Effortless perfectionism and its relationship with body dissatisfaction, and pathological eating and exercise: the mediating role of self-kindness and self-criticism

ABSTRACT

Objective

The current study investigated the interrelationships between effortless perfectionism (a desire to appear perfect without outward displays of effort) and body dissatisfaction, pathological eating, and pathological exercise. In addition, the mediating role of self-kindness and self-criticism in these relationships was explored.

Method

Using a cross-sectional design, undergraduate students (N = 497) were asked to complete a battery of questionnaires assessing these key variables.

Results

Effortless perfectionism was indirectly positively associated with more severe body dissatisfaction and pathological eating/exercise, via higher levels of self-criticism (controlling for self-kindness). By contrast, after controlling for self-criticism, self-kindness did not mediate the relationships between effortless perfectionism and body dissatisfaction, pathological eating, or pathological exercise

Conclusions

Self-criticism uniquely emerged as a mediating mechanism; the desire for effortless perfection was associated with higher self-criticism, which in turn was linked with worsened symptomatology. By contrast, self-kindness did not emerge as a unique mediating mechanism in this context. Longitudinal research would be useful to establish the effect of situational context, and possible reciprocal relationships between self-kindness and self-criticism over time.

Key Points

What is already known about this topic:

1. Body dissatisfaction, and pathological eating and exercise, are positively intercorrelated.

2. These symptoms are characterised by chronicity and physical and psychological impairment.

3. Negative perfectionism has been implicated in the onset and maintenance of eating and body-related pathology.

What this topic adds:

1. Effortless perfectionism predicted body dissatisfaction and symptoms of pathological eating/exercise.

2. Self-criticism but not self-kindness uniquely positively mediated these relationships.

3. Findings identified new paths linking negative perfectionism and disordered eating.

KEYWORDS:

• Self-compassion
• perfectionism
• body image
• disordered eating
• exercise addiction

Introduction

Pathological eating (i.e., dietary restraint, binge eating, purging) has been situated as both a consequence of body image dissatisfaction (disgruntlement with one’s physical appearance, body weight, and/or shape), and a precursor to pathological exercise (i.e., inflexible exercise obsessions with or without repetitive/compensatory exercise behaviour) (Cosh et al., 2023; Cuesta-Zamora et al., 2022; Dawson & Hammer, 2020; Ganson et al., 2022; Holsen et al., 2012; Ruiz-Turrero et al., 2022). These interconnected variables are pervasive (Ganson et al., 2022; Quittkat et al., 2019), and their consequences significant. Body dissatisfaction has been linked with unsafe substance use, self-harm, low self-esteem, and poor psychological health (Corazza et al., 2019; Dawson & Hammer, 2020; Holsen et al., 2012; Ruiz-Turrero et al., 2022). Pathological eating is characterised by high comorbidity (with mood, anxiety, and substance use disorders), chronicity (typically emerging in adolescence and persisting into adulthood), and significant physical and psychological impairment (Dalle Grave et al., 2023; Linardon, 2021; Turk & Waller, 2020; Williams & Levinson, 2022). Similarly, pathological exercise has been described as unhealthy, destructive, addictive, and psychologically burdensome (Bratland-Sanda et al., 2019). In addition to its positive relationship with body dissatisfaction and pathological eating, compulsive exercise has been linked with substance use, comorbid poor mental health, emotional dysregulation, suicidal ideation, self-injury, and poorer quality-of-life (Chamberlain & Grant, 2020; Corazza et al., 2019; Cuesta-Zamora et al., 2022; Ganson et al., 2022). Thus, given the widespread and harmful impact of these phenomena, the current study aimed to further unpack the correlates of worsened (or improved) symptomatology.

Perfectionism is one variable that has been implicated in the onset and maintenance of body dissatisfaction and pathological eating/exercise (González-Hernández et al., 2022). Perfectionism has been conceptualised as a multidimensional (i.e., not wholly dysfunctional) construct (Dunkley et al., 2006); positive perfectionism can stimulate healthy goal striving and desire for self-improvement (González-Hernández et al., 2022), whereas negative perfectionism involves holding oneself to unattainably high personal standards whilst simultaneously intensely fearing or avoiding the possibility of failure (Hamedani et al., 2023). A particularly dysfunctional sub-type of negative perfectionism is effortless perfectionism (Flett et al., 2016).

Effortless perfectionism is characterised by a desire to appear perfect (i.e., talented, intelligent, fit, popular, or attractive) to others without any outward display or appearance of effort (Flett et al., 2016; Travers et al., 2015). The desire for “natural” perfection makes this perfectionistic style especially maladaptive; effortless perfectionism encompasses the typical qualities of negative perfectionism (i.e., an inflexible expectation to be unrealistically perfect), but effortless perfectionists expect that this should be achieved without needing to try or reveal to others that any energy has been expended in the pursuit for such perfection. Even in the unlikely event that the desired level of perfection is achieved, the accomplishment is plagued by feelings of inadequacy if it required the exertion of great effort (Flett et al., 2016). As stated by Travers et al. (2015), with effortless perfectionism attainment of perfection alone is not enough; perfection without the appearance of having even tried is the ultimate outcome.

Research evidence has linked the high personal standards, inflexible adherence to rules, rigidity in behaviour, and dysfunctional self-evaluation processes characterised by negative perfectionism with higher levels of body dissatisfaction, and more dysfunctional eating and exercise behaviours (Flett et al., 2016; González-Hernández et al., 2022). Negative perfectionism features as one of four maintaining factors in the transdiagnostic theory of disordered eating (Cooper & Dalle Grave, 2017; Cooper & Fairburn, 2011; González-Hernández et al., 2022). Moreover, in relation to effortless perfectionism, active concealment of pathological eating is common (O’Connor et al., 2021), eating in secret has been identified as a key behavioural indicator of eating concern, shame, loss of control of eating, and binge/purge behaviours (Lydecker & Grilo, 2019), and pathological exercise has been conceptualised as involving adherence to strict exercise rules, inflexible exercise schedules, perpetual dissatisfaction with exercise performance, and exercise that is secretive or hidden (Ruiz-Turrero et al., 2022). Thus, the theory and evidence support the positive link between effortless perfectionism, or the desire to appear perfect to others without outward signs of work or exertion, and increased symptoms of body dissatisfaction along with pathological eating/exercise.

One mechanism that may intensify the link between effortless perfectionism and poorer outcomes in relation to body dissatisfaction and pathological eating/exercise is self-criticism. Self-criticism, in the form of constant and harsh self-evaluation and scrutiny, is a “transdiagnostic vulnerability factor that is associated with many different psychopathologies” (Krieger et al., 2019, p. 431). Self-criticism has been described as the most pathological element of perfectionism (Cuesta-Zamora et al., 2022; Dunkley et al., 2006; Williams & Levinson, 2022); it functions as both a precursor to, and a consequence of, unrelenting perfectionistic standards (Hamedani et al., 2023). If a goal is not achieved, no matter how unattainable, the result is self-criticism. Even if a goal is achieved, a negative perfectionist will self-criticise for not having set the bar higher or, in the case of effortless perfectionism, for effort that was expended in pursuit of the goal (Hamedani et al., 2023).

A stronger tendency to self-criticise or feel “entrapped” by a self-critical inner voice, also predicts heightened body dissatisfaction and pathological eating/exercise (Werner et al., 2019, p. 531). Extremely high standards and feelings of inadequacy regarding appearance or weight can increase the drive to be thinner, dissatisfaction with one’s body, and the adoption of compensatory eating and exercise behaviours (Cooper & Fairburn, 2011; Williams & Levinson, 2022). Self-criticism predicts more severe eating disorder symptomatology (Werner et al., 2019), and is the aspect of dysfunctional perfectionism most strongly linked with compulsive exercise (González-Hernández et al., 2022). Indeed, positive correlations have been observed between self-criticism and avoidance and rule-driven exercise behaviour, exercising for weight and shape reasons, and exercise rigidity (Taranis & Meyer, 2010). Thus, self-criticism likely links effortless perfectionism with worsened body dissatisfaction and pathological eating/exercise symptomatology.

A proposed antidote to self-criticism is self-compassion (Wakelin et al., 2022). Self-compassion involves being accepting and understanding towards oneself and recognising flaws and mistakes as part of being human (Cuesta-Zamora et al., 2022; Neff, 2003). By contrast to self-criticism, self-compassion operates independently of self-evaluations or the requirement to aspire towards, or conform with, appearance or performance standards (Magnus et al., 2010; Neff, 2003). Accordingly, self-compassion has been inversely correlated with negative perfectionism, whilst simultaneously being linked with strong personal standards, increased accountability and responsibility, and a self-loving desire to reduce behaviours that are harmful to health and well-being (Hamedani et al., 2023; Neff, 2003). It follows that self-compassion has been identified as a primary intervention approach for individuals with high levels of self-criticism (Wakelin et al., 2022; Werner et al., 2019), and as a protective factor against body dissatisfaction, and pathological eating and exercise behaviours (Cuesta-Zamora et al., 2022; Magnus et al., 2010).

Homan and Tylka (2015), for example, reported that higher levels of self-compassion eliminated the negative association between body-related threats (i.e., unfavourable social comparison) and body appreciation. Conversely, among those with lower self-compassion, the relationship between body-related threat and body appreciation was strong and negative. Linardon (2021) further revealed that higher body satisfaction and self-compassion reduced the likelihood of future emergence of pathological eating symptomatology after eight-months. Moreover, Magnus et al. (2010) reported a negative relationship between self-compassion and obligatory exercise, which was conceptualised as excessive exercise frequency and intensity. Theoretically, self-compassion has been argued to reduce dysfunctional coping strategies such as avoidance, suppression, and compensatory behaviours (i.e., pathological eating and exercise), instead increasing capacity to regulate negative thoughts and emotions and behave in ways that support health and well-being (Edlund et al., 2022; Linardon, 2021; Magnus et al., 2010; Turk & Waller, 2020).

It could be reasonably surmised from the available evidence that self-criticism and self-compassion may mediate the relationship between effortless perfectionism and body dissatisfaction and pathological eating/exercise. However, recent evidence challenges conceptualisations of self-compassion as encompassing both positive (i.e., self-kindness) and negative responding (i.e., self-judgement), indicating instead that self-compassion and self-criticism are independent constructs that do not represent opposing ends of a single continuum (Ferrari et al., 2022; Halamová & Kanovský, 2019; Krieger et al., 2019; López et al., 2015). It is therefore important to disentangle these constructs to explore their independent impact. The most precise delineation in the current context is to investigate the unique functions of self-criticism and a key component of self-compassion (self-kindness), whilst controlling for the influence of the other (Neff, 2003). Accordingly, in the current study we explored the mediating role of self-kindness (controlling for self-criticism) and the mediating role of self-criticism (controlling for self-kindness) in the relationship between effortless perfectionism and the outcome variables: body dissatisfaction, pathological eating, and pathological exercise. These symptoms are generally more prevalent among young females but are also present in males (Dalle Grave et al., 2023; Qian et al., 2022). Many of these pathological symptoms also emerge in adolescence and young adulthood, and persist into adulthood (Dalle Grave et al., 2023). Thus, we explored these relationships in a predominantly young adult sample, with no restrictions on gender or age. We anticipated positive relationships between effortless perfectionism and each key criterion variable, an inverse relationship between self-kindness and self-criticism, along with a protective mediation pathway for self-kindness and perpetuating mediation pathway for self-criticism. Specifically, we predicted negative relationships between effortless perfectionism and self-kindness and between self-kindness and each key criterion variable; the reverse pattern was predicted for self-criticism.

Materials and methods

Participants

The sample comprised 497 undergraduate students who ranged in age from 17 to 57 years (M = 24.10, SD = 8.58), 71.31% of whom were young adults ≤24 years of age. Participants identified as male (n = 141, 28%), female (n = 351, 71%), or non-binary (n = 5, 1%). Some participants had a certificate (14%) or Bachelor degree (11%), however, for most participants the highest educational attainment reported was Year 12 (54%). Most participants classified their ethnicity as Caucasian/White (77%), were either single (48%) or in a relationship (32%), and were full-time students (34%) or working in part-time or casual employment (46%). The large majority were not currently dieting (n = 404, 81%). After removing four univariate outliers, participants were in the healthy weight range, on average, according to Body Mass Index (BMI), (M = 24.14, SD = 4.63). Using adult cut-offs for participants ≥ 18, and adjusted cut-offs for participants 17 years old (n = 15, Cole et al., 2000, 2007), our sample were 5.5% underweight, 62.1% healthy weight, 21.1% overweight, and 10.8% obese.

Measures

Effortless perfectionism

The Effortless Perfectionism Scale (EPS; Travers et al., 2015) consists of 10-items (e.g., “I try to make my achievements look effortless”). Participants indicate the extent to which they endorsed each item on a scale from 1 (not at all) to 5 (extremely) and a total score was calculated by summing the items. Higher scores represented higher interpersonal expression of effortless perfectionism. Internal consistency was excellent (α = .86).

Self-kindness

The 5-item self-kindness subscale from the Self-Compassion Scale (SCS; Neff, 2003) asked participants to rate how frequently they engaged in each statement (e.g., “I’m kind to myself when I’m experiencing suffering”) on a scale from 1 (almost never) to 6 (almost always). A mean score was calculated whereby higher scores represented higher self-kindness. Internal consistency was excellent (α = .87).

Self-criticism

The inadequate-self (9-items) and hated-self (5-items) subscales of the Forms of Self-Criticising/Attacking and Self-Reassuring Scale (FSCRS; Gilbert et al., 2004) were used to measure self-criticism. Participants were asked to indicate how much each of the statements (e.g., “There is a part of me that puts me down”) represented them on a scale from 0 (not at all like me) to 4 (extremely like me). Responses were summed to create a total self-criticism score, where higher scores represented higher criticism. Internal consistency of the 14-items comprising the inadequate-self and hated-self subscales was excellent (α = .94).

Body image dissatisfaction

Body image dissatisfaction was measured using the 4-item Body Image Satisfaction Scale (BISS). Each item (e.g., “I would like to change a good deal about my body”) was rated on a scale from 1 (does not apply at all) to 6 (applies exactly) (Holsen et al., 2012). Two items were reverse scored, and a total score was calculated with higher scores representing more body image dissatisfaction. Internal consistency was excellent (α = .89).

Pathological eating

The Eating Disorder Examination Questionnaire (EDE-Q; Fairburn & Beglin, 1994) consists of 22-items asking about the frequency of pathological eating behaviours during the past 28-days. Participants responded to each question (e.g., “Have you had a definite fear of losing control over eating?) on a scale from 0 (no days) to 6 (every day). A global score was obtained by calculating a grand mean from the four subscale means (restraint, eating concern, shape concern, and weight concern). Higher scores represented higher pathological eating symptomatology and internal consistency for the EDE-Q global score was excellent (α = .95).

Pathological exercise

The 8-item avoidance and rule-driven behaviour subscale from the Compulsive Exercise Test (CET; Taranis et al., 2011) was used to measure pathological exercise. Participants were asked to indicate how true each statement was of them (e.g., “I usually continue to exercise despite injury unless I am very ill or too injured”) on a scale from 0 (never true) to 5 (always true). A total score was calculated by summing the items, with higher scores indicating higher pathological exercise symptomatology. Internal consistency was excellent (α = .91).

Procedure

Ethics approval was received from the Griffith University HREC (GU Ref No: 2020/221) before commencing data collection. Participants accessed the Qualtrics study link via a password protected research participation system only accessible to current undergraduate psychology students. Interested participants were first asked to read an information sheet describing the study, after which they completed demographic questions and the counterbalanced battery of questionnaires. Participation took approximately 30 minutes and participants received partial course credit as compensation for their time. The study was anonymous and informed consent was implied through survey completion.

Statistical analysis

Analyses were conducted using IBM SPSS Version 29.0. There was a small amount of missing data (0.63%) which, according to Little’s MCAR test was missing completely at random, χ²(142) = 162.95, p = .11. Given this, as well as the very small amount of missing data, we decided to remove this missingness using listwise deletion. The assumptions of mediated regression were tested; there were no violations of normality of residuals for the main variables. There were four univariate outliers for BMI, and one multivariate outlier, which were removed from the main analyses. A heteroskedasticity-consistent standard error estimator (HC3) was applied (Hayes & Cai, 2007), and there were no other violations of any assumptions.

To test the main hypotheses, Model 4 of the PROCESS macro (Hayes, 2022) was used to run six mediation analyses; three analyses explored whether self-kindness (M) mediated the relationships between effortless perfectionism (X) and body dissatisfaction, pathological eating, and pathological exercise (Y’s) whilst controlling for self-criticism. A further three analyses explored whether self-criticism (M) mediated the same X-Y relationships whilst controlling for self-kindness. All six mediation analyses also included age, gender (dummy coded), and BMI as additional covariates to control for the influence of these factors in our sample. Bootstrapping (n = 5000) with 95% bias-corrected confidence intervals was used to test indirect effects; a confidence interval that did not include zero provided evidence of a mediation effect. The final sample size in all mediation analyses (n = 481) provided sufficient statistical power; as indicated by the modelling of Fritz and MacKinnon (2007), a sample size of 462 enables detection of a bias-corrected bootstrapped indirect effect when both “a” and “b” paths are small (1-β = 0.80).

Results

Descriptive statistics (M and SD) and bivariate correlations for the key variables are reported in Table 1. Effortless perfectionism, self-criticism, body dissatisfaction, and pathological eating and exercise were all positively intercorrelated, and all (except for pathological exercise) were inversely associated with self-kindness. Partial correlations (controlling for age, gender, BMI, and self-kindness) between self-criticism and the key variables were consistent with bivariate associations. However, the partial correlations (controlling for age, gender, BMI, and self-criticism) between self-kindness and both effortless perfectionism and pathological eating were non-significant (r = −.06, p = .20 and r = .06, p = .17, respectively) and the partial correlation between self-kindness and pathological exercise became positive and significant (r = .13, p = .005). The partial correlation between self-kindness and body dissatisfaction reduced in size (relative to the bivariate association) but remained significant and negative (r = −.15, p < .001).

Table 1. Descriptive statistics and bivariate correlations for key variables.

Variable	M (SD)	1.	2.	3.	4.	5.
1. Effortless perfectionism	23.16 (7.73)	–
2. Self-kindness	3.29 (1.04)	−.29*	–
3. Self-criticism	23.45 (12.72)	.48*	−.50*	–
4. Body dissatisfaction	11.77 (4.48)	.28*	−.40*	.64*	–
5. Pathological eating	1.98 (1.43)	.28*	−.21*	.58*	.63*	–
6. Pathological exercise	15.32 (9.32)	.21*	.00	.20*	.19*	.33*

Mediation models for body dissatisfaction

Self-kindness as a mediator (controlling for self-criticism)

In path “a”, effortless perfectionism was not significantly associated with self-kindness, B_a = −0.01, SE = 0.01, 95% CI [−0.020, 0.004]. For path “b”, self-kindness was negatively associated with body dissatisfaction, B_b = −0.59, SE = 0.18, 95% CI [−0.936, −0.234]. Self-criticism was negatively associated with self-kindness, B_a1 = −0.04, SE = 0.004, 95% CI [−0.048, −0.032], and positively associated with body dissatisfaction, B_b1 = 0.19, SE = 0.02, 95% CI [0.161, 0.224]. In relation to the other covariates, being female (relative to male) positively predicted self-kindness, B_a4 = 0.218, SE = 0.09, 95% CI [0.042, 0.394], and BMI positively predicted body dissatisfaction, B_b3 = 0.18, SE = 0.04, 95% CI [0.100, 0.250].

The direct effect, c’, was not significant, B_c’ = −0.02, SE = 0.02, 95% CI [−0.065, 0.031], and the indirect effect revealed no evidence of mediation, B_a×b = 0.005, SE = 0.004, 95% CI [−0.003, 0.013].

Self-criticism as a mediator (controlling for self-kindness)

Effortless perfectionism was positively associated with self-criticism, B_a = 0.56, SE = 0.06, 95% CI [0.443, 0.685], and self-criticism was positively associated with body dissatisfaction, B_b = 0.19, SE = 0.02, 95% CI [0.161, 0.224]. Self-kindness was negatively associated with self-criticism, B_a1 = −4.77, SE = 0.47, 95% CI [−5.689, −3.841], and body dissatisfaction, B_b1 = −0.59, SE = 0.18, 95% CI [−0.936, −0.234]. For the remaining covariates, being female (relative to male), B_a4 = 4.19, SE = 0.97, 95% CI [2.285, 6.085], non-binary (relative to male), B_a5 = 5.74, SE = 2.54, 95% CI [0.752, 10.735], younger in age, B_a2 = −0.19, SE = 0.05, 95% CI [−0.288, −0.083], and having a higher BMI, B_a3 = 0.39, SE = 0.12, 95% CI [0.155, 0.633], all predicted higher self-criticism. Again, BMI positively predicted body dissatisfaction, B_b3 = 0.18, SE = 0.04, 95% CI [0.100, 0.250].

Effortless perfectionism was not directly associated with body dissatisfaction, B_c’ = −0.02, SE = 0.02, 95% CI [−0.065, 0.031], but was indirectly positively associated with body dissatisfaction via self-criticism, B_a×b = 0.11, SE = 0.02, 95% CI [0.080, 0.140]. This mediation model is displayed in Figure 1.

Figure 1. Self-criticism as a mediator of the relationship between effortless perfectionism and body dissatisfaction (*p < .05).

Mediation models for pathological eating

Self-kindness as a mediator (controlling for self-criticism)

Effortless perfectionism was not related to self-kindness, B_a = −0.01, SE = 0.01, 95% CI [−0.020, 0.004]. Self-kindness was not associated with pathological eating, B_b = 0.08, SE = 0.06, 95% CI [−0.036, 0.199]. The covariate, self-criticism, was negatively associated with self-kindness, B_a1 = −0.04, SE = 0.004, 95% CI [−0.048, −0.032], and positively associated with pathological eating, B_b1 = 0.06, SE = 0.01, 95% CI [0.049, 0.070]. For the other covariates, being female (relative to male) predicted self-kindness, B_a4 = 0.22, SE = 0.09, 95% CI [0.042, 0.394] and pathological eating, B_b4 = 0.66, SE = 0.11, 95% CI [0.445, 0.870]. BMI also positively predicted pathological eating, B_b3 = 0.07, SE = 0.01, 95% CI [0.048, 0.097].

There was no direct effect, B_c’ = 0.01, SE = 0.01, 95% CI [−0.007, 0.025], or indirect effect, B_a×b = −0.001, SE = 0.001, 95% CI [−0.003, 0.001].

Self-criticism as a mediator (controlling for self-kindness)

As revealed in Figure 2, there was a positive association between effortless perfectionism and self-criticism, B_a = 0.56, SE = 0.06, 95% CI [0.443, 0.685], and self-criticism was positively associated with pathological eating, B_b = 0.06, SE = 0.01, 95% CI [0.049, 0.070]. The covariate, self-kindness, was negatively associated with self-criticism, B_a1 = −4.77, SE = 0.47, 95% CI [−5.689, −3.841], but was not associated with pathological eating, B_b1 = 0.08, SE = 0.06, 95% CI [−0.036, 0.199]. The other covariates also predicted self-criticism; being female (relative to male), B_a4 = 4.19, SE = 0.97, 95% CI [2.285, 6.085], non-binary (relative to male), B_a5 = 5.74, SE = 2.54, 95% CI [0.752, 10.735], younger, B_a2 = −0.19, SE = 0.05, 95% CI [−0.288, −0.083], and having a higher BMI, B_a3 = 0.39, SE = 0.12, 95% CI [0.155, 0.633], all predicted more self-criticism. Being female (relative to male), B_b4 = 0.66, SE = 0.11, 95% CI [0.445, 0.870], and having a higher BMI, B_b3 = 0.07, SE = 0.01, 95% CI [0.048, 0.097], also positively predicted pathological eating.

Figure 2. Self-criticism as a mediator of the relationship between effortless perfectionism and pathological eating (*p < .05).

There was no direct effect, B_c’ = 0.01, SE = 0.01, 95% CI [−0.007, 0.025]. However, the indirect effect was positive and significant, B_a×b = 0.03, SE = 0.01, 95% CI [0.025, 0.043].

Mediation models for pathological exercise

Self-kindness as a mediator (controlling for self-criticism)

Effortless perfectionism was not significantly related to self-kindness, B_a = −0.01, SE = 0.01, 95% CI [−0.020, 0.004]. However, self-kindness was positively associated with pathological exercise, B_b = 1.36, SE = 0.47, 95% CI [0.429, 2.295]. The covariate, self-criticism, was negatively associated with self-kindness, B_a1 = −0.04, SE = 0.004, 95% CI [−0.048, −0.032], and positively associated with pathological exercise, B_b1 = 0.16, SE = 0.05, 95% CI [0.072, 0.253]. Being female (relative to male) positively predicted self-kindness, B_a4 = 0.22, SE = 0.09, 95% CI [0.042, 0.394], and negatively predicted pathological exercise, B_b4 = −2.10, SE = 0.91, 95% CI [−3.888, −0.312].

The direct effect between effortless perfectionism and pathological exercise was positive and significant, B_c’ = 0.15, SE = 0.06, 95% CI [0.034, 0.272]. The indirect was not significant, B_a×b = −0.01, SE = 0.01, 95% CI [−0.033, 0.005].

Self-criticism as a mediator (controlling for self-kindness)

As Figure 3 shows, effortless perfectionism was positively related to self-criticism, B_a = 0.56, SE = 0.06, 95% CI [0.443, 0.685], and self-criticism was positively related to pathological exercise, B_b = 0.16, SE = 0.05, 95% CI [0.072, 0.253]. The covariate, self-kindness, was negatively associated with self-criticism, B_a1 = −4.77, SE = 0.47, 95% CI [−5.689, −3.841], and positively associated with pathological exercise, B_b1 = 1.36, SE = 0.47, 95% CI [0.429, 2.295]. All of the remaining covariates predicted self-criticism; being female (relative to male), B_a4 = 4.19, SE = 0.97, 95% CI [2.285, 6.085], non-binary (relative to male), B_a5 = 5.74, SE = 2.54, 95% CI [0.752, 10.735], younger, B_a2 = −0.19, SE = 0.05, 95% CI [−0.288, −0.083], and higher in BMI, B_a3 = 0.39, SE = 0.12, 95% CI [0.155, 0.633], were associated with higher self-criticism. Being female (relative to male) also negatively predicted pathological exercise, B_b4 = −2.10, SE = 0.91, 95% CI [−3.888, −0.312].

Figure 3. Self-criticism as a mediator of the relationship between effortless perfectionism and pathological exercise (*p < .05).

There was a positive direct effect between effortless perfectionism and pathological exercise, B_c’ = 0.15, SE = 0.06, 95% CI [0.034, 0.272], and as well as a positive indirect effect via self-criticism, B_a×b = 0.09, SE = 0.03, 95% CI [0.040, 0.149].

Discussion

The current study explored the relationships between effortless perfectionism and body dissatisfaction, pathological eating, and pathological exercise, as well as the unique mediating roles of self-kindness and self-criticism. As anticipated, there was a moderate-sized negative bivariate correlation between self-criticism and self-kindness. For self-criticism, the bivariate and partial correlations (controlling for self-kindness) were consistent and as anticipated; self-criticism was positively correlated with effortless perfectionism and all three outcomes. However, after controlling for shared variance with self-criticism, self-kindness was not significantly associated with effortless perfectionism or pathological eating, and there was a positive (albeit small) correlation between self-kindness and pathological exercise. The negative partial correlation between self-kindness and body dissatisfaction remained significant and negative but reduced in size relative to the bivariate correlation. In relation to the mediation hypotheses, contrary to expectations, self-kindness (controlling for self-criticism) did not mediate the relationships between effortless perfectionism and any of the three criterion variables. However, as hypothesised, self-criticism (controlling for self-kindness) was found to positively mediate the relationships between effortless perfectionism and body dissatisfaction, pathological eating, and pathological exercise.

Findings support existing evidence linking negative perfectionism with body dissatisfaction, and pathological eating/exercise (González-Hernández et al., 2022; Hamedani et al., 2023). However, our investigation has provided new evidence regarding effortless perfectionism, an underexplored form of negative perfectionism (Flett et al., 2016). In the mediation models, which accounted for the influence of self-kindness and self-criticism (along with the additional covariates age, gender, and BMI), effortless perfectionism was positively and directly associated only with pathological exercise. This suggests a uniqueness about pathological exercise in the desire for effortless perfection. Here we measured pathological exercise as a compulsion to strictly follow exercise rules (i.e., continue to exercise even when sick, injured, or damaging to health) to avoid experiencing negative affect (i.e., guilt) from not exercising (Taranis et al., 2011). A stronger desire for “natural” perfection in the context of exercise may perpetuate obsessive exercise-related thoughts, more intense negative affectivity in response to a missed exercise session, and a stronger compulsion to adhere to a strict exercise regime (Bratland-Sanda et al., 2019). If an exercise session is missed, one must work even harder (i.e., expend more effort) to maintain or regain one’s fitness level or physical appearance, a scenario an effortless perfectionist would be compelled to avoid (Ruiz-Turrero et al., 2022; Taranis et al., 2011).

As hypothesised, effortless perfectionism was also indirectly positively associated with all three outcomes. This indirect association occurred via self-criticism, rather than via self-kindness. This finding supports evidence that self-criticism is intricately linked with unrealistic perfectionistic standards, and that self-criticism is the unequivocal outcome of perfectionistic desire whether one falls short of expectations (in which case one has outright failed), or whether a goal is achieved (in which case the goal will be judged as not having been challenging enough) (Cuesta-Zamora et al., 2022; Dunkley et al., 2006; Hamedani et al., 2023; Williams & Levinson, 2022). Our findings provide unique evidence that this association between negative forms of perfectionism and self-criticism holds true for effortless perfectionism, whereby criticism of the self is likely related to the requirement to expend effort in the pursuit of perfection and the belief that one should not need to try to be perfect (Hamedani et al., 2023). The higher the bar for perfection, the higher the requirement for effort to achieve the standard, thus perpetuating the positive link between effortless perfection and self-criticism.

By contrast, self-kindness was not correlated with effortless perfectionism, nor did self-kindness mediate the relationships between effortless perfectionism and body dissatisfaction and pathological eating/exercise. Our findings introduce the possibility that self-kindness may not be uniquely protective against pathological eating/exercise; previously reported associations between self-compassion and reduced pathology may be partially explained by the variance self-kindness shares with self-criticism (Muris et al., 2019). When interpreting these results, it is also important to consider what the self-kindness construct represents after having accounted for shared variance with self-criticism (Lynam et al., 2006). Criticism-free self-kindness may not temper aspirations for perfection. Self-compassion has been linked with high standards, accountability, and personal responsibility (Neff, 2003), and self-kindness (controlling for self-criticism) was not found to covary with effortless perfectionism in the current study. Instead, self-kindness (keeping self-criticism stable) may manifest as support for having done our best in the pursuit of goals and understanding towards ourselves for not having yet reached performance standards (and for individuals higher in perfectionism, these standards may be relentlessly demanding). In this way, higher self-kindness may encourage personal accountability and further goal striving. Consistent with previous research revealing that strict and inflexible exercise habits are more frequent among males (Murray et al., 2017), we found that being female (relative to male) negatively predicted pathological exercise. In addition, our finding of a weak positive association between self-kindness and pathological exercise further supports this suggestion that (at least in a non-clinical sample and whilst controlling for self-criticism), self-kindness may predict a more steadfast commitment to exercise.

Theoretically, our findings are consistent with the idea that individuals self-relate in complex ways depending on context and can experience “multiple inner voices”, sometimes kind and sometimes harsh (Ferrari et al., 2022, p. 1653). The results also provide further evidence that self-kindness does not perfectly covary with self-criticism and that it can function as an independent construct in the presence, or absence, of high levels of self-criticism (Ferrari et al., 2022; Halamová & Kanovský, 2019). Our findings also support existing literature demonstrating that the negative aspects of self-compassion (i.e., judgement and criticism) have clustered separately to the positive aspects (i.e., understanding and kindness) in factor analyses, and that the negative components more strongly predict psychopathology (Ferrari et al., 2022). Specifically, our results have provided complementary evidence for the pathological nature of self-criticism (Cuesta-Zamora et al., 2022; Dunkley et al., 2006; Hamedani et al., 2023; Krieger et al., 2019; Williams & Levinson, 2022); self-criticism perpetuated the link between effortless perfectionism and pathological body-related and eating/exercise outcomes. By contrast, self-kindness was weakly negatively associated with body dissatisfaction but did not emerge as a negative predictor of pathological eating/exercise, nor did it mediate their association with effortless perfectionism.

In addition to theoretical contribution, the findings have potential practical implications. Results indicate that self-criticism may be a possible target for future intervention research (Williams & Levinson, 2022). Interestingly, boosting self-compassion has been identified as a counteractive method through which to “defend oneself against one’s own self-attacks” (Gilbert et al., 2004, p. 33; Wakelin et al., 2022; Werner et al., 2019). The suggestion is that stimulating self-compassion (i.e., by increasing the salience of feelings of self-reassurance, warmth, and safety) prevents activation of the threat-based system and blocks self-critical thoughts (Ferrari et al., 2022; Werner et al., 2019). However, importantly, self-kindness may demonstrate its impact when implemented in-the-moment in direct response to a personally relevant evaluative event, but this possibility was not directly explored in the current trait-based investigation (Ferrari et al., 2022). There are also alternative methods that may reduce self-criticism including chair work in Gestalt Therapy, mindfulness-based approaches, and emotion-focused approaches (Gilbert et al., 2004; Halamová & Kanovský, 2019; Krieger et al., 2019). Transdiagnostic interventions, such as Enhanced Cognitive Behavioural Therapy (CBT-E), have also proven effective in addressing clusters of interrelated pathological symptomatology, including among young people (Dalle Grave et al., 2023); CBT-E may be relevant in this context given it was designed originally to target pathological eating, and directly addresses the underlying processes known to contribute to such pathology including higher negative perfectionism and self-criticism (Dalle Grave, 2023; Hamedani et al., 2023).

Our investigation has some limitations. In relation to measurement, although it has been argued that effortless perfectionism is distinguishable from trait perfectionism and other forms of negative perfectionism (Travers et al., 2015), the measurement of this construct and the face validity of the EPS has been queried (Flett et al., 2016). Further clarity around the independence of the effortless perfectionism construct, and its unique consequences, remains important in future investigations. Individuals across the spectrum of exercise frequency have also differed in interpretation of the CET items, our measure of pathological exercise (Alcaraz-Ibáñez et al., 2020). For a typical (or infrequent) exerciser, disappointment in oneself for missing an exercise session or attempting to make up for a missed exercise session may be quite reasonable; a degree of commitment and accountability is required to stick to an exercise regime. However, this can make it difficult to delineate healthy from problematic exercise behaviour, and to distinguish when, and for whom, certain exercise behaviours are pathological (Chamberlain & Grant, 2020). Future investigations should incorporate an assessment of exercise level to account for this as a potential moderating factor and to further clarify the positive link we observed between self-kindness and pathological exercise.

Although eating disorders are more prevalent in Western countries and among young females (Dalle Grave et al., 2023; Qian et al., 2022), our predominantly female undergraduate student sample may limit the generalisability of our results. Our data support earlier evidence that eating disorder presentations can differ between males and females (Murray et al., 2017), and our use of a non-clinical sample means that the patterns observed could differ among individuals with a diagnosed eating disorder. Cross-sectional analyses also limit our capacity to make conclusions regarding the directionality of relationships. As well as further investigating gender differences in symptomatology, experimental research exploring in-the-moment self-kindness, the efficacy of chair work, or measuring self-kindness and self-criticism as mediators of change in transdiagnostic approaches such as CBT-E remain useful avenues for further research; these approaches are valid to explore as both a treatment approach among individuals with severe symptomatology and as a preventative approach among at-risk individuals (Dalle Grave, 2023; Dalle Grave et al., 2023; Ferrari et al., 2022; Halamová & Kanovský, 2019; Hamedani et al., 2023; Krieger et al., 2019). Importantly, given that pathological eating and exercise behaviours typically first emerge at a young age, CBT-E has demonstrated efficacy among young people (14–25 years of age) with anorexia nervosa (Dalle Grave et al., 2023). Longitudinal research would also be valuable to investigate whether self-kindness and self-criticism uniquely predict various outcomes over time, and to test for longer-term reciprocal relationships between these constructs (Ferrari et al., 2022; Linardon, 2022).

In conclusion, given the pervasive and damaging impact of body dissatisfaction and pathological eating/exercise, we have endeavoured to provide further insight into the correlates of worsened or improved symptomatology. Effortless perfectionism has emerged as a direct positive predictor of heightened pathological exercise. Effortless perfectionism also indirectly predicted worsened symptoms for all three outcomes via self-criticism as a mediating mechanism. These results have further emphasised the importance of disentangling the unique effects of self-criticism and self-kindness in future research. Although self-kindness (uniquely from self-criticism) did not emerge as a mediating mechanism, future research is important to determine whether this was due to the cross-sectional and trait-based measurement approach adopted in the current study. Longitudinal research, further disentangling gender disparities in presentation, and exploring the efficacy of transdiagnostic interventions which target combinations of perpetuating processes (i.e., negative perfectionism and self-criticism), are also considerations to advance research in this important area.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The authors do not have permission to share data.