The role of clinical perfectionism and psychological flexibility in distress and wellbeing

ABSTRACT

Objectives

Clinical perfectionism, characterised by stringent standards and persistent striving, correlates with self-criticism and poor wellbeing. In this study we explored the mediating roles of psychological flexibility and inflexibility, and self-compassion, in the relationships of clinical perfectionism with distress, and clinical perfectionism and wellbeing.

Methods

We used a cross-sectional correlational study design. A non-clinical sample of 210 Australian residents aged between 18 to 65 years old completed an online survey. Data were analysed using correlational, multiple regression, and parallel mediation analysis.

Results

Heightened clinical perfectionism was associated with lower wellbeing and increased distress. Regression models revealed significant mediation by psychological inflexibility processes (self-as-content, experiential avoidance, cognitive fusion) and self-compassion. Cognitive fusion and inaction mediated clinical perfectionism-distress, while self-compassion and inaction mediated clinical perfectionism-psychological wellbeing.

Conclusions

These findings support clinical perfectionism as a transdiagnostic predictor of distress and reduced wellbeing. Those with high clinical perfectionism tend to rigidly respond to perfectionistic thoughts, avoid uncomfortable emotions, and engage in value-disconnected inaction. Associations between psychological flexibility and inflexibility processes emphasise their multidimensional nature, distinct yet interrelated. Inaction emerges as a common process in clinical perfectionism-distress and clinical perfectionism-wellbeing, while cognitive fusion specifically influences distress, and limited self-compassion affects wellbeing.

Key Points

What is already known about this topic:

1. Clinical perfectionism is linked to psychopathology and transdiagnostic mechanisms, impacting distress and psychological wellbeing.

2. Psychological inflexibility provides a transdiagnostic framework for understanding how rigid and self-limiting responses to feelings and thoughts are maintained despite negative consequences.

3. Clinical perfectionism may be a form of psychological inflexibility, with rigid responses to perfectionistic thoughts, avoidance of uncomfortable emotions, and value-disconnected inaction.

What this topic adds:

1. Cognitive fusion and inaction are crucial for understanding clinical perfectionism’s impact on distress.

2. Self-compassion and inaction play a key role in understanding the relationship between clinical perfectionism and wellbeing.

3. Understanding the functions of inaction is vital for fostering flexible responses to perfectionistic thoughts.

KEYWORDS:

• Perfectionism
• self-compassion
• psychological flexibility
• wellbeing

Perfectionism

Perfectionism is traditionally understood as a personality characteristic prevalent across the lifespan in the general population (Fink, 2020). It is associated with striving for excellence and flawlessness, but can also involve feelings of shame and failure (Afshar et al., 2011). Definitions of perfectionism commonly describe a combination of setting exceptionally high personal standards and excessive self-criticism (Burns, 1980; Flett & Hewitt, 2002; Frost et al., 1990). Perfectionism can impact many life domains, including interpersonal relationships, occupational and academic functioning, and psychological wellbeing (i.e., the combination of positive affect and optimal functioning in individual and social life; Deci & Ryan, 2008; Moroz & Dunkley, 2019; Tennant et al., 2007). Furthermore, perfectionism is implicated in significant clinical problems, including psychological distress (Egan et al., 2011). Given the prevalence and impact of perfectionism, it is important to understand what influences its expression. Research suggests perfectionism is a complex behaviour combining self-appraisal, goal motivations, and rigid behavioural patterns, existing on a continuum (Broman-Fulks et al., 2008).

Shafran et al. (2003) proposed clinical perfectionism as a construct to better identify an extreme form of perfectionism associated with greater specificity to psychopathology and transdiagnostic mechanisms. Clinical perfectionism involves combinations of excessive self-criticism and rigid striving towards unrealistic, self-imposed, and personally demanding standards (Shafran et al., 2002). A defining feature of clinical perfectionism that distinguishes it from perfectionism in general is the continual striving towards unrealistic standards even when adverse consequences are experienced (Shafran et al., 2002). Shafran et al. (2003) presented clinical perfectionism from a transdiagnostic perspective, arguing that focussing on specific, self-oriented mechanisms maintaining clinical perfectionism will facilitate successful treatment of clinical perfectionism-related psychopathology (e.g., self-harm, suicidality, and distress; Egan et al., 2011; Shafran et al., 2003).

Subsequent research has evaluated a theory-driven, cognitive-behavioural intervention for clinical perfectionism (CBT for perfectionism: Shafran et al., 2023), that seeks to address dysfunctional beliefs related to personal standards, self-criticism, and cognitive biases through cognitive restructuring and behavioural experiments, and influence associated behaviours of checking, avoidance and procrastination. CBT for perfectionism has demonstrated changes in perfectionism through treatment and significant effects for outcomes (eating disorder symptoms, depression, anxiety: for a meta-analytic review see Galloway et al., 2022). Recent work in the development of CBT for perfectionism has included self-compassion to target self-criticism (Shafran et al., 2018).

Research has indicated that broadening one’s self-evaluations and fostering greater flexibility in self-evaluations, such as through processes of self-compassion and self-acceptance, may be a component of treatment for clinical perfectionism and perfectionism-related psychopathology (Adams et al., 2023; Ong, Barney, et al., 2019). Self-compassion has been demonstrated to have a protective effect on the impacts of perfectionism (e.g., Ferrari et al., 2018; Fletcher et al., 2019). In a cross-sectional study Adams et al. (2023) observed a moderating role for self-compassion for the association between clinical perfectionism and psychological distress for adults with mild symptoms of anxiety and depression. Specifically, greater self-compassion corresponded to lower correlations between clinical perfectionism and psychological distress. Ong, Barney, et al. (2019) in a study of mediators and moderators of acceptance and commitment therapy (ACT) for clinical perfectionism observed that self-compassion mediated the intervention effect for concern over mistakes; baseline self-compassion also predicted better outcomes for ACT. In the present study we measured self-oriented mechanisms, specifically self-as-content and self-compassion, to inform hypotheses about relationships with related constructs such as distress, wellbeing (positive mental health: Tennant et al., 2007) and psychological flexibility and inflexibility.

Perfectionism, distress, and psychological wellbeing

Perfectionism has been observed as a transdiagnostic mechanism across many psychopathologies (Limburg et al., 2017). Anxious and depressive symptomatology is often characteristic of distress associated with perfectionism (Ridner, 2004). Perfectionism may also be associated with wellbeing, as evidenced by research conducted by Kanten and Yesıltas (2015). While these findings support relationships between perfectionism and general states of wellbeing, fewer studies have investigated the relationship of perfectionism with positive wellbeing states (e.g., optimism and positive affect; Lyubomirsky et al., 2005; Ryan & Deci, 2001). As Suh et al. (2017) have found associations with adaptive perfectionism and positive wellbeing states (Kamushadze et al., 2021; Suh et al., 2017) the current study explored positive wellbeing alongside distress (consistent with Winefield et al., 2012).

Perfectionism and psychological flexibility/inflexibility

The psychological flexibility model (Hayes et al., 2012) identifies a set of transdiagnostic processes theorised to promote wellbeing and personal effectiveness. Psychological flexibility is the capacity to consciously engage with the present moment, effectively adjust behaviours and cognitive strategies to suit situational demands, and pursue personally meaningful values despite experiencing distress (Kashdan & Rottenberg, 2010). Hayes et al. (2012) describe six interrelated processes promoting psychological flexibility – acceptance, cognitive defusion, self-as-context, committed action, values, and present moment awareness (Rolffs et al., 2016).

The psychological flexibility model also implicates psychological inflexibility processes of rigid responding to unwanted internal experiences (i.e., cognitions, emotions, or bodily sensations) at personal cost (Hayes et al., 2006), associated with poorer functioning and distress across various psychopathologies and problems (Levin et al., 2014). This inflexibility is engendered by processes including experiential avoidance (EA), cognitive fusion (CF), self-as-content, inaction, and lack of contact with values and the present moment (Rolffs et al., 2016).

Reviews suggest psychological flexibility and inflexibility as multidimensional, separate-yet-related constructs (Cherry et al., 2021; Tyndall et al., 2020). Ciarrochi et al. (2014) have observed that people can simultaneously demonstrate pursuit of goals and values despite high CF, suggesting that individuals may differ across psychological flexibility dimensions. Psychological flexibility and inflexibility seem related to perfectionism: it can be observed that some perfectionists can show greater flexibility in adapting emotional and behavioural responses to suit diverse situations (Clark, 2019). However, there are also forms of perfectionism that seem consistent with psychological inflexibility: responding to strivings with high self-criticism, control and experiential avoidance, resulting in a rigidity in emotional and behavioural responses employed to manage different situations (Dunkley et al., 2000; Moroz & Dunkley, 2015; Nazarzadeh et al., 2015; Santanello & Gardner, 2007). As an observation it seems that defining features of clinical perfectionism, such as rigid pursuits of achievement to avoid unwanted feelings (i.e., EA) despite this reducing engagement in personally meaningful activities (i.e., inaction) are akin to psychological inflexibility processes (Clark, 2019; Moroz & Dunkley, 2015). Other defining features of clinical perfectionism, such as preoccupations with self-critical thinking, rules around striving and self-worth (i.e., CF) and rigid self-identifications fused with the content of life experiences, failure, and negative self-evaluations (i.e., self-as-content) may result in individuals with high levels of perfectionism disconnecting from intentional contact with the present moment (i.e., lack of contact with the present moment) or from personal qualities (i.e., values) they find important (Clark, 2019; Shafran et al., 2002). This reinforces the association between clinical perfectionism and psychological inflexibility processes, highlighting the importance of further understanding this relationship.

In both circumstances, this may limit the individual’s contact with other sources of influence on their behaviours, including the consequences or unworkability of their actions, maintaining their perfectionistic tendencies (e.g., Clark, 2019; Gilman et al., 2014). These forms of psychological inflexibility may impair wellbeing (Crouse et al., 2020): understanding relationships between psychological flexibility, psychological inflexibility and wellbeing may elucidate the effects of enhancing psychological flexibility and reducing psychological inflexibility on wellbeing. Due to correlations between perfectionism and wellbeing, understanding these effects can highlight how changes in wellbeing can impact clinical perfectionism. Limited research has investigated the impact of psychological inflexibility processes: inaction (i.e., patterns of behaviour motivated by avoidance and disconnected from personal values; Rolffs et al., 2016), lack of contact with values (i.e., disconnection from chosen life directions; Rolffs et al., 2016) and lack of present moment awareness (i.e., limited mindfulness) on the clinical perfectionism-distress and clinical perfectionism-psychological wellbeing relationships.

The psychological flexibility model also considers self-oriented mechanisms, including self-as-context and self-compassion, which promote flexible observing of, and responding to, self-critical thoughts with acceptance and openness (Przezdziecki & Sherman, 2016). Self-as-context involves viewing one’s inner experiences from a detached observer’s perspective, while self-compassion involves self-kindness, common humanity, and mindfulness (Clark, 2019; Neff & Germer, 2013). There are indications these self-processes may be important in perfectionism: individual’s with maladaptive levels of perfectionism engage with self-as-content through rigid self-identifications with thoughts and life experiences involving negative self-views, self-criticism, personal failure and ideas that they cannot achieve expected standards (Moroz & Dunkley, 2015). Lo and Abbott (2019) found that those with adaptive levels of perfectionism held more positive and less negative beliefs about their personality compared to those distressed and struggling with perfectionism. Self-compassion may also mediate the perfectionism-distress relationship, with Mehr and Adams (2016) finding self-compassion partially mediated associations between maladaptive perfectionism and distress (operationalised as depressive symptomatology). Furthermore, Abdollahi et al. (2020) found that self-compassion moderated the perfectionism-distress relationship.

Based on the literature explored thus far, it appears that past research has elucidated relationships between adaptive and maladaptive perfectionism, distress, and wellbeing, and has examined whether self-compassion and psychological inflexibility processes (e.g., EA, CF, and self-as-content) mediate these relationships. However, limited research has investigated these relationships in the context of clinical perfectionism (Galloway et al., 2022).

The present study

This study investigated associations between clinical perfectionism, wellbeing, and distress (operationalised as stress, depression, and anxiety symptoms). As discussed, individuals with maladaptive levels of perfectionism may demonstrate high EA, CF, and self-as-content. Since the maladaptive perfectionism literature suggests associations exist with these psychological inflexibility processes, the current study established whether these processes were observed with clinical perfectionism in a non-clinical sample and investigated their influence on clinical perfectionism-distress and clinical perfectionism-wellbeing relationships. This study also investigated whether self-compassion would mediate these two relationships.

Aims and hypotheses

This study examined the relationships between clinical perfectionism, distress, and wellbeing, and investigate the effects of CF, self-as-content, and EA on these relationships.

It was hypothesised:

1. Greater clinical perfectionism would be associated with greater distress and lower wellbeing;

2. psychological inflexibility processes (self-as-content, CF, and EA) and self-compassion would mediate the clinical perfectionism-distress and clinical perfectionism-wellbeing relationships.

We also investigated whether the remaining three psychological inflexibility components – inaction, lack of present moment awareness, and lack of contact with values - demonstrated influence on the relationship between clinical perfectionism and distress or clinical perfectionism and psychological wellbeing. No hypotheses were made due to exploratory nature of this investigation.

Design

This study used a cross-sectional design with participants completing an anonymous online survey.

Method

Participants

Participants included 210 volunteers aged 18–65 years, residing in Australia and fluent in English, recruited through advertising on social media and university student networks.

Materials

Demographic Questionnaire

Participants were asked about their age, gender, English fluency, highest education attainment, employment status, and whether they considered themselves part of a minority regarding sexuality, gender expression, ethnicity, or religion.

The Clinical Perfectionism Questionnaire (CPQ)

The CPQ (Fairburn et al., 2003) is a 12-item scale measuring clinical perfectionism on a four-point Likert-scale. Participants responded to items based on their feelings and behaviours over the past month, with higher scores indicating greater clinical perfectionism (Dickie et al., 2012). Egan et al. (2016) support the CPQ’s discriminative validity and reliability (α = 0.71–0.83). Two items of the CPQ are reverse scored, which research has found to be problematic since this can adversely impact psychometric properties due to respondent confusion over double negative, response biases, and other sources of method variance (Prior et al., 2018). These two items were thus not included in this questionnaire.

The Depression Anxiety Stress Scale (DASS-21)

The DASS-21 (Lovibond & Lovibond, 1995) is a 21-item questionnaire assessing distress by measuring depression, anxiety, and stress experienced over the past week. Each seven-item subscale uses a 4-point Likert-scale, with higher scores indicating greater depression, anxiety, or stress (González-Rivera et al., 2019). Lovibond and Lovibond (1995) support validity and reliability of the DASS-21 (α for depression, anxiety, and stress being 0.91, 0.84, and 0.90, respectively). This study used the DASS-21 total score as the distress measure.

The Multidimensional Psychological Flexibility Inventory (MPFI)

The MPFI (Rolffs et al., 2016) is a 60-item questionnaire using a 6-point Likert-scale. It consists of six subscales (i.e., Acceptance, Present Moment Awareness, Self-as-context, Defusion, Values, and Committed Action) that measure psychological flexibility processes and six subscales (i.e., Experiential Avoidance, Lack of Contact with the Present Moment, Self-as-content, Fusion, Lack of Contact with Values, and Inaction) that measure psychological inflexibility processes. The MPFI has demonstrated strong reliability and discriminant validity (α = 0.84–0.96; Rogge et al., 2019; Rolffs et al., 2016). This study used all six psychological inflexibility subscales to test hypotheses and for the exploratory investigations presented above.

The Self-Compassion Scale-Short Form (SCSSF)

The SCSSF (Raes et al., 2011) is a 12-item scale measuring self-compassion demonstrated during difficult situations on a 5-point Likert-scale. Raes et al. (2011) support the validity and reliability of the SCSSF (α ≥ 0.86).

The Short Warwick-Edinburgh Mental Well-being Scale (SWEMWBS)

The SWEMWBS (Stewart-Brown et al., 2009) measures positive aspects of mental health experienced over the past two weeks through 7 items, each rated on a 5-point Likert-scale. Higher scores indicate greater wellbeing. Vaingankar et al. (2017) support the validity and reliability of the SWEMWBS (α = .90).

Procedure

This study was approved by the La Trobe University Human Research Ethics Committee (HEC21118) and was advertised through social media and university student networks. Australian residents fluent in English and aged between 18–65 years old were invited to participate. Participants consenting completed a 30-minute anonymous, online survey on the QuestionPro platform. On survey completion participants were invited to register for a chance to win one of seven, $A50.00 gift vouchers.

Planned data analysis

It was planned to conduct correlation, multiple regression, and parallel mediation analyses (using the PROCESS macro model number 4; A. F. Hayes, 2018) to test the study hypotheses. The study’s sample size was informed by Fritz and MacKinnon’s (2007) recommendations for percentile bootstrap mediation analysis: based on this a sample of 162 was estimated as needed to detect if the mediated effects hypothesised were present, with another 30% added to account for incomplete or spurious responding, for a final target sample size of 210.

Results

Data preparation

Data cleaning

Although 250 volunteers completed the online survey, 40 invalid responders were identified (speed of responding, failed attention checks) and excluded from the analyses, with a final sample size of 210. Assumptions of normality, linearity, and homoscedasticity were met.

Descriptive statistics and scale reliabilities

Table 1 presents mean (M) and SD (SD) values, and internal consistency (Cronbach’s α) coefficients for the study measures.

Table 1. Mean (M) and Standard Deviation (SD) values, and Cronbach’s α coefficients for the study measures.

Scale	M	SD	Cronbach’s α
DASS-21 Depression	6.79	5.46	.92
DASS-21 Anxiety	4.12	3.50	.80
DASS-21 Stress	9.03	4.90	.89
DASS-21 Total	19.94	11.85	.93
MPFI Experiential Avoidance	17.41	6.13	.94
MPFI Lack of Contact with Present Moment	14.45	5.68	.96
MPFI Self as Content	14.71	6.53	.94
MPFI Fusion	16.90	6.77	.95
MPFI Lack of Contact with Values	13.42	5.82	.94
MPFI Inaction	14.31	6.42	.95
MPFI Inflexibility Total	91.20	27.32	.96
SCSSF	2.83	.86	.91
SWEMWBS	22.20	5.01	.86
CPQ	23.24	6.08	.87

N=210. DASS-21=Distress, MPFI=Psychological Flexibility and Inflexibility, SCSSF=Self-compassion, SWEMWBS=Psychological Wellbeing, CPQ=Clinical Perfectionism.

Demographic characteristics

Demographic information is summarised in Table 2. This sample of 210 participants (M age = 37.2, SD = 13.4) comprised of 41 males (19.5%), 164 females (78.1%) and 5 non-binary individuals (2.4%). The sample was highly educated (68.8% having university education); most participants described being in employment or education (83.8%); nearly one third of the sample described themselves as part of a minority (31%). Analyses of differences in distress and wellbeing for these demographics are reported in the Supplementary Material.

Table 2. Demographic information for the sample in this present study.

Highest level of education	Number (Frequency, %)
10^th grade	2 (1)
11^th − 12^th grade	6 (2.9)
High school graduate	14 (6.7)
TAFE qualification/trade certificate	43 (20.5)
Bachelor degree	95 (45.2)
Masters degree	43 (20.5)
PhD or other doctorate degree	7 (3.3)
Current employment status
Full-time	61 (29)
Part-time	38 (18.1)
Casual	26 (12.4)
Self-employed	12 (5.7)
Student	39 (18.6)
Homemaker	3 (1.4)
Retired	6 (2.9)
Unemployed, looking for work	6 (2.9)
Unemployed, not looking for work	9 (4.3)
Other	10 (4.8)
Minority
Yes	65 (31)
No	145 (69)

Hypothesis testing

Relationship between variables

Correlations between clinical perfectionism, distress, wellbeing, self-compassion, and processes of psychological inflexibility appear in Table 3

Table 3. Pearson product-moment correlation coefficients between variables.

	1.	2.	3.	4.	5.	6.	7.	8.	9.
1. Clinical perfectionism	–
2. DASS-21	.55**	–
3. SWEMWBS	−.46**	−.66**	–
4. SCSSF	−.60**	−.58**	.67**	–
5. MPFI_EA	.33**	.26**	−.10	−.20**	–
6. MPFI_LOPMA	.20**	.38**	−.37**	−.33**	.21**	–
7. MPFI_SAContent	.49**	.62**	−.53**	−.65**	.33**	.32**	–
8. MPFI_CF	.47**	.71**	−.58**	−.71**	.32**	.33**	.73**	–
9. MPFI_LCV	.40**	.51**	−.49**	−.44**	.14*	.47**	.49**	.45**	–
10. MPFI_Inaction	.50**	.74**	−.59**	−.60**	.26**	.39**	.67**	.77**	.63**

**p < 0.01 (2-tailed).

*p <0.05 (2-tailed).

DASS-21=Distress, SCSSF=Self-compassion, SWEMWBS=Psychological Wellbeing, MPFI_EA = Experiential Avoidance, MPFI_LOPMA = Lack of Present Moment Awareness, MPFI_SAContent = Self-as-content, MPFI_CF = Cognitive Fusion, MPFI_LCV = Lack of Contact with Values.

It can be observed from Table 3 that clinical perfectionism demonstrated significant moderate associations with distress and wellbeing in the expected directions. There were moderate associations between self-compassion with clinical perfectionism, distress and wellbeing. Finally, the psychological inflexibility variables all showed significant, expected direction associations with the other study variables, aside from experiential avoidance and wellbeing (non-significant association). While the psychological inflexibility variables showed significant associations with each other the pattern ranged from weak to strong correlations.

Multiple regression analyses

Two MRAs determined variance levels in wellbeing and distress accountable by clinical perfectionism, EA, self-as-content, CF, and inaction. Inaction was included in the regression and mediation analyses given its strong correlation with clinical perfectionism, wellbeing, and distress.

Predictors of psychological wellbeing

Table 4 displays unstandardised (B) and standardised (β) regression coefficients, squared semi-partial correlations (sr²), and t values for each significant predictor in the MRA model. In predicting wellbeing, inaction, EA, self-as-content, clinical perfectionism, and CF significantly accounted for 43.7% of the model variance, R² = .437, adjusted R² = .43, F (5, 244) = 37.83, p < .001. Inaction, EA, self-as-content, clinical perfectionism, and CF uniquely contributed 35.3%, 18%, and 24.8% to the model, respectively.

Table 4. Regression model predicting psychological wellbeing.

Variable	B [95% CI]	β	sr^2	t
CPQ	−.17 [−.26, −.07]	−.20	.03	−3.49**
MPFI_CF	−.18 [−.30, −.05]	−.24	.02	−2.83**
MPFI_Inaction	−.19 [−.31, −.07]	−.24	.02	−3.07**
MPFI_EA	.12 [.04, .21]	.15	.02	2.88**
MPFI_SAContent	−.12 [−.23, −.00]	−.15	.01	−2.02**

N = 201. CPQ = Clinical Perfectionism, MPFI_CF = Cognitive Fusion, MPFI_EA = Experiential Avoidance, MPFI_SAContent = Self-as-content, CI = Confidence Interval, **p ≤ .001.

Predictors of distress

Table 5 displays unstandardised (B) and standardised (β) regression coefficients, squared semi-partial correlations (sr²), and t values for each significant predictor in the MRA model. It was found that inaction, clinical perfectionism, and CF significantly accounted for 79.7% of the variability in distress, R² = .797, adjusted R² = .64, F (5, 244) = 84.81, p < .001. Inaction, clinical perfectionism, and CF uniquely contributed 37.1%, 19.5%, and 34.6% to the model, respectively. In this model EA and self-as-content were non-significant predictors of distress.

Table 5. Regression model predicting distress.

Variable	B [95% CI]	β	sr^2	t
CPQ	.40 [.22, .58]	.20	.03	4.32**
MPFI_CF	.48 [.24, .71]	.27	.02	4.03**
MPFI_Inaction	.72 [.49, .95]	.39	.06	6.12**

N = 210. CPQ = Clinical Perfectionism, MPFI_CF = Cognitive Fusion, CI = Confidence Interval, **p ≤ .001.

Parallel mediation analyses

Using 5000 bootstrapped samples, two parallel mediation analyses determined whether CF, inaction, self-as-content, self-compassion, and EA had significant parallel roles in mediating the clinical perfectionism-distress and clinical perfectionism-wellbeing relationships.

The effect of clinical perfectionism on distress: mediating variables

The first parallel mediation analysis revealed that together, clinical perfectionism, CF, inaction, self-as-content, self-compassion, and EA significantly accounted for 63.5% of the variability in distress R² = .635, p < .001. A significant direct effect was found between clinical perfectionism and distress (b = .42, 95%CI [.22, .62], p < .001. There were significant paths between CF and distress (b = .51, 95%CI [.26, .76], p < .001), and inaction and distress (b = .72, 95%CI [.48, .95], p < .001). Non-significant paths were found between self-as-content and distress (b = .15, 95%CI [−.07, .37], p = .18), EA and distress (b = −.05, 95%CI [−.21, .11], p = .55), and self-compassion and distress (b = .57, 95%CI [−1.13, 2.28], p = .51). These findings show that clinical perfectionism, CF, and inaction were significant predictors of distress, whereas self-as-content, EA, and self-compassion were not. The analysis found significant indirect effects of clinical perfectionism on distress through CF (b = .26, 95%CI[.12, .43]) and inaction (b = .38, 95%CI [.23, .55]), as well as non-significant indirect effects of clinical perfectionism on distress through self-as-content (b = .08, 95%CI [−.04, .22]), EA (b = −.02, 95%CI [−.08, .04]), and self-compassion (b = −.05, 95%CI [−.21, .09]). Figure 1 displays a diagram of the parallel mediation analysis model.

Figure 1. A statistical diagram of the parallel mediator model for the effect of clinical perfectionism on distress.

Note. ** = statistically significant pathway at the .001 level

The effect of clinical perfectionism on psychological wellbeing: mediating variables

The second parallel mediation analysis revealed that together, clinical perfectionism, CF, inaction, self-as-content, self-compassion, and EA significantly accounted for 51.2% of the variability in wellbeing R²= .512, p < .001. The direct effect between clinical perfectionism and wellbeing (b = −.05, 95%CI [−.14, .05], p = .36) was non-significant. Significant paths were found between inaction and wellbeing (b=−.20, 95%CI [−.31, −.08], p < .001), self-compassion and wellbeing (b = 2.58, 95%CI [1.75, 3.42], p < .001) and EA and wellbeing (b = .08, 95%CI [−.00, .16], p=.05). Non-significant paths were found between CF and wellbeing (b = −.03, 95%CI [−.15, .09], p=.65) and self-as-content and wellbeing (b = −.04, 95%CI [−.15, .07], p = .47). This suggests that while inaction, self-compassion, and EA were significant predictors of wellbeing, clinical perfectionism, CF, and self-as-content were not. Moreover, the analysis found significant indirect effects of clinical perfectionism on wellbeing through inaction (b = −.10, 95%CI [−.17, −.05]) and self-compassion (b = −.22, 95%CI [−.32, −.13]). Results also found non-significant indirect effects of clinical perfectionism on wellbeing through CF (b=−.01, 95%CI [−.08, .05]), self-as-content (b = −.02, 95%CI [−.08, .03]), and EA (b = .03, 95%CI [−.00, .06]). Figure 2 displays a diagram of the parallel mediation analysis model.

Figure 2. A Statistical Diagram of the Parallel Mediator Model for the effect of Clinical Perfectionism on Psychological Wellbeing.

Note. **= statistically significant pathway at the .001 level

Discussion

This study investigated the associations of clinical perfectionism with distress and positive wellbeing in a non-clinical sample, and whether self-compassion and psychological inflexibility processes – cognitive fusion, experiential avoidance and self-as-content – play mediating roles in these relationships. We also explored whether the three remaining psychological inflexibility processes, inaction, present moment awareness and lack of contact with values influenced these relationships. The study findings support previous observations of clinical perfectionism’s associations with distress and wellbeing, while also providing insights into relationships with transdiagnostic processes relating to self-relating, preoccupation with thinking, experiential avoidance, and rigid inaction.

Relationships between clinical perfectionism, distress, wellbeing, and psychological inflexibility

As hypothesised, we found that higher levels of clinical perfectionism were associated with greater distress and lower wellbeing. This is consistent with previous findings demonstrating associations between maladaptive and clinical perfectionism and greater distress (Adams et al., 2023; Limburg et al., 2017) and for reduced wellbeing (Kanten & Yesıltas, 2015). While results partially supported predictions that CF, self-as-content, and EA mediate the clinical perfectionism-distress relationship, our findings did not support predictions that this combination of processes mediate the clinical perfectionism-wellbeing relationship. However, we found inaction played a mediating role in the effect that clinical perfectionism has on both distress and wellbeing.

Our study makes a contribution in understanding how clinical perfectionism’s direct and indirect effects impact both distress and wellbeing for a non-clinical sample, through investigating hypothesised mediators from the psychological flexibility model. Consistent with this model we found that people reporting greater clinical perfectionism tend to experience less self-compassion, take their cognitions literally, avoid and rigidly self-identify with uncomfortable internal experiences, and behave less congruently with personal values.

These findings are also consistent with Shafran et al. (2002) cognitive-behavioural model of clinical perfectionism. This model describes how clinical perfectionists tend to engage in self-criticism when self-imposed standards are not met, strive towards achieving unrealistic and personally demanding goals to avoid self-criticism, and become entangled with fears of failure, which encourage behaviours (e.g., repeated checking) that help them achieve self-imposed standards. The clinical perfectionism-self-as-content relationship in our study is consistent with Shafran and Mansell (2001), in that failure to achieve self-imposed standards can see individuals with perfectionism developing and inflexibly identifying with negative self-views. The clinical perfectionism-inaction relationship observed in our findings may be related to perfectionistic fear of failure precipitating inaction through avoidance, procrastination or worry, which subsequently, may result in the surrendering of pursuits towards set goals (Mergen, 2020).

Psychological inflexibility processes, distress, and wellbeing

Although CF and inaction significantly predicted distress and wellbeing in the context of clinical perfectionism, our hypotheses implicating self-as-content and EA as predictors in the regression models were not supported. While these were unexpected findings, this is also consistent with arguments for the multidimensional measurement of psychological inflexibility processes (Rolffs et al., 2016), which may lead to more refined understandings. When measuring psychological inflexibility in a multidimensional way, some psychological inflexibility processes will be observed to be more influential than others for particular contexts. Previous studies suggest EA plays a role in maladaptive perfectionism (e.g., Moroz & Dunkley, 2015), however investigations have not usually examined other measures of psychological inflexibility, or the assessment of clinical perfectionism. By doing both our findings expand on this area.

Our findings partially supported hypotheses that psychological inflexibility processes mediate the clinical perfectionism-distress relationship, with CF being a significant mediator, alongside inaction (not hypothesised, but a psychological inflexibility process). Self-compassion was a non-significant mediator within this relationship. Similarly EA was not found to be a mediator, a finding that appears inconsistent with the emphasis on avoidant and unhelpful responses to inner experiences in theorising about perfectionism from a psychological flexibility model perspective (e.g., Ong, Barney, et al., 2019). It was interesting to observe in this study that inaction was a stronger predictor of distress and wellbeing than CF.

For the clinical perfectionism-wellbeing relationship predictions that EA, CF and self-as-content were mediators were not supported, however, inaction and self-compassion were found to be pathways in this model. This is consistent with research demonstrating self-compassion as a mediator within the relationship between perfectionism and wellbeing (i.e., positive and negative affect; Stoeber et al., 2020). This finding further supports the position that psychological inflexibility and flexibility should be investigated multidimensionally (Cherry et al., 2021; Rolffs et al., 2016). The use of unitary measure of psychological inflexibility might obscure the nuance observed here: that clinical perfectionism may impact wellbeing through both a similar pathway to distress (inaction), but also through a different process (limited self-acceptance and kindness). This also supports Winefield et al. (2012) contention that both distress and wellbeing should be researched together as the processes influencing each may be different.

Overall, these mediation analyses suggest reducing CF and inaction may also decrease clinical perfectionism’s impact on distress, while decreasing inaction and promoting self-compassion can reduce clinical perfectionism’s impact on wellbeing. These findings contribute to our understanding of transdiagnostic psychological factors that may contribute the impact of clinical perfectionism on distress and wellbeing, at least for a non-clinical population, and point to further research and theory directions.

Clinical implications

Our findings are derived from cross-sectional analyses and a non-clinical sample, so provide limited scope for informing clinical interventions. However, our findings do provide indications of areas for further investigation, such as further extending the investigation of psychological inflexibility processes and self-compassion with clinical populations, and in using approaches to observing the dynamic interplay of the study variables over time (longitudinal designs, such as experience sampling methods), and through experimental paradigms. The cognitive-behavioural model of perfectionism has been progressive empirically (Shafran et al., 2023) with efficacious interventions developed from a specified clinical model of perfectionism based on cognitive processes (Shafran et al., 2016). We have indicated where our findings are consistent with this model, however we also think there is value in drawing upon transdiagnostic perspectives from contextual behavioural science (e.g., Hayes et al., 2012; Ong, Barney, et al., 2019) that may possibly refine this model or support development of alternate models of clinical perfectionism. The contextual perspective provides an alternate view of how cognition and behaviour is understood and influenced, informing interventions that promote psychological flexibility as a core process through cognitive decentring, self-compassion, acceptance of perfectionistic cognitions, self-as-context, and committed action (Ong, Lee, et al., 2019). The current study is an effort to understand clinical perfectionism from this lens, with the findings suggesting that facets of psychological inflexibility are implicated in the impact on distress and wellbeing. However, further research is needed to observe whether the pathways suggested in this study are influential for people clinically distressed and impacted significantly by perfectionism. Our findings highlight the advantages of taking a multidimensional approach towards measuring psychological flexibility and inflexibility, as doing so may support a nuanced understanding of whether such processes influence the cognitions and behaviours of an individual with perfectionism.

Strengths, limitations, and directions for future research

This study presented several strengths. Firstly, the sample comprised participants from a diverse age-band (18–65 years old), a range of employment statuses, with a sizeable proportion reporting minority identities. Employing the CPQ and MPFI also allowed problematic aspects of perfectionism to be compared to findings from other perfectionism measures, and facilitated exploration of each psychological inflexibility processes to understand which were more influential in understanding how clinical perfectionism may effect distress and wellbeing. This study also demonstrated sound internal validity by implementing measures demonstrating strong reliability and validity in previous research and this study.

Although inviting participants to complete an online anonymous survey reduced the likelihood that experimenter bias impacted results, nevertheless, using an online survey was a limitation as females and highly educated individuals tend to respond more frequently than males and individuals with lower education levels (Kwak & Radler, 2002; Yun & Trumbo, 2000). Our sample reflected this bias: the findings may be less generalisable to the latter populations. Another limitation included implementation of self-report measures, as participant responses may not have accurately reflected true experiences or behaviours related to clinical perfectionism, distress, self-compassion, wellbeing, and psychological inflexibility. A cross-sectional design does not allow determination of temporal, causal relationships with the variables, so the statistical mediation models we tested are useful primarily for further hypothesis generation (A. F. Hayes, 2018). Employing a non-clinical sample for this study also limits the generalisability of results to clinical samples.

Future research should therefore employ non-web-based methodologies to seek participant responses, rather than implementing online surveys. This could increase participation from males and individuals with lower education levels, which may enhance applicability of findings. Alternatively, future studies implementing online surveys could utilise different recruitment strategies to obtain samples representative of the general population. As we indicate above, the use of longitudinal designs to infer causal or temporal relationships between clinical perfectionism, distress, wellbeing, self-compassion, and psychological inflexibility will be informative.

Conclusions

Overall, this study found that for non-clinical sample that the associations of clinical perfectionism with distress and wellbeing can be understood in the context of psychological inflexibility and self-compassion. The study used a multi-dimensional approach to measuring psychological inflexibility in the context of clinical perfectionism, allowing for the observation that slightly different mediatory pathways were implicated with distress (cognitive fusion, inaction) than with wellbeing (self-compassion, inaction). Replication is required to support and expand on these findings, and further research to see if these relationships are observed in clinical populations and longitudinally, which may usefully inform interventions for clinical perfectionism.

Contribution statement

Hung Nguyen: conceptualisation (equal); methodology (equal); writing – original draft (lead); formal analysis (equal); writing – review and editing (equal).

Eric M.J. Morris: conceptualisation (equal); methodology (equal); formal analysis (equal); writing – original draft (supporting); writing – review and editing (equal); supervision (lead).

Disclosure statement

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

Supplementary material

Supplemental data for this article can be accessed at https://doi.org/10.1080/13284207.2024.2362440