comparison orientation. Upon learning about their intervention, participants found the Caregiving intervention to be moderately credible, as credible as the Competition intervention, and significantly more so than the control condition. Across conditions, participants had moderate expectations about their intervention immediately after learning about it, believing that their body dissatisfaction would improve by approximately 40% if they committed to their assigned strategy for over a month. Overall levels of self-reported compliance and effort were high across conditions. These findings suggest that, despite its relative novelty, the Caregiving intervention was as acceptable and feasible as other strategies that may be used to target appearance comparisons and their harms.

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

The intervention condition to which participants were assigned did not influence the change in body dissatisfaction, restrained eating, and eating disorder-related comparison orientation they experienced over the 48 hours. Furthermore, when holding all other variables constant (i.e., condition, pre-scores, and BMI), there was no evidence of change across conditions from pre to post. It is nevertheless worth noting in Table 2 that while our sample’s mean scores on dependent variable measures were closer to those of clinical samples than typical community samples at baseline, these scores were well-within 1 SD of community sample by the end of the intervention period (see Measures section for community and clinical norms). These numbers suggest that participants across conditions may have experienced clinically significant improvements over the 48-h study period even if these changes were not statistically significant. However, this is only a tentative observation and replication is warranted. 4.1. Social comparison orientation as a moderator Our central results indicated that as hypothesized, the relative efficacy of our Caregiving versus Competition interventions depended on the personality trait of social comparison orientation. For women with higher levels of social comparison orientation, the Caregiving intervention led to lower body dissatisfaction, restrained eating, and eating disorder-related comparison orientation over the intervention period as compared to the Competition intervention. However, for those with lower levels of this trait, the Competition intervention led to lower body dissatisfaction and restrained eating than the Caregiving intervention. It appears that for young women who were generally more focused on comparing themselves to others, responding to appearance comparisons by cultivating compassion and loving-kindness towards comparison targets was more beneficial than generating downward comparisons towards targets on non-appearance domains; however, the reverse was true for those who were generally less oriented toward comparing themselves to others. Individuals higher in social comparison orientation are more attuned to others in their environment—attending to them as sources of information from which to assess their own self-worth or performance in various domains; they also engage in more social comparisons (Buunk & Gibbons, 2006; Gibbons & Buunk, 1999), and are more negatively affected by the comparisons they make (Buunk & Gibbons, 2006; Vogel, Rose, Okdie, Eckles, & Franz, 2015). A complete shift from this comparison-oriented mindset to one of caregiving may have helped broaden these participants’ view of other females from ‘threats’ and ‘rivals’ or perpetuators of standards to achieve, to one that considers their shared humanity. With repeated practice, this mindset shift may have reduced the competitive instinct to make body, eating, and exercise-related comparisons, as well as the body dissatisfaction and restrained eating that are often tied to these comparisons. For individuals higher in social comparison orientation, making downward comparisons may have perpetuated their competitive mindset, which might explain why their eating disorder-related comparisons and body image concerns were less impacted by the competition condition. Future research should examine these potential mechanisms and others. Among women with lower social comparison orientation, who fared significantly better in the Competition condition than their counterparts in the Caregiving condition, the novelty of making (downward) comparisons may have played a salient role in helping them. It may also be that making comparisons is an inherently less harmful process for those who are lower in social comparison orientation (Buunk & Gibbons, 2006). Another potential explanation might be drawn from Wood’s (1989) theory that social comparisons can be made in the pursuit of self-evaluation, self-improvement,

159

and/or self-enhancement. Perhaps for women with lower social comparison orientation, appearance comparisons may be made more often in the service of self-enhancement—the motivation to maintain or increase self-esteem. As a result, the downward comparisons made in the Competition intervention may have been a more effective means of accomplishing this goal, as opposed to the compassion and well-wishes generated in the Caregiving intervention, which may only improve self-esteem indirectly, if at all (Dijkstra, Gibbons, & Buunk, 2011). Even though the Competition intervention seemed to benefit those low in social comparison orientation, one must be cautious about endorsing the use of a downward comparison strategy for these individuals. The long-term effects of encouraging a competitive mentality among those who are not naturally comparison-oriented could ultimately be harmful. While these downward comparisons may offer them temporary emotional benefits over a short time period, as in our study, the routine adoption of a competitive orientation can be harmful to one’s physical, psychological, and social well-being. (e.g., Leahey et al., 2007; Leahey et al., 2011; Lin & Soby, 2016; Myers & Crowther, 2009). Future research should continue to explore the advantages and disadvantages of this strategy over time. It may be especially intriguing to separate the components of this strategy – that is, the aspect that involves reminding oneself of one’s positive qualities and the aspect that involves viewing the other as inferior and oneself as superior. Perhaps performing the former part of the strategy in the absence of the latter would be a more advantageous, and less “dangerous” approach for individuals to use when they are making upward appearance comparisons. Interestingly, within the Caregiving condition, women with lower social comparison orientation did not experience as many benefits to body dissatisfaction and restrained eating as women with higher social comparison orientation experienced. As suggested above, perhaps cultivating caring, compassionate feelings does not satisfy the goal of restoring self-esteem in the way that downward comparisons do when these women make appearance comparisons. 4.2. Theoretical and practical implications The present study was the first to our knowledge to apply social mentalities theory to develop interventions for harmful appearance comparisons. Our findings suggest that this theory is a valuable lens through which we can understand and intervene with appearance comparisons. Of note, this study is the first to demonstrate that cultivating compassion for others may benefit one’s own body image and eating disorder-related behavior. That a relatively brief intervention manipulating an outward interpersonal orientation can have such a profound impact on complex internally-driven processes is remarkable. This impact is even more striking when considering the clinical significance of participants’ improvements. Such findings are an innovative addition to a fast-developing body of work establishing the benefits of compassion for others on one’s psychological (e.g., Cosley et al., 2010; Seppala et al., 2014) and physical well-being (e.g., Brown et al., 2003; Konrath et al., 2012) in both community and clinical samples. Moreover, although the interventions in this study only targeted appearance comparisons, the measure used to gauge appearance comparisons took into account comparisons based on body, eating, and exercise. It is thus rather notable that for a subset of our participants (those higher in social comparison orientation), the Caregiving intervention was found to be beneficial in reducing the frequency of a variety of comparisons implicated in eating pathology. Such promising results suggest that it may be worth examining whether the cultivation of compassion for others can effectively target other significant issues

160

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

in the body image and eating domain, such as fat talk, internalization of the thin ideal, binge/purge symptoms, or exercise dependence. Although replication is needed, from a practical perspective, findings from this study suggest that young women should be taught to cope with appearance comparisons and their associated harms differently, depending on their social comparison orientation. Clinicians and prevention or treatment programs working with individuals with body dissatisfaction and disordered eating should therefore assess individuals’ social comparison orientation. Our study suggests that clients who are less likely to compare themselves with others in general may benefit most from thinking about appearance comparisons from the perspective of social comparison theory and from making more downward comparisons. Nevertheless, more research is needed to determine whether this strategy can provide sustainable benefits to these individuals, and if so, which aspects are most helpful and least harmful (e.g., attending to one’s positive qualities but not focusing on one’s superiority). Furthermore, a nuanced approach in which practitioners do not inadvertently foster overreliance on comparing oneself to others is crucial. Our preliminary results suggest that for those with higher social comparison orientation, a different perspective and strategy than those offered by social comparison theory may be advantageous. Specifically, by cultivating feelings of compassion and well-wishes towards the people with whom they compare, these individuals may experience greater reductions in body dissatisfaction, restrained eating, and eating disorder-related comparisons. Moreover, the practice may help them to extract themselves from a deeply entrenched mentality where others in their social world are seen as threats, rivals, and information sources about the self, and instead promote a greater sense of connectedness and shared humanity with others. Given that individuals with disordered eating and body dissatisfaction experience a range of social struggles (Hinrichsen, Wright, Waller, & Meyer, 2003; Striegel-Moore, Silberstein, & Rodin, 1993; Tiller et al., 1997), such a contribution would be nontrivial. While the present study has brought to light an interesting collection of results, it naturally leads to questions about how adopting a caregiving mentality toward others ultimately brings about changes in one’s own body and eating-related cognitions and behaviors. Future research should investigate the mechanisms at play, especially for those who are highly comparison-oriented. In this vein, one promising line of exploration may be the construct of self-compassion. Gilbert et al. (2009) suggests that the caregiving mindset, when activated, should promote feelings of connectedness and warmth in one’s relationships, as opposed to the insecurity and distress in the relationships that are typical of the competitive mentality. This newfound sense of safeness in one’s social environment should foster a more stable form of self-acceptance that is not predicated on where one stands relative to others, but is rather derived from a sense of shared humanity. The construct of self-compassion may be one such form of self-acceptance, given its characterization as the tendency to respond to personal distress and inadequacies with kindness, non-judgment, and the recognition that others also face similar struggles (Neff, 2003). Also encouraging is the large body of evidence that shows self-compassion to be a significant contributor to well-being (Zessin et al., 2015) and to fewer body image concerns and greater body image acceptance (e.g., Albertson, Neff, & Dill-Shackleford, 2015; Braun, Park, & Gorin, 2016; Kelly, Vimalakanthan, & Carter, 2014; Kelly, Vimalakanthan, & Miller, 2014). Given that self-compassion is linked to both the caregiving mindset and to benefits in the body image domain, further exploration as to whether the caregiving mindset works to indirectly improve body- and eating-related outcomes by

improving self-compassion may be a promising avenue of future research.

4.3. Limitations and future directions This study had a number of limitations that should be addressed in future research. First, we had a relatively small and homogeneous participant sample composed of young females who were predominantly students. Although the study was limited to female participants on the basis of research pointing to a higher frequency of upward appearance comparisons among women and subsequently, a stronger link to body image dysfunction (Strahan et al., 2006; Thompson & Heinberg, 1993), it remains the case that both men and women make appearance comparisons. Interestingly, extant literature also suggests that women may be more focused on emotional expressions of care relative to men, who may be more instrumentally-oriented (Gilligan, 1982). In consequence, while our intervention may have been well-matched to the population towards which it was targeted, in future work it would be interesting to see how well it generalizes to men, although the measures used would have to differ; many of the measures used in this study were developed for use with and psychometrically validated on predominantly female samples. Future work should also broaden the generalizability of the present study’s findings by investigating these interventions’ effectiveness among younger and older populations, as well as individuals recruited from community-based settings and clinical settings. Second, this study relied wholly on self-report measures; as a result, some participants may have been affected by selfpresentation concerns when completing questionnaires. Future work should include more objective measures, such as behavioral measures (e.g., restrained eating task), in the study design to address such concerns. Furthermore, in investigating dependent variables, we used trait-level measures. Although the instructions were amended to ask participants to consider only the pertinent time frame (e.g., the 48-h contractual period), using state measures in future research that ask about momentary experiences may provide an alternative approach for investigating the effects of these interventions. In addition, while the naturalistic nature of the contract paradigm is a strength of this study in terms of generalizability, it is also the case that we cannot be certain about participants’ accuracy in employing their assigned strategy during the contractual period, beyond what they self-reported. Third, to make our interventions feasible for participants to perform, this study was conducted over a relatively short-term period of 48 h. In future studies, more time may help participants become used to their assigned intervention, and as a result, more fully benefit from its effects. This may be particularly true of the Caregiving intervention, which was likely a novel strategy for the participants to whom it was assigned. Future work should also explore the sustainability of the different strategies and their effects over time through subsequent follow-ups. For example, while the Competition intervention was demonstrably advantageous to a subset of women in this study, its effectiveness as a strategy may degrade over time, which could explain the mixed research on the effectiveness of downward comparisons (e.g., Fitzsimmons-Craft, 2017; Leahey et al., 2011; Lin & Kulik, 2002; Lin & Soby, 2016). Fourth, our analytic approach forced us to exclude participants who did not complete post-intervention measures, leaving us with a sample that is potentially biased. However, attrition was low (7 out of 136) and did not differ significantly across conditions. Furthermore, no significant differences were found in baseline scores or scores of intervention credibility and expectancy between participants who completed the study and participants who did not.

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

5. Conclusions This study was the first to our knowledge to apply social mentalities theory to the problem of appearance comparisons, and the first to explore the benefits of adopting a caregiving mindset vis-à-vis the targets of one’s unfavorable social comparisons. Findings suggest that this compassion-based approach is feasible and acceptable to young women, and is especially helpful for those who are most comparison-oriented, and arguably most vulnerable to the threat of appearance comparisons. Funding statement This work was supported by the Social Sciences and Humanities Research Council of Canada, the Government of Ontario, and the University of Waterloo. References Albertson, E. R., Neff, K. D., & Dill-Shackleford, K. E. (2015). Self-compassion and body dissatisfaction in women: A randomized controlled trial of a brief meditation intervention. Mindfulness, 6, 444–454. http://dx.doi.org/10.1007/ s12671-014-0277-3 Bailey, S. D., & Ricciardelli, L. A. (2010). Social comparisons, appearance related comments, contingent self-esteem and their relationships with body dissatisfaction and eating disturbance among women. Eating Behaviors, 11, 107–112. http://dx.doi.org/10.1016/j.eatbeh.2009.12.001 Blechert, J., Nickert, T., Caffier, D., & Tuschen-Caffier, B. (2009). Social comparison and its relation to body dissatisfaction in bulimia nervosa: Evidence from eye movements. Psychosomatic Medicine, 71, 907–912. http://dx.doi.org/10.1097/ PSY.0b013e3181b4434d Boone, L., Soenens, B., Vansteenkiste, M., & Braet, C. (2012). Is there a perfectionist in each of us? An experimental study on perfectionism and eating disorder symptoms. Appetite, 59, 531–540. http://dx.doi.org/10.1016/j.appet.2012.06. 015 Bowlby, J. (1982). (2nd ed.). Attachment and loss (Vol. 1) New York, NY: Basic Books. Braun, T. D., Park, C. L., & Gorin, A. (2016). Self-compassion, body image, and disordered eating: A review of the literature. Body Image, 17, 117–131. http:// dx.doi.org/10.1016/j.bodyim.2016.03.003 Breines, J. G., & Chen, S. (2013). Activating the inner caregiver: The role of support-giving schemas in increasing state self-compassion. Journal of Experimental Social Psychology, 49, 58–64. http://dx.doi.org/10.1016/j.jesp. 2012.07.015 Brown, S. L., Nesse, R. M., Vinokur, A. D., & Smith, D. M. (2003). Providing social support may be more beneficial than receiving it: Results from a prospective study of mortality. Psychological Science, 14, 320–327. http://dx.doi.org/10. 1111/1467-9280.14461 Buunk, A. P., & Gibbons, F. X. (2006). Social comparison orientation: A new perspective on those who do and those who don’t compare with others. In S. Guimond (Ed.), Social comparison and social psychology: understanding cognitions, intergroup relations, and culture (pp. 15–32). Cambridge, UK: Cambridge University Press. Buunk, B. R., Collins, R. L., Taylor, S. E., Van Yperen, N. W., & Dakof, G. A. (1990). The affective consequences of social comparison: Either direction has its ups and downs. Journal of Personality and Social Psychology, 59, 1238–1249. http://dx. doi.org/10.1037/0022-3514.59.6.1238 Carson, J. W., Keefe, F. J., Lynch, T. R., Carson, K. M., Goli, V., Fras, A., & Thorp, S. R. (2005). Loving-kindness meditation for chronic low back pain: Results from a pilot trial. Journal of Holistic Nursing, 23, 287–304. http://dx.doi.org/10.1177/ 0898010105277651 Collins, R. L. (1996). For better or worse: The impact of upward social comparison on self-evaluations. Psychological Bulletin, 119, 51–69. http://dx.doi.org/10. 1037/0033-2909.119.1.51 Cosley, B. J., McCoy, S. K., Saslow, L. R., & Epel, E. S. (2010). Is compassion for others stress buffering? Consequences of compassion and social support for physiological reactivity to stress. Journal of Experimental Social Psychology, 46, 816–823. http://dx.doi.org/10.1016/j.jesp.2010.04.008 Devilly, G. J., & Borkovec, T. D. (2000). Psychometric properties of the Credibility/Expectancy Questionnaire. Journal of Behavior Therapy and Experimental Psychiatry, 31, 73–86. http://dx.doi.org/10.1016/S00057916(00)00012-4 Dijkstra, P., Gibbons, F. X., & Buunk, A. P. (2011). Social comparison theory. In J. E. Maddux & J. P. Tangney (Eds.), Social psychological foundations of clinical psychology (pp. 195–211). New York, NY: The Guilford Press. Evans, C., & Dolan, B. (1993). Body Shape Questionnaire: Derivation of shortened alternate forms. International Journal of Eating Disorders, 13, 315–321. http:// dx.doi.org/10.1002/1098-108X(199304)13:3<315:AID-EAT2260130310>3.0. CO;2-3 Fairburn, C. G., & Beglin, S. J. (1994). Assessment of eating disorders: Interview or self-report questionnaire? International Journal of Eating Disorders, 16,

161

363–370. http://dx.doi.org/10.1002/1098-108X(199412)16:4<363:AIDEAT2260160405>3.0.CO;2-# Fairburn, C. G. (2008). Cognitive behavior therapy and eating disorders. New York, NY: Guilford Press. Feldman, C., & Kuyken, W. (2011). Compassion in the landscape of suffering. Contemporary Buddhism, 12, 143–155. http://dx.doi.org/10.1080/14639947. 2011.564831 Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117–140. http://dx.doi.org/10.1177/001872675400700202 Fitzsimmons-Craft, E. E., Bardone-Cone, A. M., & Harney, M. B. (2012). Development and validation of the Body, Eating, and Exercise Comparison Orientation Measure (BEECOM) among college women. Body Image, 9, 476–487. http://dx.doi.org/10.1016/j.bodyim.2012.07.007 Fitzsimmons-Craft, E. E. (2017). Eating disorder-related social comparison in college women’s everyday lives. International Journal of Eating Disorders, 50, 893–905. http://dx.doi.org/10.1002/eat.22725 Fredrickson, B. L., Cohn, M. A., Coffey, K. A., Pek, J., & Finkel, S. M. (2008). Open hearts build lives: Positive emotions, induced through loving-kindness meditation, build consequential personal resources. Journal of Personality and Social Psychology, 95, 1045–1062. http://dx.doi.org/10.1037/a0013262 Gibbons, F. X. (1986). Social comparison and depression: Company’s effect on misery. Journal of Personality and Social Psychology, 51, 140–148. http://dx.doi. org/10.1037/0022-3514.51.1.140 Gibbons, F. X., & Buunk, B. P. (1999). Individual differences in social comparison: Development of a scale of social comparison orientation. Journal of Personality and Social Psychology, 76, 129–142. http://dx.doi.org/10.1037/0022-3514.76.1. 129 Gilbert, P., McEwan, K., Bellew, R., Mills, A., & Gale, C. (2009). The dark side of competition: How competitive behavior and striving to avoid inferiority are linked to depression, anxiety, stress and self-harm. Psychology and Psychotherapy, 82, 123–136. http://dx.doi.org/10.1348/147608308X 379806 Gilbert, P. (1989). Human nature and suffering. Hove, UK: Psychology Press. Gilbert, P. (2000). Social mentalities: Internal social conflict and the role of inner warmth and compassion in cognitive therapy. In P. Gilbert & K. G. Bailey (Eds.), Genes on the couch: Explorations in evolutionary psychotherapy (pp. 118–150). New York, NY: Brunner-Routledge. Gilbert, P. (2005). Compassion: Conceptualisations, research and use in psychotherapy. Hove, UK: Routledge. Gilbert, P. (2010). Compassion focused therapy: Distinctive features. London, UK: Routledge. Gilbert, P. (2014). The origins and nature of compassion focused therapy. British Journal of Clinical Psychology, 53, 6–41. http://dx.doi.org/10.1111/bjc.12043 Gilligan, C. (1982). In a different voice: Psychological theory and women’s development. Cambridge, MA: Harvard University Press. Gollwitzer, P. M. (1999). Implementation intentions: Strong effects of simple plans. American Psychologist, 54, 493–503. http://dx.doi.org/10.1037/0003-066X.54.7. 493 Groesz, L. M., Levine, M. P., & Murnen, S. K. (2002). The effect of experimental presentation of thin media images on body satisfaction: A meta-analytic review. International Journal of Eating Disorders, 31, 1–16. http://dx.doi.org/10. 1002/eat.10005 Hakmiller, K. L. (1966). Threat as a determinant of downward comparisons. Journal of Experimental Social Psychology, 1, 32–39. http://dx.doi.org/10.1016/00221031(66)90063-1 Helgeson, V. S., & Taylor, S. E. (1993). Social comparisons and adjustment among cardiac patients. Journal of Applied Social Psychology, 23, 1171–1195. http://dx. doi.org/10.1111/j.1559-1816.1993.tb01027.x Hermanto, N., & Zuroff, D. C. (2016). The social mentality theory of self-compassion and self-reassurance: The interactive effect of care-seeking and caregiving. Journal of Social Psychology, 156, 523–535. http://dx.doi.org/10.1080/00224545. 2015.1135779 Hinrichsen, H., Wright, F., Waller, G., & Meyer, C. (2003). Social anxiety and coping strategies in the eating disorders. Eating Behaviors, 4, 117–126. http://dx.doi. org/10.1016/S1471-0153(03)00016-3 Hutcherson, C. A., Seppala, E. M., & Gross, J. J. (2008). Loving-kindness meditation increases social connectedness. Emotion, 8, 720–724. http://dx.doi.org/10. 1037/a0013237 Kelly, A. C., Vimalakanthan, K., & Carter, J. C. (2014). Understanding the roles of self-esteem, self-compassion, and fear of self-compassion in eating disorder pathology: An examination of female students and eating disorder patients. Eating Behaviors, 15, 388–391. http://dx.doi.org/10.1016/j.eatbeh.2014.04.008 Kelly, A. C., Vimalakanthan, K., & Miller, K. E. (2014). Self-compassion moderates the relationship between body mass index and both eating disorder pathology and body image flexibility. Body Image, 11, 446–453. http://dx.doi.org/10.1016/ j.bodyim.2014.07.005 Konrath, S., Fuhrel-Forbis, A., & Lou, A. (2012). Motives for volunteering are associated with mortality risk in older adults. Health Psychology, 31, 87–96. http://dx.doi.org/10.1037/a0025226 Leahey, T. M., Crowther, J. H., & Mickelson, K. D. (2007). The frequency, nature, and effects of naturally occurring appearance-focused social comparisons. Behavior Therapy, 38, 132–143. http://dx.doi.org/10.1016/j.beth.2006.06.004 Leahey, T. M., Crowther, J. H., & Ciesla, J. A. (2011). An ecological momentary assessment of the effects of weight and shape social comparisons on women with eating pathology, high body dissatisfaction, and low body dissatisfaction. Behavior Therapy, 42, 197–210. http://dx.doi.org/10.1016/j.beth.2010.07.003

162

K. Vimalakanthan et al. / Body Image 25 (2018) 148–162

Lemyre, L., & Smith, P. M. (1985). Intergroup discrimination and self-esteem in the minimal group paradigm. Journal of Personality and Social Psychology, 49, 660–670. http://dx.doi.org/10.1037/0022-3514.49.3.660 Lew, A. M., Mann, T., Myers, H., Taylor, S., & Bower, J. (2007). Thin-ideal media and women’s body dissatisfaction: Prevention using downward social comparisons on non-appearance dimensions. Sex Roles, 57, 543–556. http://dx.doi.org/10. 1007/s11199-007-9274-5 Lin, L. F., & Kulik, J. A. (2002). Social comparison and women’s body satisfaction. Basic and Applied Social Psychology, 24, 115–123. http://dx.doi.org/10.1207/ S15324834BASP2402 Lin, L. F., & Soby, M. (2016). Appearance comparisons styles and eating disordered symptoms in women. Eating Behaviors, 23, 7–12. http://dx.doi.org/10.1016/j. eatbeh.2016.06.006 Luce, K. H., & Crowther, J. H. (1999). The reliability of the Eating Disorder Examination—Self-Report Questionnaire Version (EDE-Q). International Journal of Eating Disorders, 25, 349–351. http://dx.doi.org/10.1002/(SICI)1098108X(199904)25:3<349:AID-EAT15 > 3.0.CO;2-M Mond, J. M., Hay, P. J., Rodgers, B., Owen, C., & Beumont, P. J. (2004). Validity of the Eating Disorder Examination Questionnaire (EDE-Q) in screening for eating disorders in community samples. Behavior Research and Therapy, 42, 551–567. http://dx.doi.org/10.1016/S0005-7967(03)00161-X Morrison, T. G., Kalin, R., & Morrison, M. A. (2004). Body-image evaluation and body-image investment among adolescents: A test of sociocultural and social comparison theories. Adolescence, 39, 571–592. Morse, S., & Gergen, K. J. (1970). Social comparison, self-consistency, and the concept of self. Journal of Personality and Social Psychology, 16, 148–156. http:// dx.doi.org/10.1037/h0029862 Murnen, S. K., & Smolak, L. (2015). Gender and eating disorders. In L. Smolak & M. P. Levine (Eds.), The Wiley handbook of eating disorders (pp. 352–366). Chichester, UK: John Wiley & Sons. Myers, T. A., & Crowther, J. H. (2009). Social comparison as a predictor of body dissatisfaction: A meta-analytic review. Journal of Abnormal Psychology, 118, 683–698. http://dx.doi.org/10.1037/a0016763 Neff, K. D., & Pommier, E. (2013). The relationship between self-compassion and other-focused concern among college undergraduates, community adults, and practicing meditators. Self and Identity, 12, 160–176. http://dx.doi.org/10.1080/ 15298868.2011.649546 Neff, K. (2003). Self-compassion: an alternative conceptualization of a healthy attitude toward oneself. Self and Identity, 2, 85–101. http://dx.doi.org/10.1080/ 15298860309032 O’Brien, K. S., Caputi, P., Minto, R., Peoples, G., Hooper, C., Kell, S., & Sawley, E. (2009). Upward and downward physical appearance comparisons: Development of scales and examination of predictive qualities. Body Image, 6, 201–206. http://dx.doi.org/10.1016/j.bodyim.2009.03.003 Paxton, S. J., Eisenberg, M. E., & Neumark-Sztainer, D. (2006). Prospective predictors of body dissatisfaction in adolescent girls and boys: A five-year longitudinal study. Developmental Psychology, 42, 888–899. http://dx.doi.org/ 10.1037/0012-1649.42.5.888 Pinto-Gouveia, J., Ferreira, C., & Duarte, C. (2014). Thinness in the pursuit for social safeness: An integrative model of social rank mentality to explain eating psychopathology. Clinical Psychology & Psychotherapy, 21, 154–165. http://dx. doi.org/10.1002/cpp.1820 Poulin, M. J., Brown, S. L., Dillard, A. J., & Smith, D. M. (2013). Giving to others and the association between stress and mortality. American Journal of Public Health, 103, 1649–1655. http://dx.doi.org/10.2105/AJPH.2012.300876 Rancourt, D., Schaefer, L. M., Bosson, J. K., & Thompson, J. K. (2016). Differential impact of upward and downward comparisons on diverse women’s disordered eating behaviors and body image. International Journal of Eating Disorders, 49, 519–523. http://dx.doi.org/10.1002/eat.22470

Salovey, P., & Rodin, J. (1984). Some antecedents and consequences of social-comparison jealousy. Journal of Personality and Social Psychology, 19, 780–792. http://dx.doi.org/10.1037/0022-3514.47.4.780 Salzberg, S. (1995). Loving-kindness: The revolutionary art of happiness. Boston, MA: Shambala Publications. Salzberg, S. (2011). Mindfulness and loving-kindness. Contemporary Buddhism, 12, 177–182. http://dx.doi.org/10.1080/14639947.2011.564837 Seppala, E. M., Hutcherson, C. A., Nguyen, D. T., Doty, J. R., & Gross, J. J. (2014). Loving-kindness meditation: A tool to improve healthcare provider compassion, resilience, and patient care. Journal of Compassionate Health Care, 1, 5. http://dx.doi.org/10.1186/s40639-014-0005-9 Smith, K. E., Mason, T. B., Murray, S. B., Griffiths, S., Leonard, R. C., Wetterneck, C. T., . . . & Lavender, J. M. (2017). Male clinical norms and sex differences on the Eating Disorder Inventory (EDI) and Eating Disorder Examination Questionnaire (EDE-Q). International Journal of Eating Disorders, 50, 769–775. http://dx.doi.org/10.1002/eat.22716 Snoek, H. M., van Strien, T., Janssens, J. M. A. M., & Engels, R. C. M. E. (2008). Restrained eating and BMI: A longitudinal study among adolescents. Health Psychology, 27, 753–759. http://dx.doi.org/10.1037/0278-6133.27.6.753 Strahan, E. J., Wilson, A. E., Cressman, K. E., & Buote, V. M. (2006). Comparing to perfection: How cultural norms for appearance affect social comparisons and self-image. Body Image, 3, 211–227. http://dx.doi.org/10.1016/j.bodyim.2006. 07.004 Striegel-Moore, R. H., Silberstein, L. R., & Rodin, J. (1993). The social self in bulimia nervosa: Public self-consciousness, social anxiety, and perceived fraudulence. Journal of Abnormal Psychology, 10, 297–303. http://dx.doi.org/10.1037/0021843X.102.2.297 Thompson, J. K., & Heinberg, L. J. (1993). Preliminary test of two hypotheses of body image disturbance. International Journal of Eating Disorders, 14, 59–63. http://dx.doi.org/10.1002/1098-108X(199307)14:1<59::AIDEAT2260140108>3.0.CO;2-T Tiller, J. M., Sloane, G., Schmidt, U., Troop, N., Power, M., & Treasure, J. L. (1997). Social support in patients with anorexia nervosa and bulimia nervosa. International Journal of Eating Disorders, 21, 31–38. http://dx.doi.org/10.1002/ (SICI)1098-108X(199701)21:1<31:AID-EAT4>3.0.CO;2-4 van den Berg, P., & Thompson, J. K. (2007). Self-schema and social comparison explanations of body dissatisfaction: A laboratory investigation. Body Image, 4, 29–38. http://dx.doi.org/10.1016/j.bodyim.2006.12.004 Vogel, E. A., Rose, J. P., Okdie, B. M., Eckles, K., & Franz, B. (2015). Who compares and despairs? The effect of social comparison orientation on social media use and its outcomes. Personality and Individual Differences, 86, 249–256. http://dx. doi.org/10.1016/j.paid.2015.06.026 Wasylkiw, L., MacKinnon, A. L., & MacLellan, A. M. (2012). Exploring the link between self-compassion and body image in university women. Body Image, 9, 236–245. http://dx.doi.org/10.1016/j.bodyim.2012.01.007 Wills, T. A. (1981). Downward comparison principles in social psychology. Psychological Bulletin, 90, 245–271. http://dx.doi.org/10.1037/0033-2909.90.2. 245 Wood, J. V., Taylor, S. E., & Lichtman, R. R. (1985). Social comparison in adjustment to breast cancer. Journal of Personality and Social Psychology, 49, 1169–1183. http://dx.doi.org/10.1037/0022-3514.49.5.1169 Wood, J. V. (1989). Theory and research concerning social comparisons of personal attributes. Psychological Bulletin, 106, 231–248. http://dx.doi.org/10.1037/ 0033-2909.106.2.231 Zessin, U., Dickhäuser, O., & Garbade, S. (2015). The relationship between self-compassion and well-being: A meta-analysis. Applied Psychology: Health and Well-Being, 7, 340–364. http://dx.doi.org/10.1111/aphw.12051