The Relationship Between Perfectionism and Psychopathology:

A Meta-Analysis
Karina Limburg,1,2 Hunna J. Watson,1,3,4,5 Martin S. Hagger,1 and Sarah J. Egan1
1
School of Psychology and Speech Pathology, Curtin University, Perth, Australia
2
Department of Psychosomatic Medicine and Psychotherapy, Klinikum rechts der Isar,
Technische Universität München, Munich, Germany
3
Eating Disorders Program, Specialised Child and Adolescent Mental Health Service, Perth,
Australia
4
School of Paediatrics and Child Health, The University of Western Australia, Perth, Australia
5
Department of Psychiatry, University of North Carolina at Chapel Hill, United States
Objective: The clinical significance of 2 main dimensions of perfectionism (perfectionistic strivings
and perfectionistic concerns) was examined via a meta-analysis of studies investigating perfectionism
in the psychopathology literature. Method: We investigated relationships between psychopathol-
ogy outcomes (clinical diagnoses of depression, anxiety disorders, obsessive-compulsive disorder,
and eating disorders; symptoms of these disorders; and outcomes related to psychopathology, such
as deliberate self-harm, suicidal ideation, and general distress) and each perfectionism dimension.
The relationships were examined by evaluating (a) differences in the magnitude of association of the
2 perfectionism dimensions with psychopathology outcomes and (b) subscales of 2 common measures
of perfectionism. Results: A systematic literature search retrieved 284 relevant studies, resulting
in 2,047 effect sizes that were analysed with meta-analysis and meta-regression while accounting for
data dependencies. Conclusion: Findings support the notion of perfectionism as a transdiagnostic
factor by demonstrating that both dimensions are associated with various forms of psychopathology.
C 2016 Wiley Periodicals, Inc. J. Clin. Psychol. 73:1301–1326, 2017.

Keywords: perfectionism; psychopathology; meta-analysis; perfectionistic strivings; perfectionistic

concerns

Introduction
Perfectionism has a critical role in psychopathology. Many studies have linked perfectionism to
affective disorders, anxiety disorders, obsessive-compulsive disorder (OCD), eating disorders,
and other mental health problems. Egan, Wade, and Shafran’s (2011) narrative review concluded
that perfectionism is a “transdiagnostic” risk and maintaining factor for multiple psychological
disorders.
Definitions of perfectionism center on the pursuit of high standards and self-criticism over
not meeting standards, and perfectionism has generally been conceptualized as multidimen-
sional.The two most widely used measures of perfectionism are the Frost Multidimensional
Perfectionism Scale (FMPS; Frost, Marten, Lahart, & Rosenblate, 1990) and the Hewitt
Multidimensional Perfectionism Scale (HMPS; Hewitt & Flett, 1991b), each comprising vari-
ous subscales. Factor analysis of the scales typically results in a two-factor solution comprising

We thank Rachael Glassey for providing assistance with the original literature search. We also thank
Professor Thomas Ehring for his helpful comments on a previous version of this paper.

Please address correspondence to: Karina Limburg, Department of Psychosomatic Medicine and Psy-
chotherapy, Klinikum rechts der Isar, Technische Universität München, Langerstraße 3, 81675 Munich,
Germany. E-mail: [email protected]

JOURNAL OF CLINICAL PSYCHOLOGY, Vol. 73(10), 1301–1326 (2017) 

C 2016 Wiley Periodicals, Inc.

Published online in Wiley Online Library (wileyonlinelibrary.com/journal/jclp). DOI: 10.1002/jclp.22435

 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1302 Journal of Clinical Psychology, October 2017

perfectionistic strivings and perfectionistic concerns (Bieling, Israeli, & Antony, 2004; Frost,
Heimberg, Holt, & Mattia, 1993). Perfectionistic concerns has been suggested to be more strongly
related to maladaptive outcomes, such as negative affect, depression, stress, and anxiety, and
perfectionistic strivings with adaptive outcomes, such as positive affect (Bieling, Israeli, et al.,
2004; Frost et al., 1993). However, there is argument that perfectionistic strivings is also associ-
ated with maladaptive outcomes (e.g., Egan et al., 2011).
Although the MPS scales have been widely used, demonstrated satisfactory reliability and
validity, and have the advantage of enabling cross-study comparisons given their widespread
use, there has been some criticism of the scales. One criticism has been that the FMPS’ doubts
about actions subscale has substantive, nontrivial overlap with symptoms of OCD because
the majority of items on the subscale were derived from a measure of OCD symptoms, the
Maudsley Obsessive-Compulsive Inventory (Hodgson & Rachman, 1977). Consequently, it has
been argued that the doubts about actions subscale primarily reflects checking symptoms of
OCD, rather than perfectionism per se (Shafran & Mansell, 2001). It has also been argued that
the parental expectations and parental criticism subscales of the FMPS potentially confound the
aetiological factors of perfectionism with the measurement of the construct because the subscale
focuses on developmental aspects and the reporting of past experiences with parents (Rhéaume
et al., 2000).
Further, Shafran, Fairburn, and Cooper (2003) argued that the widespread use of the MPS
scales has led to reduced focus on understanding the maintaining aspects and the clinical
relevance of perfectionism, which is why they proposed “clinical perfectionism” referring to the
pursuit of high standards despite negative consequences and basing self-worth on achievement
(Fairburn, Cooper, & Shafran, 2003a). This definition of clinical perfectionism has been used
as a focus in the development of cognitive-behavioural treatments for perfectionism (see Egan,
Wade, Shafran, & Antony, 2014), which have evidence for efficacy (Lloyd, Schmidt, Khondoker,
& Tchanturia, 2015). An overview of the existing perfectionism scales and their categorisation
into the two main domains can be seen in Table 1.
Our hypotheses are based on the accumulating evidence from studies that have shown that
dimensions of perfectionism are significantly higher in clinical samples with a range of disorders,
compared to controls, and associated with psychopathology in nonclinical samples.
The perfectionistic concerns dimension has consistently been shown to be significantly higher
than controls in individuals with clinical disorders like depression (Enns, Cox, & Borger, 2001;
Hewitt & Flett, 1991a; Huprich, Porcerelli, Keaschuk, Binienda, & Engle, 2008; Norman, Davies,
Nicholson, Cortese, & Malla, 1998; Sassaroli et al., 2008); social anxiety disorder (Antony,
Purdon, Huta, & Swinson, 1998; Juster et al., 1996; Saboonchi, Lundh, & Ost, 1999), panic
disorder (Antony et al., 1998; Iketani et al., 2002); and OCD (Antony et al., 1998; Buhlmann,
Etcoff, & Wilhelm, 2008; Frost & Steketee, 1997; Sassaroli et al., 2008). Perfectionistic concerns
have also been linked to generalized anxiety disorder (Handley, Egan, Kane, & Rees, 2014) and
posttraumatic stress disorder (Egan, Hattaway, & Kane, 2014). A smaller number of studies
have found perfectionistic strivings to also be elevated in clinical disorders such as depression
(Hewitt & Flett, 1991a) and OCD (Antony et al., 1998; Buhlmann et al., 2008; Frost & Steketee,
1997; Sassaroli et al., 2008) and associated with generalized anxiety disorder (Handley et al.,
2014).
Perfectionism is a particularly strong risk and maintaining factor in eating disorders. Clinical
perfectionism is one of several core-maintaining mechanisms in Fairburn’s transdiagnostic model
of eating disorders (Fairburn et al., 2003a), which is the theoretical basis for cognitive-behavioral
treatment for eating disorders. Perfectionism is also a central variable in the three-factor
model of bulimia nervosa (Bardone-Cone, Abramson, Vohs, Heatherton, & Joiner, 2006) and
the cognitive-interpersonal model of anorexia nervosa (Schmidt & Treasure, 2006). Individ-
uals with anorexia nervosa and bulimia nervosa have significantly higher scores on perfec-
tionistic strivings and perfectionistic concerns than controls (e.g., Cockell et al., 2002; Halmi
et al., 2000; Lilenfeld et al., 2000; Moor, Vartanian, Touyz, & Beumont, 2004; Sassaroli et al.,
2008).
Similar patterns have been found in nonclinical populations, in which perfectionistic concerns
are positively correlated with depressive symptoms (Bieling, Israeli, et al., 2004; Enns, Cox,
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Perfectionism and Psychopathology 1303

Table 1
Scales Measuring Perfectionism Along With the Classification of Their Subscales Into the Two
Major Dimensions of Perfectionism
Scale Perfectionistic concerns Perfectionistic strivings
Multidimensional Concern over mistakes (CM): Personal standards (PS):
Perfectionism Scale (FMPS; tendency to show negative striving for high standards
Frost et al., 1990) reactions to mistakes and to
interpret mistakes as a failure
Doubts about actions (DA): Organization (O): need for
concern that tasks have not order and neatness
been completed properly
Parental expectations (PE): high
expectations that respondent’s
parents placed on his/her
performance
Parental criticism (PC): parents
being overly critical
Multidimensional Socially prescribed perfectionism Self-oriented perfectionism
Perfectionism Scale (HMPS; (SPP): tendency to expect (SOP): tendency to set high
Hewitt & Flett, 1991b) others to have extremely high standards for oneself while
standards for him/her and to also reflecting the intrinsic
constantly evaluate him/her for motivation to reach those
what he/she achieves standards
Other-oriented perfectionism
(OOP): having
unrealistically high
standards for significant
others
Almost Perfect Scale-Revised Discrepancy High standards
(APS-R; Slaney et al., 2001)
Perfectionism Questionnaire Negative consequences of Perfectionistic tendencies
(PQ; Rhéaume et al., 2000) perfectionism
Children and Adolescent Socially prescribed perfectionism
Perfectionism Scale (CAPS;
Flett, Hewitt, Boucher,
Davidson, & Munro, 1997)
Adaptive/Maladaptive Sensitivity to mistakes
Perfectionism Scale (AMPS;
Rice, Kubal, & Preusser,
2004)
Compulsiveness
Need for admiration
Dysfunctional Attitudes Scale Self-criticism/self-critical
(DAS; Weissman & Beck, perfectionism
1978)
Clinical Perfectionism Perfectionism
Questionnaire (CPQ;
Fairburn, Cooper, &
Shafran, 2003b)
Positive and Negative Negative perfectionism
Perfectionism Scale
(PANPS; Terry-Short,
Owens, Slade, & Dewey,
1995)
Obsessive Beliefs Perfectionism
Questionnaire (Obsessive
Compulsive Cognitions
Working Group [OCCWG],
2001)
Perfectionism subscale of the Socially prescribed perfectionism Self-oriented perfectionism
Eating Disorder Inventory
(EDI-P;Garner, Olmstead,
& Polivy, 1983)
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1304 Journal of Clinical Psychology, October 2017

Sareen, & Freeman, 2001; Graham et al., 2010) and trait anxiety (Gnilka, Ashby, & Noble,
2012). Further, perfectionistic strivings is positively related to depressive symptoms (Lombardo,
Mallia, Battagliese, Grano, & Violani, 2013). In nonclinical populations with symptoms of eating
disorders, many subscales of perfectionism measures from both perfectionism dimensions have
been found to be related to pathology (e.g., Brannan & Petrie, 2008; Miller-Day & Marks, 2006;
Welch, Miller, Ghaderi, & Vaillancourt, 2009).
There are also prospective studies that have linked perfectionism to the development of
depression. These longitudinal studies are important: They give stronger evidence for the di-
rectional relation between perfectionism and depression because the majority of research is
cross-sectional. For example, socially prescribed perfectionism has been found to predict onset
of depressive symptoms at follow-up (Békés et al., 2015; Hewitt, Flett, & Ediger, 1996); and
perfectionistic strivings predict higher depressive symptoms at 1-year follow-up in a clinical
sample (Békés et al., 2015).
Further, patients with clinical depression were followed over 3-year (Dunkley, Sanislow,
Grilo, & McGlashan, 2006) and 4-year (Dunkley, Sanislow, Grilo, & McGlashan, 2009) follow-
up periods and the perfectionism subscale of the Dysfunctional Attitudes Scale (DAS-SC;
Weissman & Beck, 1978), typically referred to as self-critical perfectionism, predicted increases
in depressive symptoms. These findings have been further corroborated in a recent meta-analysis
of 10 longitudinal studies of perfectionism and depression in which perfectionistic strivings and
perfectionistic concerns had small, positive relationships with depressive symptoms at follow-up
(Smith et al., 2016).
The existing body of research has some limitations. The concept of perfectionism and its
measurement has been vigorously debated and has changed over time. The investigations of the
associations between perfectionism and psychopathology have largely been disorder specific,
yet recently there has been growing interest in clinical psychology in transdiagnostic processes.
Transdiagnostic processes are aspects of cognition or behavior that contribute to the maintenance
of more than one psychological disorder (Harvey, Watkins, Mansell, & Shafran, 2004) and have
been referred to as being the points of intersection between personality and psychopathology
(Rodriguez-Seijas, 2015). Thus, they may hold an important key to improving treatment efficacy.
The findings from individual studies would benefit from contextualization to a transdiagnostic
perspective.
Although narrative reviews on the topic of perfectionism and psychopathology exist in eating
disorders (Bardone-Cone et al., 2007), across various disorders (Egan et al., 2011; Shafran &
Mansell, 2001), single meta-analyses on treatment outcomes for perfectionism (Lloyd et al.,
2015), and longitudinal studies of depression (Smith et al., 2016), there has been no quantitative
synthesis of relations between perfectionism and psychopathological outcomes across disorders
and symptoms using meta-analytic techniques. There have also been mixed study findings, with
some studies reporting an association between perfectionistic strivings and psychopathology and
other studies finding a null effect. Individual studies are limited in their ability to resolve these
contradictions. To test the association between perfectionistic concerns and perfectionistic striv-
ings with psychopathology, it is necessary to consider measures of each perfectionism domain.

The Present Study

The aim of this study was to investigate the relationship of perfectionistic strivings and perfection-
istic concerns with psychopathology across studies using meta-analytic techniques. Specifically,
we aimed to test whether the literature supports the hypothesis that perfectionistic concerns can
be considered a transdiagnostic process across disorders in clinical samples and psychopathol-
ogy in nonclinical samples, as proposed by (Egan et al., 2011), and whether perfectionistic
strivings is associated with psychopathology. The clinical relevance of understanding the link
between perfectionism and psychopathology is that if perfectionism is found to be relevant across
disorders (i.e., transdiagnostic), then it may be important to target in an attempt to reduce the
symptoms of a range of disorders (Egan, Wade, et al., 2014).
The second aim was to investigate the relative contribution of the subscales of the two
most commonly used scales, the FMPS and HMPS, in the prediction of psychopathology, to
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Perfectionism and Psychopathology 1305

address the question of whether there are certain subscales that show a stronger relationship to
psychopathology than others. By identifying the scales that are most associated with
psychopathology, it may be possible to recommend which subscales should be used to assess
perfectionism in the context of psychopathology.
A meta-analysis of the extant literature may assist in providing a better understanding of
the relationship between perfectionism and psychopathology across disorders and symptoms
because it will provide bias-corrected estimates of the size and pattern of effects that cannot
be gained from narrative reviews. Critically, we will examine the unique contribution of the
perfectionism dimensions on psychopathology. Much of the research on perfectionism has
examined zero-order effects of different perfectionism dimensions on outcomes. This does not
account for the unique effects of the dimension on the outcome when accounting for effects of
other perfectionism domains.
Given that different dimensions of perfectionism have been shown to be significantly cor-
related, it is possible that overlap in the dimensions may give a misleading representation of
the true effects of the dimensions. Because of this overlap, the zero-order effects of perfection-
ism dimensions will give a misleading, most likely inflated, representation of the true effects
of the perfectionism dimensions. We will therefore test the unique effects of the perfectionism
dimensions using meta-analytic path analysis of the weighted averaged correlations between
perfectionistic dimensions and the psychopathological outcomes.
The analysis will also permit the assessment of the degree of variability in effects across stud-
ies that cannot be attributed to the methodological artefacts corrected for in the analysis (i.e.,
sampling error). Identification of substantive heterogeneity in links between perfectionism and
psychopathology will catalyze a search for key moderators to resolve the heterogeneity, a key goal
of meta-analysis. We will evaluate the effects of potential moderators of the relationship between
perfectionism and psychopathology outcomes, such as age, gender, and, in case of follow-up stud-
ies, the time between baseline assessment of perfectionism and the assessment of the outcome.

Method
Search Strategy
We used several strategies to identify eligible studies. First, we searched the databases ERIC,
Embase, ISI Web of Science (Science Citation Index Expanded, Social Science Citation Index
Expanded), Medline, PsycINFO, PsycARTICLES, and Scopus for all years covered through to
July 2013. The following key words were used: perfectionism, mental health, outcome, behav-
ior/behavior, intervention, and psychopathology. We conducted manual searches of reference
lists from prior literature reviews. We queried the electronic mailing list Perfectionism Network
Mailing List to identify studies that were accepted to a peer-reviewed journal, but not published
at the time of the literature search. Active researchers in the field of perfectionism who had
previously published two or more relevant articles on perfectionism and psychopathology were
contacted to request additional citations.

Relevant Outcomes
The area of psychopathology contains a variety of outcomes, and as such there was a need
to classify them into appropriate, meaningful categories. Consequently, we identified three
broad categories of outcome: clinical disorders, symptoms of disorders, and outcomes related
to psychopathology. Through scanning the existing literature, four subcategories of frequently
evaluated clinical disorders in the perfectionism literature were identified for inclusion in subse-
quent analyses: depression, anxiety disorders, OCD, and eating disorders (anorexia nervosa and
bulimia nervosa). The relevant symptoms of disorders had to fall into one of the proposed
disorder categories.
Because studies reported various subsets of symptoms that represent the same symp-
tom category, conceptually related symptoms were aggregated into analyzable subcategories.
For example, symptoms of social phobia that were measured in studies included fear of
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1306 Journal of Clinical Psychology, October 2017

Figure 1. Categories and subcategories of outcomes that were analyzed in the meta-analysis.

communication situation, fear of negative evaluation, shyness, social anxiety, and social in-
teraction anxiety, all of which were aggregated into the subcategory symptoms of social phobia.
In addition, the OCD symptom compulsions was formed by integrating various reported com-
pulsions, such as checking, cleaning, ordering, and washing. Additional clinical outcomes that
cannot be directly related to a disorder such as suicidal ideation and self-harming behavior were
subsumed as outcomes related to psychopathology. The first author and the final author (Egan),
who is an experienced clinical psychologist, conducted this process of aggregation according
to symptoms to ensure the classification was clinically meaningful. A summary of all analyzed
categories is presented in Figure 1.

Inclusion and Exclusion Criteria

The search strategy resulted in the identification of studies relevant to the relationship between
perfectionism and different forms of psychopathology. Eligible studies were required to assess
perfectionism using a validated self-report measure of trait perfectionism and a relevant outcome
(see Figure 1). All outcomes had to be assessed with validated measures. No restriction was
placed on study characteristics regarding participant age, gender, race, or ethnicity; results
from clinical and nonclinical samples were included, and studies from any nation and any time
period were considered relevant. Studies had to be printed or accepted in peer-reviewed journals.
Dissertations or unpublished data were excluded to avoid the risk of retrieving duplicate effect
sizes. Adequate detail of method, results, and data to calculate effect sizes had to be present
for a study to be included. Eligible research designs included correlational studies and studies
reporting a group comparison, for example, between a clinical and a nonclinical group. Reasons
why studies were identified as not eligible were coded. The most common reasons for studies not
being eligible are presented in Figure 2.

Study Identification
To determine study eligibility, two independent judges, who both held a bachelor’s degree
in psychology, examined titles and abstracts of all identified studies. If differing assessments
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Perfectionism and Psychopathology 1307

Identified in systematic search (n = 2,497)

Excluded (n = 1,522)
Duplicates (n = 33)
Did not satisfy criteria (n = 1,489)

Retrieved for eligibility (n = 975)

Excluded (n = 691)
No relevant outcome reported (n = 298)
Data is not sufficient or applicable to calculate effect size (n = 142)
Not focusing on perfectionism (n = 58)
Not using an explicit, validated self-report measure of perfectionism (n = 54)
Printed in a language other than English or German (n = 54)
Dissertation (n = 23)
Editorial/abstract only (n = 18)
No quantitative data available (n = 16)
Not using a continuous, validated measure for the respective outcome (n = 13)
Review article (n = 9)
Sample duplicate (n = 3)
Book chapter (n = 3)

Eligible and included in meta-analysis

(n = 284 studies containing 323 tests)

Figure 2. Process of study selection.

occurred, then those cases were discussed until consensus was reached. The first author assessed
all remaining studies in full text and coded them.

Coding Procedures
The following data were coded for each of the eligible study reports: sample size, mean age of
participants (years), proportion of female participants, sample type (i.e., clinical, nonclinical),
diagnosis (if applicable), perfectionism measure, and effect sizes. Because some studies did not
report the exact mean age of participants, it was estimated using valid indicators. For example, if
data were reported on a sample of undergraduate students, then it was estimated to be 19 years.

Formation of Perfectionistic Strivings and Perfectionistic Concerns Perfectionism

Different views exist as to which subscales of which perfectionism measures should be subsumed
under the two main perfectionism dimensions. For the purpose of this review, the formation of
perfectionistic strivings and perfectionistic concerns follows the suggestions of Stoeber and Otto
(2006), who conducted a review of different conceptualizations of perfectionism and proposed
recommendations on how to form the two dimensions based on theoretical considerations. In
addition to the subscales of FMPS and HMPS, the authors considered the subscales of the
Almost Perfect Scale-Revised (APS-R; Slaney, Rice, Mobley, Trippi, & Ashby, 2001) and the
Perfectionism Questionnaire (PQ; Rhéaume et al., 2000) in their classification.
Based on empirical evidence we determined a number of additional instruments to be included
in the two dimensions, they can be found in Table 1. The decisions whether measures are valid
for inclusion or exclusion of studies from the meta-analysis were made in an expert consensus
process after considering the items of the various measures and aligning them to subscales
of more established measures of perfectionism such as the FMPS and HMPS, and on the
basis of correlations between these scales and psychopathology in reviews in the literature
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1308 Journal of Clinical Psychology, October 2017

(Egan et al., 2011). Following Stoeber and Otto’s (2006) recommendation, FMPS-organization
and HMPS-other-oriented perfectionism have been omitted because findings as to whether these
scales represent perfectionistic concerns or perfectionistic strivings were unclear. Furthermore,
FMPS-parental expectations and parental criticism have been disregarded because these scales
may not reflect core aspects of perfectionism but preceding factors that emerge during upbringing
(Stoeber & Otto, 2006).

Statistical Methods
Effect size estimation procedure. Relevant primary studies reported either (a) zero-order
correlation coefficients, r, between perfectionism and the relevant outcome or (b) group com-
parisons between a clinical and a comparison group regarding perfectionism using parametric
tests of difference, such as t tests or analysis of variance models. Given the aim of examining
the relationship between perfectionism and an outcome and a high number of studies report-
ing correlation coefficients, the zero-order correlation coefficient was selected as the effect size
metric. Thus, the correlation coefficients reported in primary studies were extracted. Because
the variance of the correlation coefficient depends on the correlation, standardization of the
effect size using Fisher’s r to Z transformation is recommended (Borenstein, Hedges, Higgins,
& Rothstein, 2009).
All analyses were performed using the transformed values; the results were then back-
transformed (Borenstein et al., 2009). When correlation coefficients were not reported, effect
sizes were calculated from other statistics, such as Cohen’s d, the standardized mean difference
score (Cohen, 1988). This was obtained through calculating the difference between the per-
fectionism means for the clinical and comparison conditions divided by the pooled standard
deviation (Borenstein et al., 2009). The standardized mean difference (d) was then converted
into a correlation (r) (Borenstein et al., 2009).

Statistical analysis. Two main sets of analyses were conducted in this study. The first set of
analyses aimed to obtain weighted average effect sizes for the relationship between perfectionism
and various outcomes related to psychopathology, thus gaining a general understanding of the
size of the effect for the relations and the degree of heterogeneity associated with the effects.
The second set of analyses aimed to further investigate the relationship by implementing meta-
regression models.
An important issue that had to be considered in all analyses was the occurrence of data
dependencies. Because of the fact that the majority of studies usually reported more than one
outcome and these outcomes were often belonging to the same subcategory, there were many
cases of nested effect sizes within studies. To address this concern, two meta-analytic methods
were applied in the first set of analyses (the calculation of weighted average effect sizes).
First, we used the conventional Hedges-Olkin random-effects model (Hedges & Olkin, 1985)
when effect sizes to be combined were not nested. Second, we used Hedges’ robust variance
estimation model (Hedges, Tipton, & Johnson, 2010) when effect sizes to be combined contained
nested effects. The robust variance estimation model is advantageous because it takes into
account the within-study dependencies by introducing an estimate of the mean correlation (ρ)
between all pairs of nested effect sizes. This estimate is involved in the calculation of the between
study sampling variance estimate (τ²). Because the robust variance estimator does not require
information on the true correlation in the data, τ² was estimated with ρ = 0.80 in all analyses, as
recommended by (Tanner-Smith, Wilson, & Lipsey, 2013).
Weighting of the studies was conducted by calculating inverse-variance weights for all analyses.
Heterogeneity in the effect sizes could be estimated via evaluation of the τ²-statistic in the
context of the robust standard error estimation technique. To determine whether the observed
heterogeneity is substantial or large, a prediction interval around the mean effect size (μ) can be
calculated (after Black, 2009).
Second, to evaluate the relative contribution of the effects of perfectionism dimensions on
psychopathological outcomes, we used the zero-order averaged weighted correlations between
the dimensions and each psychopathology outcome as input into a meta-analytic path analysis.
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Perfectionism and Psychopathology 1309

In each path analysis, the outcome of interest was regressed on to the perfectionism dimensions.
The models were estimated using a maximum likelihood estimation method, with the average
sample size as the input sample size (Viswesvaran & Ones, 1995). Given the number of analyses
and sample sizes, we used a stringent probability level (p < .01) to indicate a statistically
significant effect and 95% confidence intervals of the parameter estimates to test whether the
relative contribution of each perfectionism dimension differed.
Finally, we applied nested meta-regression modelling to further investigate the relationship
between perfectionism and psychopathology by comparing this relationship among various
outcomes and sample types. In addition, to ensure sufficient statistical power, we set our criterion
for the minimum number of primary studies per moderator group to 10, as recommended by
Dalton and Dalton (2008). Only moderator variables evaluated in at least 10 tests were considered
in the meta-regression. To rule out type II errors, a sensitivity analysis was performed for each
model after the meta-regression.

Tests for data censoring. Two forms of bias in the effect sizes, including variance that
could be attributed to publication bias (Rosenthal, 1979) and funnel plots were visually inspected
to detect asymmetry (Borenstein et al., 2009) and the fail-safe N-method was used to calculate
the number of null results that would lead to a nonsignificant effect size if added to the analysis
(Rosenthal, 1979). Egger’s asymmetry test was also used to formally test for small study biases
that could be attributed to publication bias (Egger, Smith, Schneider, & Minder, 1997; Hagger
& Chatzisarantis, 2014).

Statistical programs. The data were extracted and coded in comprehensive meta-analysis
(version 2.0); the program was also used to standardize and convert among effect sizes. IBM
SPSS statistics (version 21.0) and MPlus (version 7.31) analysis package were used for the
analyses.

Results
Description of Studies
The process of study selection is displayed in Figure 2. In sum, 284 studies containing
323 independent samples with effect size data and a total of 57,200 participants were included,
18 of which used a longitudinal design. The majority of participants were female (74.0%) and
mean sample age across studies was 25.06 years (standard deviation [SD] = 8.13). The ma-
jority of primary studies (65.1%) reported data from nonclinical samples. The psychological
disorders most evaluated in relation to perfectionism were anxiety disorders (6.2%), followed
by eating disorders (4.6%), OCD (3.9%), and depression (1.6%). The symptoms most evalu-
ated were depressive symptoms (28.0%), followed by symptoms of anxiety disorders (19.8%),
OCD (18.1%), and eating disorders (12.0%). Related outcomes including deliberate self-harm,
suicidal behavior and ideation, and general psychological distress were evaluated in 5.8% of all
tests.
The majority of studies reported data from nonclinical samples and thus focused on symptoms
of psychopathology rather than clinically diagnosed disorders. As a consequence, the focus for
the current review is on disorder symptoms rather than clinically diagnosed disorders. The
perfectionism measure most utilized was the FMPS (48.1%). The HMPS was used in 27.8%
of all studies, followed by EDI-P (7.9%), APS-R (5.3%), OBQ-P (4.2%), CAPS (2.1%), DAS
(1.9%), and CPQ (0.7%), with a small minority using other scales. Overall, 2,047 effect sizes
across these outcomes were included in the analysis.

Weighted Average Effect Sizes

To investigate the presence of heterogeneity, the preliminary meta-analysis did not distinguish
between perfectionism dimensions or psychopathological outcome (i.e., clinical disorders, symp-
toms of disorders, outcomes related to psychopathology). All effect sizes from all studies were
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1310 Journal of Clinical Psychology, October 2017

Figure 3. Funnel plot of the standard error by the effect size Fisher’s Z.

combined in a single analysis, and the weighted mean effect size was estimated using weighted
random-effects analyses with robust variance estimates; this resulted in an overall effect size
for the association between perfectionism and all psychopathological outcomes. The weighted
average effect size was 0.26 (n = 2,047, k = 323, p < 0.001). The average τ2 of 0.05 (SD =
0.26, p < 0.001) indicated the presence of additional heterogeneity in the effect size estimates
unattributed to the methodological artefacts corrected for in the analysis. To determine whether
this heterogeneity was substantial, the prediction interval was calculated to estimate the range in
which a new estimated effect size would fall in 95% of new studies. The prediction interval was
−0.54 to 3.0, indicating a wide range in which a new effect size could fall and thus substantial
heterogeneity was assumed.
This finding confirmed our expectation given that the literature refers to distinct influences
of perfectionism in the context of different psychological disorders, symptoms, and outcomes
related to psychopathology. Therefore, subsequent analyses of effect size exploring the modera-
tion of the effect size by perfectionism dimensions and by separate psychopathological outcomes
were justified.

Small Study Bias

A funnel plot on the overall set of studies was investigated for asymmetry to test for small study
bias that might be indicative of publication bias (see Figure 3). Visual inspection of the plot
appeared slightly asymmetric. More importantly, Egger’s test for asymmetry based on the funnel
plot indicated an absence of substantial bias.

Perfectionism dimensions and psychopathology. We used weighted random-effects anal-

yses with robust variance estimates to investigate the relationships between the two main
perfectionism dimensions and psychological disorders, symptoms, and related outcomes. To
draw basic conclusions about specific patterns of perfectionism for each outcome, effect
sizes were pooled for the respective outcome and perfectionism dimension and evaluated
separately.
In addition to investigating the overall correlations between the two main perfectionism di-
mensions and psychopathology outcomes, another research objective was to assess the relative
importance of perfectionistic concerns and perfectionistic striving. Previous studies have exam-
ined the overlap of the 95% confidence intervals (CIs) of the mean effect size for each outcome
group (Tanner-Smith et al., 2013). However, this approach gives a misleading representation
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Perfectionism and Psychopathology 1311

Perfectionistic
Concerns

Psychopathology
Outcome

Perfectionistic
Strivings

Figure 4. Diagram depicting generalized meta-analytic path model for effects of perfectionism dimensions
on psychopathological outcomes.

of the relative effects because it focuses on the zero-order correlations, which are, in essence,
separate analyses. Instead, we conducted a series of meta-analytic path analyses in which each
outcome was regressed on the two main perfectionism dimensions–perfectionistic concerns and
perfectionistic strivings. The generalized model is depicted in Figure 4. Consistent with two-
variable path analytic models, the two predictors were correlated.
The average weighted zero-order (r) and unique (β) effect sizes between perfectionism dimen-
sions and psychopathological outcomes are presented in Table 2. Regarding clinical disorders,
there was evidence that both perfectionistic concerns and perfectionistic strivings were signifi-
cantly related to depression, anxiety disorders, OCD, and bulimia nervosa. Only perfectionistic
strivings was significantly related to anorexia nervosa and not perfectionistic concerns, and the
variability in the latter effect size was substantial as illustrated by the wide confidence intervals
that included the value of zero. Examination of the effect sizes from the path analyses provided
detail on the relative contribution of each dimension in the prediction of psychopathological
outcomes. For all outcomes, the unique effect (β) for perfectionistic concerns was larger than
the effect for strivings with no overlap in the confidence intervals of the effect size.
Importantly, the effects for strivings were much smaller than the averaged zero-order correla-
tion, indicating that these effects were relatively trivial in comparison to the effects of concerns.
CIs indicated that the effect for depression included the value of zero and the effect for OCD
was approaching zero.
Regarding symptoms of disorders, the weighted average zero-order effect sizes were signifi-
cant for the associations between the two main domains of perfectionism and various symptom
outcomes, with the exception of the effect for social phobia symptoms and perfectionistic striv-
ings. Examination of the unique effects from the path analyses revealed that the effects for
perfectionistic concerns were substantially larger for many of the outcomes including depressive
symptoms, anxiety, social phobia symptoms, worry, OCD symptoms, obsessive beliefs, global
eating pathology, binge eating, and body dissatisfaction. For these outcomes, the CIs revealed
significant differences in the size of the effects for perfectionistic concerns relative to perfection-
istic strivings.
In contrast, there was overlap in the CIs for the effect sizes for perfectionism dimensions on
dietary restraint, drive for thinness, and thin ideal internalization. The effect for perfectionistic
concerns and perfectionistic strivings was no different in the magnitude of the effects, indicating
a relatively equal contribution to explaining variance in eating disorder outcomes.
 Table 2
1312

Weighted Averaged Zero-Order (r) and Unique (β) Effect Sizes With Confidence Intervals for Relations Between the Two Main Dimensions of Perfectionism and
Psychological Disorders, Symptoms, and Psychopathological Outcomes

Perfectionistic concerns Perfectionistic strivings

95% CI of r 95% CI of β 95% CI of r 95% CI of β

Outcome k n r LL UL β LL UL k n r LL UL β LL UL

Psychological Depression 9 12 .40*** .29 .50 .40*** .32 .48 7 8 .18* .04 .32 .01 −.08 .09
disorders Anxiety disorders 20 49 .30*** .24 .36 .33*** .29 .37 16 29 .07** .01 .12 −.08*** −.12 −.03
OCD 14 32 .35*** .24 .45 .37*** .32 .43 10 15 .11** .04 .18 −.06* −.11 −.01
Anorexia nervosa 5 8 .81 −.20 .99 .70*** .65 .75 4 4 .56** .23 .78 .25*** .21 .30
Bulimia nervosa 5 9 .45** .22 .64 .36*** .22 .51 4 4 .36*** .24 .47 .20** .06 .35
Symptoms of Depressive symptoms 151 256 .39*** .37 .41 .42*** .41 .43 118 162 .11*** .09 .14 −.08*** −.09 −.07
disorders Anxiety 69 149 .35*** .33 .38 .36*** .34 .37 48 104 .14*** .11 .17 −.02*\ −.03 −.01
Social phobia symptoms 14 38 .39*** .31 .47 .46*** .45 .47 12 26 .05 −.03 .13 −.15*** −.16 −.14
Worry 10 11 .47*** .42 .52 .44*** .40 .49 5 5 .26*** .11 .40 .07** .03 .11
OCD symptoms 29 79 .30*** .25 .35 .30*** .28 .32 15 34 .14*** .09 .20 .01 −.01 .03
Obsessive beliefs 12 35 .54*** .45 .61 .49*** .46 .52 1 1 .33*** .23 .43 .12*** .08 .15
Global eating pathology 19 23 .27*** .23 .31 .22*** .19 .25 18 21 .21*** .23 .31 .11*** .09 .14
Binge eating 10 18 .30*** .26 .35 .32*** .29 .35 8 13 .10*** .05 .15 −.04** −.07 −.01
Body dissatisfaction 24 35 .32*** .21 .42 .27*** .24 .29 20 24 .24** .09 .35 .12*** .10 .15
Dietary restraint 20 27 .29*** .21 .36 .21*** .19 .24 18 22 .27*** .21 .33 .18*** .15 .20
Journal of Clinical Psychology, October 2017

Drive for thinness 6 7 .22* .01 .41 .14*** .09 .20 5 5 .24*** .17 .28 .18*** .12 .24
Thin-ideal internalisation 2 2 .21*** .11 .30 .15*** .08 .22 2 2 .20*** .13 .26 .13*** .06 .21
Outcomes related Suicidal ideation 19 22 .31*** .26 .36 .34*** .30 .38 14 15 .09* .02 .15 −.06** −.10 −.02
to
psychopathology General psychological distress 9 17 .42*** .32 .51 .42*** .39 .46 9 12 .18*** .12 .23 −.01 −.04 .03

Note. k = number of tests; n = number of effect sizes; r = weighted averaged zero-order effect size (correlation) for perfectionism dimension-outcome relation from meta-analysis;
CI = confidence interval; LL = lower limit; UL = upper limit; β = parameter estimate for unique effect of perfectionism dimension on outcome from path analysis.
*p < 0.05. **p < 0.01. ***p < 0.001.

10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Perfectionism and Psychopathology 1313

Regarding outcomes related to psychopathology, the perfectionism dimensions were signifi-

cantly related to suicidal ideation and general psychological distress. The unique effects from the
path analysis and their confidence intervals indicated significantly larger effects for perfection-
istic concerns relative to perfectionistic strivings for both outcomes with the effects for strivings
not significant or relatively trivial in size.
Overall, these data support the overall hypothesis of relations between perfectionism and
psychopathology. Specifically, both perfectionistic concerns and perfectionistic strivings were
significantly associated with a range of different psychological disorders including symptoms of
psychological disorders and related outcomes within clinical and community samples. However,
tests of the unique effects of the specific perfectionism dimensions revealed that perfectionistic
concerns had the larger effect for most outcomes; the only exceptions were for outcomes related to
eating disorders, in which concerns and strivings contributed approximately equally to explaining
variance.

Perfectionism measures and psychopathology. The second aim also concerned the eval-
uation of each outcome in relation to perfectionism but was concerned with the two most
commonly used perfectionism scales: the FMPS and HMPS. The purpose was to investigate
whether certain perfectionism subscales showed a stronger relationship with some disorders,
symptoms, or related outcomes than others. As expected, average effect sizes revealed higher
overall correlations on scales previously found to measure perfectionistic concerns, such as
FMPS-concern over mistakes, doubts about actions, and HMPS-socially prescribed perfection-
ism, than on scales previously found to measure perfectionistic strivings, such as FMPS-personal
standards and HMPS-self-oriented perfectionism for depression, anxiety disorders, and OCD.
The same tendency was observed for anorexia nervosa but the differences between magnitudes
of correlation coefficients were smaller.
Regarding the remaining scales of the two measures, worthy of note was a significant pos-
itive overall correlation for FMPS-organization and anorexia nervosa. Further, the combined
dimension of the parenting related subscales of the FMPS, parental expectations and parental
criticism, showed significant positive (but small) overall correlations with anxiety disorders and
OCD. The results are presented in Table 3.
In order to examine heterogeneity due to between-studies variability, we evaluated various
indicators: In case of dependent effect sizes, we examined the τ²- statistic as a between-study
sampling variance estimate because it involves an estimate of the mean correlation (ρ) between
all pairs of nested effect sizes. In case of independent effect sizes, we evaluated the Q-statistic,
which informs about the presence of heterogeneity, and the I² index, which assesses the degree
of heterogeneity. These values indicated substantial heterogeneity, which may likely be due
to extraneous moderator variables beyond the artefacts corrected, for that may account for
the variation in the magnitude of the correlation between perfectionism and outcome. Those
moderator variables can be identified by meta-regression analyses.
In sum, various scales of both measures were differentially related to the different outcomes,
with most scales previously found to measure perfectionistic concerns showing a stronger rela-
tionship to psychopathology than those previously found to measure perfectionistic strivings.
Nevertheless, significant findings for the association between subscales measuring perfectionistic
strivings (FMPS-personal standards, FMPS-organization, HMPS-self-oriented perfectionism)
and psychopathology were found, indicating that, in contrast to the view of previous authors
(e.g., Stoeber & Otto, 2006), perfectionistic strivings may also play a role in some forms of
psychopathology.

Meta-Regression Analyses
To examine the influences that may have led to the observed heterogeneity in the effect sizes
for perfectionism dimensions on psychopathology outcomes, we implemented meta-regression
analyses. This procedure was used to explain variance in the average weighted effect size of
perfectionism and psychopathology with perfectionism dimensions and sample types as mod-
erators. Specifically, the average weighted correlation between all perfectionism dimensions
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1314 Journal of Clinical Psychology, October 2017

Table 3
Weighted Mean Effect Sizes and 95% Confidence Intervals for the Relationship Between Subscales
of FMPS and HMPS and Several Psychological Disorders

Depression Anxiety disorders

95% CI 95% CI

k n r LL UL k n r LL UL

FMPS-concern over 3 3 0.45** 0.19 0.66 10 14 0.34*** 0.25 0.42

mistakes
FMPS-doubts about 3 3 0.34** 0.10 0.54 8 13 0.25*** 0.11 0.37
actions
HMPS-socially 4 4 0.50*** 0.39 0.61 6 9 0.53*** 0.29 0.71
prescribed
perfectionism
FMPS-personal 3 3 0.10 −0.02 0.21 10 15 0.05 −0.02 0.11
standards
HMPS-self-oriented 4 4 0.26** 0.09 0.41 6 9 0.08 −0.05 0.21
perfectionism
FMPS-organization 2 2 −0.08 −0.21 0.05 8 13 0.01 −0.07 0.08
FMPS-parental 3 6 0.29 −0.14 0.63 9 25 0.16*** 0.11 0.21
expectations and
criticism

OCD Anorexia nervosa

95% CI 95% CI

k n r LL UL k n r LL UL

FMPS-concern over 10 11 0.37*** 0.21 0.52 3 3 0.92* 0.02 1.00

mistakes
FMPS-doubts about 9 10 0.54*** 0.30 0.71 3 3 0.89 −0.10 0.99
actions
HMPS-socially 1 1 0.26* 0.07 0.43 1 1 0.78*** 0.66 0.86
prescribed
perfectionism
FMPS-personal 10 11 0.10* 0.01 0.19 3 3 0.44** 0.15 0.66
standards
HMPS-self-oriented 1 1 0.16 −0.04 0.34 1 1 0.83*** 0.74 0.89
perfectionism
FMPS-organization 8 9 0.08* 0.02 0.14 3 3 0.41** 0.13 0.64
FMPS-parental 10 21 0.16* 0.06 0.25 3 5 0.36 −0.35 0.81
expectations and
criticism

Note. k = number of tests; n = number of effect sizes; CI = confidence interval; LL = lower limit; UL = upper
limit.
*p < 0.05. **p < 0.01. ***p < 0.001.

and outcomes was predicted in a series of nested meta-regression models, with perfectionism
dimension and sample type as predictors. Each meta-regression model aimed at answering a
specific research question.
The moderators were by dummy-coded contrast variables representing membership of the
moderator groups. For moderators with more than two categories, a reference category had to
be determined. The reference category was determined as the characteristic with the largest
number of effect sizes. The first regression model aimed to predict the overall effect of
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Perfectionism and Psychopathology 1315

Table 4
Coefficients and Robust Standard Errors From Nested Meta-Regression Models Predicting Effect
Sizes of the Relationship Between Perfectionism and Psychopathology by the Subscales of FMPS
and HMPS and Sample Characteristics With All Obtained and Relevant Effect Sizes Included
(k = 322, n = 2,012)

Model

Variable B SE β

Perfectionistic strivings −0.21** 0.06 −0.14

FMPS-concern over mistakes −0.02 0.09 −0.01
FMPS-doubts about actions 0.15 0.11 0.05
HMPS-socially prescribed perfectionism 0.10 0.14 0.06
FMPS-personal standards 0.13 0.08 0.05
HMPS-self-oriented perfectionism −0.03 0.13 −0.02
FMPS-parental expectations and criticism −0.06 0.08 −0.03
FMPS-organization −0.38 0.22 −0.09
HMPS-other-oriented perfectionism −0.35*** 0.08 −0.16
Age < 0.001 < 0.001 0.07
Gender (% female) < 0.001 < 0.001 0.05
Outcome time (months) < 0.001 < 0.001 −0.03

Note. The reference category was perfectionistic concerns. Coefficients shown for the between-study effects
of variables that varied within and between studies. Age and gender were not provided in n = 35 cases, which
is why the number of effect sizes included here is lower than the overall number of effect sizes. k = number of
tests; n = number of effect sizes; B = unstandardized regression coefficient; SE = standard error of B; β =
standardized regression coefficient; outcome time = months between baseline assessment of perfectionism
and assessment of outcome.
**p < 0.01. ***p < 0.001.

perfectionism on all psychopathological outcomes controlling for study characteristics. The

perfectionism dimensions (perfectionistic concerns, perfectionistic strivings, FMPS-concerns
over mistakes, FMPS-doubts about actions, HMPS-socially prescribed perfectionism, FMPS-
personal standards, HMPS-self-oriented perfectionism, FMPS-parental expectations and criti-
cism, FMPS-organization, HMPS-other-oriented perfectionism) and study characteristics (age,
gender, time between baseline assessment of perfectionism and follow-up assessment of outcome)
were the moderator variables. The reference category for the dummy variable perfectionism was
determined to be perfectionistic concerns. Perfectionistic strivings and the remaining subscales
measuring different aspects of perfectionism were compared with the reference category by
evaluating the polarity of their coefficient.
The results are presented in Table 4. Both perfectionistic strivings as well as HMPS-other-
oriented perfectionism accounted for a significant amount of variance in effect sizes and were
associated with a significantly smaller effect compared to perfectionistic concerns. No other
perfectionism dimension or study characteristic yielded a significant effect.
The second set of meta-regression models was implemented to evaluate possible moder-
ators as well as to investigate the subscales of FMPS and HMPS in specific sample types
(clinical vs. nonclinical). Two meta-regression models were calculated: one restricted to clin-
ical samples, the other restricted to nonclinical samples (see Table 5). Both models aimed
to predict the overall effect of perfectionism on all psychopathological outcomes controlling
for study characteristics in the respective sample type (clinical vs. nonclinical). The reference
category for the dummy-coded perfectionism dimension moderator variable was overall per-
fectionistic concerns in both models. In the moderator analysis limited to clinical samples,
HMPS-socially prescribed perfectionism accounted for a significant amount of variance in
the effect size; it was associated with a significantly higher effect compared to perfectionistic
concerns.
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1316 Journal of Clinical Psychology, October 2017

Table 5
Coefficients and Robust Standard Errors From Nested Meta-Regression Models Predicting Effect
Sizes of the Relationship between Perfectionism and Psychopathology by the Subscales of FMPS
and HMPS and Sample Characteristics, Restricted to Effect Sizes on Clinical Samples Versus
Nonclinical Samples

Model 1

Variable B SE β

Clinical samples Perfectionistic strivings −0.72 0.40 −0.19

(k = 42, n = 233) FMPS-concern over mistakes 0.25 0.88 0.04
FMPS-doubts about actions 0.46 0.93 0.08
HMPS-socially prescribed perfectionism 0.73* 0.46 0.20
FMPS-parental expectations and criticism 0.18 0.52 0.05
FMPS-organization −1.34 0.90 −0.24
HMPS-other-oriented perfectionism −0.27 0.39 −0.07
Age −0.01 0.01 −0.15
Gender (% female) < 0.001 < 0.001 0.27
Outcome time (months) < 0.001 0.02 −0.02
Nonclinical samples Perfectionistic strivings −0.21** 0.06 −0.16
(k = 216, FMPS-concern over mistakes −0.02 0.14 −0.01
n = 1,332) FMPS-doubts about actions 0.10 0.11 0.04
HMPS-socially prescribed perfectionism 0.11 0.18 0.08
FMPS-personal standards 0.15 0.08 0.08
HMPS-self-oriented perfectionism −0.10 0.18 −0.07
FMPS-parental expectations and criticism −0.05 0.08 −0.03
FMPS-organization −0.16 0.24 −0.04
HMPS-other-oriented perfectionism −0.35*** 0.09 −0.19
Age < 0.001 < 0.001 0.09
Gender (% female) < 0.001 < 0.001 0.01
Outcome time (months) < 0.001 < 0.001 −0.02

Note. The reference category in both models was perfectionistic concerns. Coefficients shown for the between-
study effects of variables that varied within and between studies. SE = standard error; k = number of tests;
n = number of effect sizes; B = unstandardized regression coefficient; SE = standard error of B; β =
standardized regression coefficient; outcome time = months between baseline assessment of perfectionism
and assessment of outcome.
**p < 0.01. ***p < 0.001.

The analysis restricted to nonclinical samples yielded significant negative regression co-
efficients for perfectionistic strivings and HMPS-other-oriented perfectionism, indicating a
weaker link of psychopathology to these two domains than to perfectionistic concerns in
nonclinical samples. It is important to note that fewer moderators could be evaluated in the
clinical group compared to nonclinical group because there were fewer studies on clinical sam-
ples. This fact may serve to explain the differences between the two groups. For all models of
the meta-regression procedures, no additional perfectionism dimension or study characteristic
accounted for a significant amount of variance in effect sizes.
Finally to evaluate the role of the dimensions of perfectionism in the context of existing
specific psychological disorders, one additional set of meta-regression models was computed.
This implied one additional meta-regression per disorder; all models aimed to predict the overall
effect of perfectionism on psychological disorders controlling for study characteristics. The only
exception was the analysis for depression, which contained fewer than 10 tests, and because of
the small sample size, it was deemed inappropriate to conduct an analysis. Anorexia nervosa
and bulimia nervosa were subsumed into a single “eating disorders” category because of a low
number of tests for each disorder in the sample of studies. The reference category in each subset
was overall perfectionistic concerns for the dummy-coded variable perfectionism. Results are
depicted in Table 6.
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Perfectionism and Psychopathology 1317

Table 6
Coefficients and Robust Standard Errors From Nested Meta-Regression Models Predicting Effect
Sizes of the Relationship Between Perfectionism and Psychopathology by the Subscales of FMPS
and HMPS and Sample Characteristics, Restricted to Effect Sizes on Samples With Anxiety
Disorders, OCD, and Eating Disorders, Respectively

Model

Variable B SE β

Anxiety disorders Perfectionistic strivings −0.60* 0.24 −0.26

(k = 20, n = 127) FMPS-concern over mistakes 1.33** 0.32 0.37
FMPS-doubts about actions 0.04 0.76 0.01
HMPS-socially prescribed perfectionism 0.42 0.28 0.19
FMPS-organization −2.03* 0.50 −0.53
HMPS-other-oriented perfectionism −0.06 0.20 −0.02
Age < 0.001 < 0.001 0.04
Gender (% female) < 0.001 < 0.001 −0.05
Outcome time (months) 0.02 0.01 0.17
OCD (k = 12, Perfectionistic strivings 1.49 0.55 0.41
n = 51) FMPS-concern over mistakes −10.48* 3.06 −1.53
FMPS-doubts about actions 8.83* 2.05 1.55
HMPS-self-oriented perfectionism −2.28 1.29 −0.28
FMPS-organization −4.77* 1.57 −1.02
Age −0.10* 0.03 −2.46
Gender (% female) −0.02 0.01 −1.26
Eating disorders Perfectionistic strivings −1.11 3.40 −0.17
(k = 12, n = 46) FMPS-concern over mistakes 0.26 2.20 0.04
FMPS-organization −1.13 5.05 −0.12
Age 0.04 0.09 0.25
Gender (% female) < 0.001 0.01 0.08
Outcome time (months) −0.27 0.33 −0.18

Note. The reference category in all models was perfectionistic concerns. Coefficients shown for the between-
study effects of variables that varied within and between studies. k = number of tests; n = number of
effect sizes; B = unstandardized regression coefficient; SE = standard error of B; β = standardized regres-
sion coefficient; outcome time = months between baseline assessment of perfectionism and assessment of
outcome.
***p < 0.001.

Perfectionistic strivings, FMPS-concern over mistakes, and FMPS-organization accounted

for a significant amount of variance in anxiety disorders, with perfectionistic strivings and
FMPS-organization exhibiting negative effects and FMPS-concern over mistakes exhibiting
positive effects compared to perfectionistic concerns. For OCD, the association of FMPS-
concerns over mistakes, FMPS-organization, and age to OCD was significantly lower than the
correlation between perfectionistic concerns and OCD, while the association between FMPS-
doubts about actions and OCD was significantly higher. Among eating disorders, none of
the perfectionism dimensions or study characteristics accounted for a significant amount of
variance.1
To determine whether the chosen reference category affected the findings, sensitivity analyses
were conducted by (a) varying the reference categories in each model and (b) not controlling for
the two broad perfectionism dimensions perfectionistic concerns and perfectionistic strivings.

1 Inaddition to the meta-regression analyses presented here, we also conducted nested meta-regression
models predicting effect sizes restricted to studies reporting on perfectionistic striving versus perfectionistic
concerns. However, these analyses did not yield any meaningful results and are thus not presented here.
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1318 Journal of Clinical Psychology, October 2017

These analyses had no major effect on the findings, i.e., the direction of relationship and the
statistical significance remained unchanged, and so the interpretations of the findings are based
on the results presented above.

Discussion
The purpose of this meta-analysis was to investigate the relationships between perfectionism
domains and psychopathology, the relative contribution of the subscales of FMPS and HMPS
in the prediction of psychopathology, and to examine the effects of candidate moderators of the
effects of perfectionism dimensions on psychopathological outcomes.

Perfectionistic Concerns Versus Perfectionistic Strivings in Relation to Psychopathology

The main finding was that both dimensions of perfectionism were associated with psychopathol-
ogy outcomes across studies. In the majority of outcomes in which the results of the path
analysis are considered, perfectionistic strivings was less related to psychopathology than per-
fectionistic concerns, particularly in nonclinical populations. This finding supports the view of
previous authors (e.g., Stoeber & Otto, 2006). Among the investigated clinical disorders, only
eating disorders were an exception because both dimensions were strongly related to pathol-
ogy, as shown by the examination of weighted averaged correlation coefficients and unique
effects.
Furthermore, the two dimensions of perfectionism were overall positively correlated. This
finding is important for theory as it indicates substantial overlap in the dimensions and in-
dicates that zero-order correlations between these dimensions and outcomes may provide a
misleading representation of the strength and pattern of effects. It highlights the need to
account for unique effects of these constructs when predicting outcomes. The findings are
also important for interventions and indicate that although both components of perfection-
ism should be targeted for outcomes relating to eating disorders, the focus of reducing symp-
toms of OCD, anxiety disorders, and depression should be on perfectionistic concerns because
these dimensions seem to contribute most in explaining variance in these psychopathological
outcomes.
The positive associations between the perfectionism dimensions and outcomes across all
domains of psychopathology are consistent with findings that perfectionism is not specific to
certain disorders or symptoms (e.g., Frost & Steketee, 1997; Rhéaume, Freeston, Dugas, Letarte,
& Ladouceur, 1995) and is a transdiagnostic process (Egan et al., 2011).

Investigating the Subscales of FMPS and HMPS

Regarding the particular subscales of FMPS and HMPS in the context of psychopathol-
ogy, several findings were observed. There were higher scores on scales that load on perfec-
tionistic concerns, including FMPS-concern over mistakes, FMPS-doubts about actions, and
HMPS-socially prescribed perfectionism than on scales loading on perfectionistic strivings, such
as FMPS-personal standards and HMPS-self-oriented perfectionism for depression, anxiety
disorders, and OCD, supporting previous research (Bieling, Israeli, et al., 2004; Frost et al.,
1993). However, this finding was not confirmed in the meta-regression likely because of the fact
that scores on some subscales were not reported in a sufficient number of scales and thus could
not be evaluated in the meta-regression. To draw final conclusions in a meta-regression analysis,
more studies reporting the effects of the different subscales of the two measures in the context
of the particular disorders are needed.
For depression, the tendency of higher scores on HMPS-self-oriented perfectionism com-
pared to FMPS-personal standards was revealed, consistent with research reporting high scores
on HMPS-self-oriented perfectionism in depression (Hewitt & Flett, 1991a) and low scores on
FMPS-personal standards (Lombardo et al., 2013). This suggests that the assessment of per-
fectionism in the context of depression may concentrate on HMPS-self-oriented perfectionism
instead of FMPS-personal standards. This conclusion is drawn with caution, however, because
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Perfectionism and Psychopathology 1319

depression was not evaluated in a sufficient number of studies to conduct a meta-regression

restricted to effect sizes regarding depression.
The meta-regression analyses restricted to eating disorders revealed a tendency for scores
on FMPS-concern over mistakes to be lower in eating disorders compared to perfectionistic
concerns, indicating that the latter combination of scales may be more suitable to assess perfec-
tionism in eating disorders instead of single subscales like FMPS-concern over mistakes. Further,
among OCD, FMPS-doubts about actions tended to be highly correlated to the presence of OCD
as hypothesized, supporting previous findings of pronounced scores on FMPS-doubts about ac-
tions in OCD (Antony et al., 1998; Frost & Steketee, 1997). This is not surprising given that
some items comprising doubts about actions were taken from a measure of OCD symptoms,
and thus this subscale likely overlaps with OCD symptoms.
An additional finding was that HMPS-other-oriented perfectionism consistently explained
significant amounts of variance; it was associated with a lower effect size in every meta-regression
analysis in which it was evaluated. Although it has been stated that HMPS-other-oriented
perfectionism is somewhat different from the other perfectionism dimensions (Stoeber & Otto,
2006), the present findings support the inclusion of HMPS-other-oriented perfectionism in the
broad dimension perfectionistic strivings, as proposed by (Bieling, Israeli, et al., 2004). However,
in the evaluation of weighted average effect sizes HMPS-other-oriented perfectionism was not
significantly correlated with most of the outcomes except for global eating pathology, dietary
restraint, and deliberate self-harm. Future research is needed to investigate the relationship of
HMPS-other-oriented perfectionism to psychopathology in more detail and determine whether
or not it can be subsumed with the remaining dimensions.
In summary, the two main dimensions of perfectionism (perfectionistic concerns and perfec-
tionistic strivings) consistently explained significant amounts of variance, whereas the subscales
from different inventories could not explain variance in most of the meta-regression models,
other than those mentioned above. Thus, the formation of the two main dimensions as involving
various subscales is supported; evaluating the single subscales separately can, however, give
insight into specific patterns of perfectionism in certain types of psychopathology.

Implications
In terms of theoretical implications, the findings suggest that perfectionism needs to be con-
sidered in the context of a variety of disorders. As outlined, a transdiagnostic process is a
one that is involved in the maintenance of multiple psychological disorders (Harvey et al.,
2004). Egan et al. (2011) argued that perfectionism is a transdiagnostic process because it
is (a) elevated across eating disorders, anxiety disorders, OCD, and depression compared
to healthy controls; (b) a risk and maintaining factor across disorders; and (c) associated
with co-occurring psychological disorders. Consistent with this is a study of 345 people with
co-occurring anxiety and mood disorders that found the number of diagnoses was posi-
tively correlated with perfectionism and that perfectionism predicted higher co-occurrence
of disorders even after controlling for symptoms (Bieling, Summerfeldt, Israeli, & Antony,
2004).
Bieling and colleagues (Bieling, Summerfeldt, et al., 2004) concluded that treating perfec-
tionism will be more beneficial in patients with co-occurring disorders than disorder spe-
cific treatments and may result in symptom reduction across multiple disorders. The cur-
rent meta-analytic findings support the assertion based on the previous narrative review of
Egan et al. (2011), that perfectionism is a shared etiological factor in OCD, anxiety disor-
ders, depression, and eating disorders. Future research should examine the reason for this,
for example, by considering factors including whether it is the relationship of perfection-
ism to comorbidity or shared symptoms between disorders and how perfectionism is an
etiological factor across disorders. There are many potential etiological factors for per-
fectionism, such as parental and cognitive factors, that may be relevant and some have
been described elsewhere (e.g., Maloney, Egan, Kane, & Rees, 2014); however, others such
as an overall deficit in one’s sense of self or core low self-esteem may also be useful to
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1320 Journal of Clinical Psychology, October 2017

investigate further to inform the theoretical understanding of how perfectionism is a trans-

diagnostic process.
The clinical implications of the study are speculative based on our data. In summary, our
findings suggest that although perfectionism is not specific to a particular form of psychopathol-
ogy, decreasing perfectionism through CBT for perfectionism could be beneficial in the context
of a variety of psychopathological outcomes given there is meta-analytic evidence for efficacy in
a range of symptoms across disorders (Lloyd et al., 2015). Further, evaluating specific patterns
of perfectionism and comparing scores on subscales of multidimensional measures could help
develop a more detailed picture of specific cognitions and inform treatment in a more targeted
manner, rather than through concentrating solely on the broad perfectionistic concerns and per-
fectionistic strivings dimensions. Current treatment approaches in CBT for perfectionism (Egan,
Wade, et al., 2014) emphasize the reduction of perfectionistic concerns but not perfectionistic
strivings.
This approach appears to be supported for clients with diagnoses of OCD, anxiety
disorders, and depression given the evidence we found for perfectionistic concerns having a
stronger relationship with these disorders than perfectionistic strivings, which had small or triv-
ial associations. However, given our findings that perfectionistic strivings was strongly associated
with eating disorder outcomes along with perfectionistic concerns, it appears that a different
approach may be required when targeting clients presenting for perfectionism treatment who
meet a diagnosis of an eating disorder. The current treatment emphasis has been on explicitly
stating to the client early in therapy that there is nothing wrong with striving for standards
in itself (i.e., perfectionistic strivings), but it is the concern over mistakes (i.e., perfectionistic
concerns) and basing ones self-worth on striving and achievement that is problematic (Egan,
Wade, et al., 2014).
Given our findings it would be useful for future research to determine if changing this
approach for those presenting with eating disorders would be more effective than the current
treatment, specifically where it is investigated if modifying treatment to reduce perfectionistic
strivings results in stronger effects in reduction of eating disorder symptoms. This would also
be an interesting line of research to examine regarding eating disorder prevention, in which
perfectionism has been a recent focus of interest, to determine if prevention programs for
perfectionism should be modified to explicitly focus on reducing both perfectionistic strivings
and perfectionistic concerns.

Strengths and Limitations

The key strength of the current analysis was the comprehensive literature search and inclusion
criteria for studies on perfectionism and the adoption of meta-analytic techniques to estimate
bias-corrected tests of relations among perfectionism and psychopathology outcomes across
the extant literature. A further advantage is the adoption of state-of-the-art meta-analytic tech-
niques using robust variance estimation (Hedges et al., 2010), which allowed including multiple
effect sizes from single studies while controlling for data dependencies. Furthermore, through
the simultaneous evaluation of a variety of psychological disorders, symptoms, and outcomes
related to psychopathology this study has provided the first meta-analysis to enable a detailed
understanding of the role of perfectionism in a range of psychopathology.
A limitation was the various measures used for perfectionism and outcomes, and thus assessing
the relationship was more difficult than if the same measures had been used. This is likely to have
introduced further methodological variance. Moreover, the variety of investigated outcomes led
to the need to categorize them into broader groups. For example, anxiety disorders had to be
subsumed into one category instead of keeping the different anxiety disorders separate. Thus,
possible differences between the anxiety disorders have not been addressed in this review. A
similar issue appeared for OCD and eating disorders: We were unable to draw conclusions on
specific subtypes of OCD or specific eating disorders like anorexia and bulimia nervosa in the
meta-regression analyses because too few studies in the current sample conducted analysis of
the respective subtypes separately.
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Perfectionism and Psychopathology 1321

In addition, the clinical meaning of the present findings is weakened by the fact that
more studies on nonclinical than on clinical samples were included. This fact may also serve
to help explaining the differences found in the meta-regression models restricted to clinical
versus nonclinical samples: The tendency of a lower influence of perfectionistic strivings in clin-
ical samples was not significant, possibly because there was a lower number of studies in the
clinical samples and because of the same reason that some subscales could not be evaluated in
clinical samples at all.
Although we examined gender as a moderator, we were unable to draw valid conclusions
because the majority of the samples (74%) were female. In particular, we were unable to examine
perfectionism in young men because there was low percentage of male samples in the studies
collected. Future research investigating relations between perfectionism and outcomes in men
is advocated to develop an evidence base to allow for better tests of gender differences in this
literature.
Another limitation was that although we argued that identifying which perfectionism measure
exhibits the strongest relationship with psychopathology is important, because it may inform
clinical interventions, a problem with this is that it is possible that the correlation between
perfectionism measures and symptoms may be due to shared method variance (i.e., both are
single informant and self-report) or some measures potentially having overlap with the specific
symptoms of the disorder. This is the case particularly when considering the strong association
between doubts about actions and OCD; as outlined previously, this perfectionism subscale has
been criticized as being highly overlapping with OCD symptoms (Shafran & Mansell, 2001).
While we do not believe this is the case for all measures of perfectionism, because some do
not overlap with psychopathology symptoms, it is possible that some of our results such as
those in OCD may have inflated associations due to this overlap and this is a limitation to be
acknowledged.
A further limitation of this study is that only 18 of the 284 studies that were included were
longitudinal, highlighting that a problem with the current literature on perfectionism, as is the
case with most psychological research, that the data for the field, on the whole, does not adopt
strong, longitudinal designs. Correlational designs hold back the field of research and limit the
inferences that researchers and practitioners can make with respect to understanding theory and
interventions.
While one recent meta-analysis exists looking at longitudinal studies of depression in perfec-
tionism (Smith et al., 2016), given we followed the guideline that ten primary studies are needed
to evaluate moderator variables (Dalton & Dalton, 2008), we were unable to conduct separate
analyses predicting longitudinal development for the single clinical disorders or symptom cat-
egories which is a limitation. Consequently, no conclusion as to whether perfectionism is as a
risk or maintenance factor for psychopathology can be made because of the fact that insufficient
longitudinal studies about the relationship between psychological disorders and perfectionism
exist. Therefore, it is not possible to say whether perfectionism leads to the onset of particular
disorders or symptom.
Thus, conclusions on whether reducing perfectionism would reduce symptoms of a disorder
are not possible. We issue a call to the field to reduce the use of cross-sectional research and
instead encourage future researchers to conduct longitudinal research, most importantly, longi-
tudinal, cross-lagged panel designs and experimental research looking to change perfectionism
dimensions and observe the effects on psychopathology outcomes.
Concerning the analysis of single subscales of FMPS and HMPS, the mostly nonsignificant
findings may be due to the fact that there was more power to detect differences with the
dimensional approach than the subscale approach because there were more studies for the
former. Another limitation was that we included only published or in press studies to minimize
potential duplication of findings. However, this approach may also mean that we may omitted
studies that have been completed but not yet published or still under review. However, it is
important to note that we inspected the funnel plot of study precision against effect size and
applied Egger et al.’s (1997) regression techniques to identify potential small study or publication
bias in the current set of studies.
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1322 Journal of Clinical Psychology, October 2017

Finally, another limitation is that although we included a measure of self-criticism (DAS-

SC; Weissman & Beck, 1978), which has been used in numerous studies examining the link
between perfectionism and psychopathology (e.g., Dunkley, Sanislow, et al., 2006; Dunkley
et al., 2009), there are other scales we did not include, such as the Depressive Experiences
Questionnaire (DEQ; Blatt, D’Aflitti, & Quinlan, 1979) and the Sociotropy-Autonomy Scale
(SAS; Beck, Epstein, Harrison, & Emery, 1983), which have been found in some factor analytic
studies to load on to a self-critical perfectionism factor (e.g., Dunkley, Blankstein, Zuroff,
Lecce, & Hui, 2006). We did not include these other scales because we focused specifically on
perfectionistic strivings and perfectionistic concerns and on measures specifically designed to
assess perfectionism (e.g. MPS; Frost et al., 1990; Hewitt & Flett, 1991b), and our inclusion
criteria reflect this. An analysis of all measures of self-criticism was beyond the scope of the
current analysis; however, we look to future research to expand current findings to include other
measures such as the DEQ and SAS, which share conceptual overlap with perfectionism.

Directions for Future Research

Because the current meta-analysis was conducted on studies that mostly used nonclinical samples
and thus concentrated on symptoms of disorders, it would be important to again investigate
the relation of perfectionism to clinical diagnoses in addition to symptoms when sufficient data
become available. Thus, although many primary studies on the role of perfectionism in the
context of specific disorders exist, further studies are needed. This would allow more meaningful
conclusions on the clinical role of perfectionism.
It would be worthwhile to establish the role of perfectionism as a risk or maintenance factor in
the context of different disorders to propose suggestions for prevention and treatment of a variety
of disorders. This could be done by further examination of longitudinal associations between
perfectionism and clinical outcomes using meta-analytic methods. As outlined, a problem in
the research area is the preponderance of cross-sectional studies relative to longitudinal studies,
and future researchers are encouraged to consider longitudinal designs to provide more robust
evidence of the relationships between perfectionism and psychopathology.
The present review did not evaluate moderators of the relationship between perfectionism and
psychopathology, such as duration of illness, treatment seeking, and methodology; further, as
stated before, age and gender could not be evaluated sufficiently because of the low proportions
of male participants and similar age groups in primary studies. More primary studies giving
information and statistical variance on these characteristics as well as studies with wider age
spans and the inclusion of male participants are needed to allow insight by way of meta-analysis.
Another direction for future research would be to consider whether being high on both per-
fectionistic concerns and perfectionistic strivings at the same time confers a “dual” vulnerability
in the sense of being elevated on both dimensions of perfectionism. This may explain why perfec-
tionistic strivings is more strongly linked to psychopathology in clinical samples, largely relating
to eating disorder outcomes, that is, individuals with clinical disorders may show elevations in
both perfectionism dimensions, and this differentiates them from control samples.
Further meta-analytic research would be useful to evaluate the efficacy of existing interven-
tions for perfectionism. While there has been one systematic review (Lloyd et al., 2015) that
examined CBT for perfectionism, which found large pooled effect sizes for reductions in perfec-
tionism (FMPS-personal standards, concern over mistakes, HMPS-self-oriented perfectionism)
and medium pooled effect sizes for reductions in anxiety and depression, only eight studies were
included; further randomized controlled trials that have now been published (e.g., Egan, Wade,
et al., 2014; Handley et al., 2014) would be useful to include in an updated meta-analysis on
treatment efficacy.
There are psychological disorders that have just started to gain attention in the context
of perfectionism and could not be included in this review because of the low numbers of
primary studies, for instance, obsessive-compulsive personality disorder, posttraumatic stress
disorder, and body dysmorphic disorder. Extending research on these would help to gain a
better understanding of the role of perfectionism in various outcomes. The scope of the present
research was not wide enough to evaluate findings specific to particular populations, such
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Perfectionism and Psychopathology 1323

as athletes, or perfectionism in other domains, such as dyadic perfectionism, and thus they
could be evaluated in future meta-analyses. Synthesis of such research will lead to far-reaching
conclusions for distinct areas, thus shedding further light on the role of perfectionism and the
need for effective prevention and intervention.

Conclusion
This is the first meta-analysis of tests of effects between perfectionism and psychopathology.
Findings indicate substantial overlap in the two perfectionism dimensions in the context of
various psychological disorders, their symptoms, and outcomes related to psychopathology.
The application of meta-analytic methods offered the chance to resolve inconsistencies observed
in the literature attributable to methodological artefacts.

References
References included in the meta-analysis are provided in the online supplementary material.
Antony, M. M., Purdon, C. L., Huta, V., & Swinson, R. P. (1998). Dimensions of perfectionism
across the anxiety disorders. Behaviour Research and Therapy, 36(12), 1143–1154. doi:10.1016/S0005-
7967%2898%2900083-7
Bardone-Cone, A. M., Abramson, L. Y., Vohs, K. D., Heatherton, T. F., & Joiner, T. E. (2006). Predicting
bulimic symptoms: An interactive model of self-efficacy, perfectionism, and perceived weight status.
Behaviour Research and Therapy, 44(1), 27–42. doi:10.1016/j.brat.2004.09.009
Bardone-Cone, A. M., Wonderlich, S. A., Frost, R. O., Bulik, C. M., Mitchell, J. E., Uppala, S., & Simonich,
H. (2007). Perfectionism and eating disorders: Current status and future directions. Clinical Psychology
Review, 27(3), 384–405.
Beck, A. T., Epstein, N., Harrison, R. P., & Emery, G. (1983). Development of the Sociotropy-Autonomy
Scale: A measure of personality factors in psychopathology. Unpublished manuscript, University of
Pennsylvania, Philadelphia.
Békés, V., Dunkley, D. M., Taylor, G., Zuroff, D. C., Lewkowski, M., Foley, J. E., & Westreich, R. (2015).
Chronic stress and attenuated improvement in depression over 1 year: The moderating role of perfec-
tionism. Behavior Therapy, 46(4), 478–492.
Bieling, P. J., Israeli, A. L., & Antony, M. M. (2004). Is perfectionism good, bad, or both? Examin-
ing models of the perfectionism construct. Personality and Individual Differences, 36(6), 1373–1385.
doi:10.1016/S0191-8869%2803%2900235-6
Bieling, P. J., Summerfeldt, L. J., Israeli, A. L., & Antony, M. M. (2004). Perfectionism as an explanatory
construct in comorbidity of Axis I disorders. Journal of Psychopathology and Behavioral Assessment,
26(3), 193–201.
Black, K. (2009). Business statistics: Contemporary decision making (6th ed.). New York: Wiley.
Blatt, S. J., D’Aflitti, J. P., & Quinlan, D. M. (1979). Depressive Experiences Questionnaire. Unpublished
manuscript. New Haven, CT.
Borenstein, M., Hedges, L., Higgins, J., & Rothstein, H. (2005). Comprehensive meta-analysis version 2.
Englewood, NJ: Biostat.
Borenstein, M., Hedges, L. V., Higgins, J. P., & Rothstein, H. R. (2009). Introduction to meta-analysis. West
Sussex, UK: John Wiley & Sons, Ltd.
Brannan, M. E., & Petrie, T. A. (2008). Moderators of the body dissatisfaction-eating disorder symp-
tomatology relationship: Replication and extension. Journal of Counseling Psychology, 55(2), 263–
275.
Buhlmann, U., Etcoff, N. L., & Wilhelm, S. (2008). Facial attractiveness ratings and perfectionism in body
dysmorphic disorder and obsessive-compulsive disorder. Journal of Anxiety Disorders, 22(3), 540–547.
doi:10.1016/j.janxdis.2007.05.004
Cockell, S. J., Hewitt, P. L., Seal, B., Sherry, S., Goldner, E. M., Flett, G. L., & Remick, R. A. (2002). Trait
and self-presentational dimensions of perfectionism among women with anorexia nervosa. Cognitive
Therapy and Research, 26(6), 745–758. doi:10.1023/A:1021237416366
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence
Earlbaum Associates.
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1324 Journal of Clinical Psychology, October 2017

Dalton, D. R., & Dalton, C. M. (2008). Meta-analyses some very good steps toward a bit longer journey.
Organizational Research Methods, 11(1), 127–147.
Dunkley, D. M., Blankstein, K. R., Zuroff, D. C., Lecce, S., & Hui, D. (2006). Self-critical and personal
standards factors of perfectionism located within the five-factor model of personality. Personality and
Individual Differences, 40(3), 409–420.
Dunkley, D. M., Sanislow, C. A., Grilo, C. M., & McGlashan, T. H. (2006). Perfectionism and
depressive symptoms 3 years later: Negative social interactions, avoidant coping, and perceived
social support as mediators. Comprehensive Psychiatry, 47(2), 106–115. doi:10.1016/
j.comppsych.2005.06.003
Dunkley, D. M., Sanislow, C. A., Grilo, C. M., & McGlashan, T. H. (2009). Self-criticism versus neuroticism
in predicting depression and psychosocial impairment for 4 years in a clinical sample. Comprehensive
Psychiatry, 50(4), 335–346.
Egan, S. J., Hattaway, M., & Kane, R. T. (2014). The relationship between perfectionism and rumination in
post traumatic stress disorder. Behavioural and Cognitive Psychotherapy, 42(2), 211–223.
Egan, S. J., Wade, T. D., & Shafran, R. (2011). Perfectionism as a transdiagnostic process: A clinical review.
Clinical Psychology Review, 31(2), 203–212. doi:10.1016/j.cpr.2010.04.009
Egan, S. J., Wade, T., Shafran, R., & Antony, M. M. (2014). Cognitive-behavioral treatement for perfection-
ism. New York: Guilford Publications.
Egger, M., Smith, G. D., Schneider, M., & Minder, C. (1997). Bias in meta-analysis detected by a simple,
graphical test. BMJ, 315(7109), 629–634.
Enns, M. W., Cox, B. J., & Borger, S. C. (2001). Correlates of analogue and clinical depression: A further
test of the phenomenological continuity hypothesis. Journal of Affective Disorders, 66(2-3), 175–183.
doi:10.1016/S0165-0327%2800%2900305-0
Enns, M. W., Cox, B. J., Sareen, J., & Freeman, P. (2001). Adaptive and maladaptive perfectionism in
medical students: A longitudinal investigation. Medical Education, 35(11), 1034–1042. doi:10.1046/
j.1365-2923.2001.01044.x
Fairburn, C. G., Cooper, Z., & Shafran, R. (2003a). Cognitive behaviour therapy for eating disorders: A
"transdiagnostic" theory and treatment. Behaviour Research and Therapy, 41(5), 509–528.
Fairburn, C. G., Cooper, Z., & Shafran, R. (2003b). Cognitive behaviour therapy for eating disorders: A
“transdiagnostic” theory and treatment. Behaviour Research and Therapy, 41(5), 509–528.
Flett, G., Hewitt, P., Boucher, D., Davidson, L., & Munro, Y. (1997). The child-adolescent perfection-
ism scale: Development, validation, and association with adjustment. Unpublished manuscript. York
University, York, UK.
Frost, R. O., Heimberg, R. G., Holt, C. S., & Mattia, J. I. (1993). A comparison of two measures of perfection-
ism. Personality and Individual Differences, 14(1), 119–126. doi:10.1016/0191-8869%2893%2990181-2
Frost, R. O., Marten, P., Lahart, C., & Rosenblate, R. (1990). The dimensions of perfectionism. Cognitive
Therapy and Research, 14(5), 449–468.
Frost, R. O., & Steketee, G. (1997). Perfectionism in obsessive-compulsive disorder patients. Behaviour
Research and Therapy, 35(4), 291–296. doi:http://dx.doi.org/10.1016/S0005-7967%2896%2900108-8
Garner, D. M., Olmstead, M. P., & Polivy, J. (1983). Development and validation of a multidimensional
eating disorder inventory for anorexia nervosa and bulimia. International Journal of Eating Disorders,
2(2), 15–34.
Gnilka, P. B., Ashby, J. S., & Noble, C. M. (2012). Multidimensional perfectionism and anxiety: Differences
among individuals with perfectionism and tests of a coping-mediation model. Journal of Counseling &
Development, 90(4), 427–436. doi:10.1002/j.1556-6676.2012.00054.x
Graham, A. R., Sherry, S. B., Stewart, S. H., Sherry, D. L., McGrath, D. S., Fossum, K. M., & Allen, S.
L. (2010). The existential model of perfectionism and depressive symptoms: A short-term, four-wave
longitudinal study. Journal of Counseling Psychology, 57(4), 423–438.
Hagger, M. S., & Chatzisarantis, N. L. (2014). It is premature to regard the ego-depletion effect as “Too
Incredible.” Frontiers in Psychology, 5.
Halmi, K. A., Sunday, S. R., Strober, M., Kaplan, A., Woodside, D. B., Fichter, M., . . . Kaye, W. H. (2000).
Perfectionism in anorexia nervosa: Variation by clinical subtype, obsessionality, and pathological eating
behavior. American Journal of Psychiatry, 157(11), 1799–1805.
Handley, A. K., Egan, S. J., Kane, R. T., & Rees, C. S. (2014). The relationships between perfectionism,
pathological worry and generalised anxiety disorder. BMC Psychiatry, 14(1).
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Perfectionism and Psychopathology 1325

Harvey, A., Watkins, E., Mansell, W., & Shafran, R. (2004). Cognitive behavioural processes across psycho-
logical disorders: A transdiagnostic approach to research and treatment. Oxford, UK: Oxford University
Press.
Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis. New York: Academic Press.
Hedges, L. V., Tipton, E., & Johnson, M. C. (2010). Robust variance estimation in meta-regression with
dependent effect size estimates. Research Synthesis Methods, 1(1), 39–65.
Hewitt, P. L., & Flett, G. L. (1991a). Dimensions of Perfectionism in Unipolar Depression. Journal of
Abnormal Psychology, 100(1), 98–101.
Hewitt, P. L., & Flett, G. L. (1991b). Perfectionism in the self and social contexts: Conceptualization,
assessment, and association with psychopathology. Journal of Personality and Social Psychology, 60(3),
456–470.
Hewitt, P. L., Flett, G. L., & Ediger, E. (1996). Perfectionism and depression: Longitudinal assessment of a
specific vulnerability hypothesis. Journal of Abnormal Psychology, 105(2), 276–280. doi:10.1037/0021-
843X.105.2.276
Hodgson, R. J., & Rachman, S. (1977). Obsessional compulsive complaints. Behaviour Research and
Therapy, 15, 389–395.
Huprich, S. K., Porcerelli, J., Keaschuk, R., Binienda, J., & Engle, B. (2008). Depressive personality dis-
order, dysthymia, and their relationship to perfectionism. Depression and Anxiety, 25(3), 207–217.
doi:10.1002/da.20290
Iketani, T., Kiriike, N., Stein, M. B., Nagao, K., Nagata, T., Minamikawa, N., . . . Fukuhara, H. (2002).
Relationship between perfectionism and agoraphobia in patients with panic disorder. Cognitive Be-
haviour Therapy, 31(3), 119–128. doi:10.1080/165060702320337997
Juster, H. R., Heimberg, R. G., Frost, R. O., Holt, C. S., Mattia, J. I., & Faccenda, K. (1996). Social
phobia and perfectionism. Personality and Individual Differences, 21(3), 403–410. doi:10.1016/0191-
8869(96)00075-x
Lilenfeld, L., Stein, D., Bulik, C., Strober, M., Plotnicov, K., Pollice, C., . . . Kaye, W. H. (2000). Personality
traits among current eating disordered, recovered and never ill first-degree female relatives of bulimic
and control women. Psychological Medicine, 30(6), 1399–1410. doi:10.1017/S0033291799002792
Lloyd, S., Schmidt, U., Khondoker, M., & Tchanturia, K. (2015). Can psychological interventions reduce
perfectionism? A systematic review and meta-analysis. Behavioural and cognitive psychotherapy, 43(6),
705–731.
Lombardo, C., Mallia, L., Battagliese, G., Grano, C., & Violani, C. (2013). Perfectionism mediates the
relationship between insomnia and depressive symptoms. Sleep and Biological Rhythms, 11(2), 90–98.
Maloney, G. K., Egan, S. J., Kane, R. T., & Rees, C. S. (2014). An etiological model of perfectionism. PloS
One, 9(5), e94757.
Miller-Day, M., & Marks, J. D. (2006). Perceptions of parental communication orientation, perfection-
ism, and disordered eating behaviors of sons and daughters. Health Communication, 19(2), 153–163.
doi:10.1207/s15327027hc1902_7
Moor, S., Vartanian, L. R., Touyz, S. W., & Beumont, P. (2004). Psychopathology of EDNOS patients: To
whom do they compare? Clinical Psychologist, 8(2), 70–75. doi:10.1080/1328420041233130436
Norman, R. M. G., Davies, F., Nicholson, I. R., Cortese, L., & Malla, A. K. (1998). The relationship of two
aspects of perfectionism with symptoms in a psychiatric outpatient population. Journal of Social and
Clinical Psychology, 17(1), 50–68. doi:10.1521/jscp.1998.17.1.50
Obsessive Compulsive Cognitions Working Group. (2001). Development and initial validation of the
Obsessive Beliefs Questionnaire and the Interpretation of Intrusions Inventory. Behaviour Research
and Therapy, 39(8), 987–1006.
Rhéaume, J., Freeston, M. H., Dugas, M. J., Letarte, H., & Ladouceur, R. (1995). Perfectionism, responsi-
bility and obsessive-compulsive symptoms. Behaviour Research and Therapy, 33(7), 785–794.
Rhéaume, J., Freeston, M. H., Ladouceur, R., Bouchard, C., Gallant, L., Talbot, F., & Vallières, A. (2000).
Functional and dysfunctional perfectionists: are they different on compulsive-like behaviors? Behaviour
Research and Therapy, 38(2), 119–128.
Rice, K. G., Kubal, A. E., & Preusser, K. J. (2004). Perfectionism and children’s self-concept: Further
validation of the Adaptive/Maladaptive Perfectionism Scale. Psychology in the Schools, 41(3), 279–290.
doi:10.1002/pits.10160
Rodriguez-Seijas, C., Eaton, N. R., & Krueger, R. F. (2015). How transdiagnostic factors of personality and
psychopathology can inform clinical assessment and intervention. Journal of Personality Assessment,
97(5), 425–435.
 10974679, 2017, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jclp.22435 by Universidad De La Laguna, Wiley Online Library on [16/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1326 Journal of Clinical Psychology, October 2017

Rosenthal, R. (1979). The file drawer problem and tolerance for null results. Psychological Bulletin, 86(3),
638.
Saboonchi, F., Lundh, L.-G., & Ost, L.-G. (1999). Perfectionism and self-consciousness in social phobia and
panic disorder with agoraphobia. Behaviour Research and Therapy, 37(9), 799–808. doi:10.1016/S0005-
7967%2898%2900183-1
Sassaroli, S., Lauro, L. J. R., Ruggiero, G. M., Mauri, M. C., Vinai, P., & Frost, R. (2008). Perfectionism in
depression, obsessive-compulsive disorder and eating disorders. Behaviour Research & Therapy, 46(6),
757–765. doi:10.1016/j.brat.2008.02.007
Schmidt, U., & Treasure, J. (2006). Anorexia nervosa: Valued and visible. A cognitive-interpersonal mainte-
nance model and its implications for research and practice. British Journal of Clinical Psychology, 45(3),
343–366.
Shafran, R., Cooper, Z., & Fairburn, C. G. (2003). “Clinical perfectionism” is not “multidimensional
perfectionism”: A reply to Hewitt, Flett, Besser, Sherry & McGee. Behaviour Research and Therapy,
41, 1217–1220.
Shafran, R., & Mansell, W. (2001). Perfectionism and psychopathology: A review of research and treatment.
Clinical Psychology Review, 21(6), 879–906. doi:10.1016/S0272-7358%2800%2900072-6
Slaney, R. B., Rice, K. G., Mobley, M., Trippi, J., & Ashby, J. S. (2001). The Revised Almost Perfect Scale.
Measurement and Evaluation in Counseling and Development, 34(3), 130–145.
Smith, M. M., Sherry, S. B., Rnic, K., Saklofske, D. H., Enns, M., & Gralnick, T. (2016). Are perfectionism
dimensions vulnerability factors for depressive symptoms after controlling for neuroticism? A meta-
analysis of 10 longitudinal studies. European Journal of Personality, 30(2), 201–212.
Stoeber, J., & Otto, K. (2006). Positive conceptions of perfectionism: Approaches, evidence, challenges.
Personality and social psychology review, 10(4), 295–319.
Tanner-Smith, E. E., Wilson, S. J., & Lipsey, M. W. (2013). The comparative effectiveness of outpatient
treatment for adolescent substance abuse: A meta-analysis. Journal of Substance Abuse Treatment,
44(2), 145–158.
Terry-Short, L., Owens, R. G., Slade, P., & Dewey, M. (1995). Positive and negative perfectionism.
Personality and Individual Differences, 18(5), 663–668.
Viswesvaran, C., & Ones, D. S. (1995). Theory testing: Combining psychometric meta-analysis and structural
equations modeling. Personnel Psychology, 48(4), 865–885.
Weissman, A. N., & Beck, A. T. (1978). Development and validation of the Dysfunctional Attitude Scale:
A preliminary investigation. Paper presented at the meeting of the Association for the Advancement of
Behavior Therapy, Chicago.
Welch, E., Miller, J. L., Ghaderi, A., & Vaillancourt, T. (2009). Does perfectionism mediate or moderate the
relation between body dissatisfaction and disordered eating attitudes and behaviors? Eating Behaviors,
10(3), 168–175. doi:10.1016/j.eatbeh.2009.05.002