Content uploaded by Martin Hagger
Author content
All content in this area was uploaded by Martin Hagger on Jan 30, 2018
Content may be subject to copyright.
PERFECTIONISM AND PSYCHOPATHOLOGY 1
Word count (excluding abstract, acknowledgements, references, figures and tables): 9,256
Short title/running head: PERFECTIONISM AND PSYCHOPATHOLOGY
The Relationship Between Perfectionism and Psychopathology: A Meta-Analysis
Keywords: perfectionism; psychopathology; meta-analysis; perfectionistic striving;
perfectionistic concerns
Karina Limburg1,2*, Hunna J. Watson1,3,4,5, Martin S. Hagger1, Sarah J. Egan1
1 School of Psychology and Speech Pathology, Curtin University, Perth, Australia
2 Clinic for Psychosomatic Medicine and Psychotherapy, University Hospital Klinikum rechts
der Isar, Technische Universität München, 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
*Corresponding author: Karina Limburg, Clinic for Psychosomatic Medicine and
Psychotherapy, University Hospital Klinikum rechts der Isar, Technische Universität
München, Langerstraße 3, 81675 Munich, Germany, (e-mail: karina.limburg@tum.de, phone:
+49 89 4140 4397).
Full citation:
Limburg, K., Watson, H. J., Hagger, M. S., & Egan, S. J. (2016). The relationship between
perfectionism and psychopathology: A meta-analysis. Journal of Clinical Psychology. doi:
10.1002/jclp.22435
PERFECTIONISM AND PSYCHOPATHOLOGY 2
Abstract
Objective: The clinical significance of two 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 psychopathology 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 two perfectionism dimensions with psychopathology outcomes and (b)
subscales of two 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.
Conclusions: Findings support the notion of perfectionism as a transdiagnostic factor by
demonstrating that both dimensions are associated with various forms of psychopathology.
Keywords: perfectionism; psychopathology; meta-analysis; perfectionistic strivings;
perfectionistic concerns
PERFECTIONISM AND PSYCHOPATHOLOGY 3
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 centre on the pursuit of high standards and self-criticism
over not meeting standards, and perfectionism has generally been conceptualised as
multidimensional. 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
consisting of various subscales. Factor analysis of the scales typically results in a two-factor
solution consisting of 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 associated with maladaptive outcomes (e.g., Egan et al.,
2011).
While 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 subscale doubts about actions has substantive, non-trivial overlap with symptoms of
OCD given the majority of items on the subscale were derived from a measure of OCD
symptoms, the MOCI (Hodgson & Rachman, 1977). Consequently it has been argued that the
doubts about actions subscale primarily reflects checking symptoms of OCD, rather than
PERFECTIONISM AND PSYCHOPATHOLOGY 4
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 given 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.
[Insert Table 1 here]
Our hypotheses are based on the accumulating evidence from studies that have shown
dimensions of perfectionism are significantly higher in clinical samples with a range of
disorders compared to controls, and associated with psychopathology in non-clinical 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 has also been linked to generalized anxiety
PERFECTIONISM AND PSYCHOPATHOLOGY 5
disorder (Handley, Egan, Kane, & Rees, 2014), and post-traumatic stress disorder (Egan,
Hattaway, & Kane, 2014). A smaller number of studies has 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). Individuals with anorexia nervosa and bulimia nervosa have
significantly higher scores on perfectionistic 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 non-clinical populations, where perfectionistic
concerns are positively correlated with depressive symptoms (Bieling, Israeli, et al., 2004;
Enns, Cox, 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 non-clinical 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 which have linked perfectionism to the
development of depression. These longitudinal studies are important as they give stronger
evidence for the directional relation between perfectionism and depression since the majority
PERFECTIONISM AND PSYCHOPATHOLOGY 6
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 predicts higher depressive symptoms at one-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, where 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 behaviour that contribute to
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) and
across various disorders (Egan et al., 2011; Shafran & Mansell, 2001), and 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
PERFECTIONISM AND PSYCHOPATHOLOGY 7
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 strivings 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 perfectionistic 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 psychopathology in non-clinical 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 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 as 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
PERFECTIONISM AND PSYCHOPATHOLOGY 8
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 correlated, it is possible that overlap in the dimensions may
give a misleading representation of the true effects of the dimensions. Due to this overlap, the
zero-order effects of perfectionism 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 studies 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 catalyse 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 studies, the time
between baseline assessment of perfectionism and the assessment of the outcome.
Method
Search Strategy
Several strategies were used to identify eligible studies. First, the databases ERIC,
Embase, ISI Web of Science (Science Citation Index Expanded, Social Science Citation
Index Expanded), Medline, PsycINFO, PsycARTICLES, and Scopus were searched for all
years covered through to July 2013. Key words used were: perfectionism, mental health,
outcome, behaviour/behavior, intervention, and psychopathology. Manual searches were
PERFECTIONISM AND PSYCHOPATHOLOGY 9
conducted of reference lists from prior literature reviews and the electronic mailing list
Perfectionism Network Mailing List was queried 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, three broad
categories of outcome were identified: 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 subsequent 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 symptom category, conceptually related
symptoms were aggregated into analysable subcategories. For example, symptoms of social
phobia that were measured in studies included fear of communication situation, fear of
negative evaluation, shyness, social anxiety and social interaction anxiety, all of these were
aggregated into the subcategory symptoms of social phobia. In addition, the OCD symptom
compulsions was formed by integrating various reported compulsions, such as checking,
cleaning, ordering, and washing. Additional clinical outcomes that cannot directly be related
to a disorder such as suicidal ideation and self-harming behaviour were subsumed as
outcomes related to psychopathology. This process of aggregation according to symptoms
was done by the first author and the final author (Egan) who is an experienced Clinical
Psychologist to ensure the classification was clinically meaningful. A summary of all
analysed categories is presented in Figure 1.
PERFECTIONISM AND PSYCHOPATHOLOGY 10
Insert Figure 1 here
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 non-clinical samples were included;
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 non-clinical 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, titles and abstracts of all identified studies were
examined by two independent judges that both held a bachelor’s degree in psychology. If
differing assessments occurred, those cases were discussed until consensus was reached. All
remaining studies were then assessed in full text and coded by the first author.
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,
non-clinical), diagnosis (if applicable), perfectionism measure, and effect sizes. Because
PERFECTIONISM AND PSYCHOPATHOLOGY 11
some studies did not report the exact mean age of participants, it was estimated using valid
indicators. For example, if data was reported on a sample of undergraduate students, 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 & Otto, 2006) who conducted a review of different conceptualisations 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 (Egan et al., 2011). Following Stoeber and Otto’s (2006) recommendation,
FMPS-Organization and HMPS-Other-Oriented Perfectionism have been omitted because of
unclear findings as to whether these scales represent perfectionistic concerns or
perfectionistic strivings. 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).
PERFECTIONISM AND PSYCHOPATHOLOGY 12
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
comparisons between a clinical and a comparison group regarding perfectionism using
parametric tests of difference such as t-tests or ANOVA models. Due to the aim of examining
the relationship between perfectionism and an outcome and high number of studies reporting
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 perfectionism 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. Due to 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
PERFECTIONISM AND PSYCHOPATHOLOGY 13
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. 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.
PERFECTIONISM AND PSYCHOPATHOLOGY 14
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 ten,
as recommended (Dalton & Dalton, 2008). Only moderator variables evaluated in at least ten
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 in order 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 non-significant 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 V2.0 (CMA; Borenstein, Hedges, Higgins, & Rothstein, 2005); the program was
also used to standardise and convert amongst effect sizes. IBM SPSS Statistics V21.0 (IBM
Corp., 2012) and MPlus version 7.31 analysis package (Muthén and Muthén, 2012) 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 these studies used a longitudinal design. The majority of participants were
female (74.0%) and mean sample age across studies was 25.06 years (SD = 8.13). The
PERFECTIONISM AND PSYCHOPATHOLOGY 15
majority of primary studies (65.1%) reported data from non-clinical 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 evaluated 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 non-
clinical 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.
[Insert Figure 2 here]
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, symptoms of disorders, outcomes related to psychopathology). All effect sizes
from all studies were 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
2 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
PERFECTIONISM AND PSYCHOPATHOLOGY 16
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, 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
moderation 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.
[Insert Figure 3 here]
Perfectionism dimensions and psychopathology. We used weighted random-effects
analyses 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 dimensions and psychopathology outcomes, another research objective was to
assess the relative importance of perfectionistic concerns and perfectionistic striving.
Previous studies have examined 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 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-
PERFECTIONISM AND PSYCHOPATHOLOGY 17
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.
[Insert Table 2 and Figure 4 here]
The average weighted zero-order (r) and unique (β) effect sizes between
perfectionism dimensions and psychopathological outcomes are presented in Table 2.
Regarding clinical disorders, there was evidence that both perfectionistic concerns and
perfectionistic strivings were significantly 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 was much
smaller than the averaged zero-order correlation indicating that these effects were relatively
trivial in comparison to the effects of concerns. Confidence intervals 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
significant 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 strivings. 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.
PERFECTIONISM AND PSYCHOPATHOLOGY 18
For these outcomes, the CIs revealed significant differences in the size of the effects for
perfectionistic concerns relative to perfectionistic 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 internalisation. The effect for perfectionistic concerns and
perfectionistic strivings were no different in the magnitude of the effects indicating a
relatively equal contribution to explaining variance in eating disorder outcomes.
Regarding outcomes related to psychopathology, the perfectionism dimensions were
significantly related to suicidal ideation and general psychological distress. The unique
effects from the path analysis and their confidence intervals indicated significantly larger
effects for perfectionistic 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, perfectionistic concerns and perfectionistic strivings were
both 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 where concerns and strivings
contributed approximately equally to explaining variance.
Perfectionism measures and psychopathology. The second aim also concerned the
evaluation 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 show 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
PERFECTIONISM AND PSYCHOPATHOLOGY 19
perfectionism 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, however the
differences between magnitudes of correlation coefficients were smaller. Regarding the
remaining scales of the two measures, worthy of note was a significant positive overall
correlation for FMPS-organisation 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 as 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 can 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 relationship 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-organisation,
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.
PERFECTIONISM AND PSYCHOPATHOLOGY 20
[Insert Table 3 here]
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, meta-regression
analyses were implemented. This procedure was used to explain variance in the average
weighted effect size of perfectionism and psychopathology with perfectionism dimensions
and sample types as moderators. Specifically, the average weighted correlation between all
perfectionism dimensions 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 determined 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 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 criticism,
FMPS-organisation, 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. Perfectionistic strivings as well as HMPS-Other-
Oriented Perfectionism accounted for a significant amount of variance in effect sizes, both
PERFECTIONISM AND PSYCHOPATHOLOGY 21
associated with a significantly smaller effect compared to perfectionistic concerns. No other
perfectionism dimension or study characteristic yielded a significant effect.
[Insert Table 4 here]
The second set of meta-regression models was implemented to evaluate possible
moderators as well as to investigate the subscales of FMPS and HMPS in specific sample
types (clinical vs. non-clinical). Two meta-regression models were calculated, one restricted
to clinical samples, the other restricted to non-clinical 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. non-clinical).
The reference category for the dummy-coded perfectionism dimension moderator variable
was overall perfectionistic 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. The analysis restricted to non-clinical samples yielded significant
negative regression coefficients for perfectionistic strivings and HMPS-other-oriented
perfectionism, indicating a weaker link of psychopathology to these two domains than to
perfectionistic concerns in non-clinical samples. It is important to note that fewer moderators
could be evaluated in the clinical group compared to non-clinical group because there were
fewer studies on clinical samples. 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.
[Insert Table 5 here]
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
PERFECTIONISM AND PSYCHOPATHOLOGY 22
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 ten tests so was deemed inappropriate to conduct an analysis due to the small sample
size. Anorexia nervosa and bulimia nervosa were subsumed into a single “eating disorders”
category due to the low numbers 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. Perfectionistic strivings, FMPS-
concern over mistakes, and FMPS-organisation accounted for a significant amount of
variance in anxiety disorders, with perfectionistic strivings and FMPS-organisation 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-
organisation, 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
[Insert Table 6 here]
To determine whether the chosen reference category had an impact on the findings,
sensitivity analyses were conducted by (1) varying the reference categories in each model and
(2) not controlling for the two broad perfectionism dimensions perfectionistic concerns and
perfectionistic strivings. These analyses had no major impact on the findings, i.e., the
direction of relationship and the statistical significance remained unchanged, so the
interpretations of the findings are based on the results presented above.
1
In addition to the meta-regression analyses presented here, we also conducted nested meta-regression models
predicting effect sizes restricted to studies reporting on perfectionistic striving vs. perfectionistic concerns.
However, these analyses did not yield any meaningful results and are thus not presented here.
PERFECTIONISM AND PSYCHOPATHOLOGY 23
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 vs. perfectionistic strivings in relation to psychopathology.
The main finding was that both dimensions of perfectionism were associated with
psychopathology outcomes across studies. In the majority of outcomes when the results of the
path analysis are considered, perfectionistic strivings was less related to psychopathology
than perfectionistic concerns, particularly in non-clinical populations. This finding supports
the view of previous authors (e.g., Stoeber & Otto, 2006). Amongst the investigated clinical
disorders only eating disorders were an exception as both dimensions were strongly related to
pathology 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 indicates 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 perfectionism should be targeted for outcomes relating to eating disorders, the
focus for reducing symptoms of OCD, anxiety disorders and depression should be on
perfectionistic concerns as this dimensions seems 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
PERFECTIONISM AND PSYCHOPATHOLOGY 24
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
psychopathology, several findings were observed. There were higher scores on scales which
load on perfectionistic 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 due to the fact that scores on some subscales were not reported in
a sufficient number of scales so that they 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
compared 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 perfectionism in the context of depression may concentrate on HMPS-self-
oriented perfectionism instead of FMPS-personal standards. This conclusion is drawn with
caution however as 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 perfectionism in eating
disorders instead of single subscales like FMPS-concern over mistakes. Further, amongst
PERFECTIONISM AND PSYCHOPATHOLOGY 25
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
actions in OCD (Antony et al., 1998; Frost & Steketee, 1997). This is not surprising given
that some items which comprise doubts about actions were taken from a measure of OCD
symptoms, 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 it was evaluated in. 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 to determine whether it can be subsumed with the
remaining dimensions or not.
In summary, the two main dimensions of perfectionism (perfectionistic concerns and
perfectionistic 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.
PERFECTIONISM AND PSYCHOPATHOLOGY 26
Implications
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.
In terms of theoretical implications, the findings suggest that perfectionism needs to
be considered in the context of a variety of disorders. As outlined, a transdiagnostic process
is a one which is involved in the maintenance of multiple psychological disorders (Harvey et
al., 2004). Egan et al. (2011) argued thatperfectionism is a transdiagnostic process because it
is (i) elevated across eating disorders, anxiety disorders, OCD and depression compared to
healthy controls; (ii) a risk and maintaining factor across disorders, and (iii) 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 positively
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
perfectionism will be more beneficial in patients with co-occurring disorders than disorder
specific treatments and may result in symptom reduction across multiple disorders. The
current 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 disorders,
depression and eating disorders. Future research should examine the reason for this, for
example by considering factors including whether it is the relationship of perfectionism to
comorbidity or shared symptoms between disorders, and how perfectionism is an etiological
factor across disorders. There are many potential etiological factors for perfectionism which
may be relevant and some have been described elsewhere such as parental and cognitive
PERFECTIONISM AND PSYCHOPATHOLOGY 27
factors (e.g., Maloney, Egan, Kane, & Rees, 2014), however others such as an overall deficit
in ones sense of self or core low self-esteem may also be useful to investigate further in order
to inform the theoretical understanding of how perfectionism is a transdiagnostic 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
psychopathology, 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 than through concentrating solely on the
broad perfectionistic concerns and perfectionistic strivings dimensions. Current treatment
approaches in CBT for perfectionism (Egan, Wade, et al., 2014) emphasise 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 trivial 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
PERFECTIONISM AND PSYCHOPATHOLOGY 28
current treatment, specifically where it is investigated if modifying treatment in order 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, where perfectionism has been a recent focus of interest, in order 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 techniques 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, 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 categorise 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 due to too few studies in
PERFECTIONISM AND PSYCHOPATHOLOGY 29
the current sample conducting analysis of the respective subtypes separately. In addition, the
clinical meaning of the present findings is weakened by the fact that more studies on non-
clinical 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 vs. non-clinical
samples: The tendency of a lower influence of perfectionistic strivings in clinical samples
was not significant, possibly due to a lower number of studies in clinical samples.
Furthermore, due to the same reason some subscales could not be evaluated in clinical
samples at all.
Further, although we examined gender as a moderator, we were unable to draw valid
conclusions as the majority of the samples (74%) were female. In particular, we were unable
to examine perfectionism in young men due to the low percentage of male samples in the
studies collected. Future research investigating relations between perfectionism and outcomes
in men is advocated in order 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 as 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 due to some measures potentially having overlap
with the specific symptoms of the disorder. This is particularly the case when considering the
strong association between doubts about actions and OCD, as outlined previously this
perfectionism subscale has been criticised as being highly overlapping with OCD symptoms
(Shafran & Mansell, 2001). While we do not believe this is the case for all measures of
perfectionism, as 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.
PERFECTIONISM AND PSYCHOPATHOLOGY 30
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 perfectionism (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 categories which is a limitation.
Consequently, no conclusion as to whether perfectionism is as a risk or maintenance factor
for psychopathology can be made due to 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,
longitudinal, 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 non-
significant 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 only included published or in press studies in
order 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
PERFECTIONISM AND PSYCHOPATHOLOGY 31
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.
Finally, another limitation is that while 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 which 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). These other scales were not included due to our
focus specifically on perfectionistic strivings and perfectionistic concerns and 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.
Future Research
As the current meta-analysis was conducted on studies that mostly used non-clinical
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 in order to propose suggestions for prevention and treatment of
a variety of disorders. This could be done by further examination of longitudinal associations
PERFECTIONISM AND PSYCHOPATHOLOGY 32
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 in
order 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 due
to 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
perfectionistic 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 perfectionistic 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
interventions for perfectionism. While there has been one systematic review (Lloyd et al.,
2015) which examined CBT for perfectionism and found large pooled effect sizes for
reductions in perfectionism (FMPS-personal standards, concern over mistakes, HMPS-self-
oriented perfectionism), and medium pooled effect size reductions for anxiety and
depression, only eight studies were included, and further RCTs that have now been published
(e.g., Egan, Wade, et al., 2014; Handley et al., 2014) which would be useful to include in an
PERFECTIONISM AND PSYCHOPATHOLOGY 33
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 due to 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 as athletes, or perfectionism in other domains, such as
dyadic perfectionism, 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.
PERFECTIONISM AND PSYCHOPATHOLOGY 34
Acknowledgements
The authors wish to thank Rachael Glassey for providing assistance with the original
literature search. Thanks to Prof. Thomas Ehring for his helpful comments on a previous
version of this paper.
PERFECTIONISM AND PSYCHOPATHOLOGY 35
References
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,USA.
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 perfectionism. . Behavior Therapy, 46(4), 478-492.
Bieling, P. J., Israeli, A. L., & Antony, M. M. (2004). Is perfectionism good, bad, or both?
Examining 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.): Wiley.
Blatt, S. J., D'Aflitti, J. P., & Quinlan, D. M. (1979). Depressive Experiences Questionnaire.
Unpublished manuscript. New Haven, CT.
PERFECTIONISM AND PSYCHOPATHOLOGY 36
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 symptomatology 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.
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
PERFECTIONISM AND PSYCHOPATHOLOGY 37
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., Shafran, R., & Antony, M. M. (2014). Cognitive-behavioral treatement
for perfectionism. New York: Guilford Publications.
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
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.
PERFECTIONISM AND PSYCHOPATHOLOGY 38
Flett, G., Hewitt, P., Boucher, D., Davidson, L., & Munro, Y. (1997). The child-adolescent
perfectionism scale: Development, validation, and association with adjustment.
Unpublished manuscript.
Frost, R. O., Heimberg, R. G., Holt, C. S., & Mattia, J. I. (1993). A comparison of two
measures of perfectionism. 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,
PERFECTIONISM AND PSYCHOPATHOLOGY 39
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).
Harvey, A., Watkins, E., Mansell, W., & Shafran, R. (2004). Cognitive behavioural processes
across psychological 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 disorder, dysthymia, and their relationship to perfectionism. Depression
and Anxiety, 25(3), 207-217. doi: 10.1002/da.20290
PERFECTIONISM AND PSYCHOPATHOLOGY 40
IBM Corp. (2012). IBM SPSS Statistics for Windows, Version 21.0. Armonk, NY: IBM
Corp.
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 Behaviour 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,
perfectionism, and disordered eating behaviors of sons and daughters. Health
Communication, 19(2), 153-163. doi: 10.1207/s15327027hc1902_7
PERFECTIONISM AND PSYCHOPATHOLOGY 41
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.
Rhéaume, J., Freeston, M. H., Dugas, M. J., Letarte, H., & Ladouceur, R. (1995).
Perfectionism, responsibility 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.
Rosenthal, R. (1979). The file drawer problem and tolerance for null results. Psychological
Bulletin, 86(3), 638.
PERFECTIONISM AND PSYCHOPATHOLOGY 42
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 maintenance 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.
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.
PERFECTIONISM AND PSYCHOPATHOLOGY 43
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.
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
PERFECTIONISM AND PSYCHOPATHOLOGY 44
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 Perfectionism Scale
(FMPS; Frost et al., 1990)
Concern over Mistakes (CM): tendency to show negative
reactions to mistakes and to interpret mistakes as a
failure
Personal Standards (PS): striving for high standards
Doubts about Actions (DA): concern that tasks have not
been completed properly
Organisation (O): need for order and neatness
Parental Expectations (PE): high expectations that
respondent’s parents placed on his/her performance
Parental Criticism (PC): parents being overly critical
Multidimensional Perfectionism Scale
(HMPS; Hewitt & Flett, 1991b)
Socially Prescribed Perfectionism (SPP): tendency to
expect others to have extremely high standards for
him/her and to constantly evaluate him/her for what
he/she achieves
Self-Oriented Perfectionism (SOP): tendency to set high
standards for oneself while also reflecting the intrinsic
motivation to reach those standards
Other-Oriented Perfectionism (OOP): having
unrealistically high standards for significant others
Almost Perfect Scale – Revised (APS-R;
Slaney et al., 2001)
Discrepancy
High Standards
Perfectionism Questionnaire (PQ;
Rhéaume et al., 2000)
Negative Consequences of Perfectionism
Perfectionistic Tendencies
Children and Adolescent Perfectionism
Scale (CAPS; Flett, Hewitt, Boucher,
Davidson, & Munro, 1997)
Socially Prescribed Perfectionism
Adaptive/Maladaptive Perfectionism
Scale (AMPS; Rice, Kubal, & Preusser,
2004)
Sensitivity to Mistakes
Compulsiveness
Need for Admiration
Dysfunctional Attitudes Scale (DAS;
Weissman & Beck, 1978)
Self-Criticism/Self-Critical Perfectionism
PERFECTIONISM AND PSYCHOPATHOLOGY 45
Clinical Perfectionism Questionnaire
(CPQ; Fairburn, Cooper, & Shafran,
2003b)
Perfectionism
Positive and Negative Perfectionism
Scale (PANPS; Terry-Short, Owens,
Slade, & Dewey, 1995)
Negative Perfectionism
Obsessive Beliefs Questionnaire
(Obsessive Compulsive Cognitions
Working Group [OCCWG], 2001)
Perfectionism
Perfectionism subscale of the Eating
Disorder Inventory (EDI-P;Garner,
Olmstead, & Polivy, 1983)
Socially Prescribed Perfectionism
Self-Oriented Perfectionism
PERFECTIONISM AND PSYCHOPATHOLOGY 46
Table 2
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
Outcome
k
n
Perfectionistic concerns
k
n
Perfectionistic strivings
r
95% CI of r
β
95% CI of β
r
95% CI of r
β
95% CI of β
LL
UL
LL
UL
LL
UL
LL
UL
Psychological
disorders
Depression
9
12
.40***
.29
.50
.40***
.32
.48
7
8
.18*
.04
.32
.01
-.08
.09
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
disorders
Depressive
symptoms
151
256
.39***
.37
.41
.42***
.41
.43
118
162
.11***
.09
.14
-.08***
-.09
-.07
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
PERFECTIONISM AND PSYCHOPATHOLOGY 47
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
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 to
psychopathology
Suicidal ideation
19
22
.31***
.26
.36
.34***
.30
.38
14
15
.09*
.02
.15
-.06**
-.10
-.02
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.
PERFECTIONISM AND PSYCHOPATHOLOGY 48
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
k
n
r
95% CI
k
n
r
95% CI
LL
UL
LL
UL
FMPS-Concern over Mistakes
3
3
0.45**
0.19
0.66
10
14
0.34***
0.25
0.42
FMPS-Doubts about Actions
3
3
0.34**
0.10
0.54
8
13
0.25***
0.11
0.37
HMPS-Socially Prescribed
Perfectionism
4
4
0.50***
0.39
0.61
6
9
0.53***
0.29
0.71
FMPS-Personal Standards
3
3
0.10
-0.02
0.21
10
15
0.05
-0.02
0.11
HMPS-Self-Oriented
Perfectionism
4
4
0.26**
0.09
0.41
6
9
0.08
-0.05
0.21
FMPS-Organisation
2
2
-0.08
-0.21
0.05
8
13
0.01
-0.07
0.08
FMPS-Parental Expectations
and Criticism
3
6
0.29
-0.14
0.63
9
25
0.16***
0.11
0.21
OCD
Anorexia Nervosa
k
n
r
95% CI
k
n
r
95% CI
LL
UL
LL
UL
FMPS-Concern over Mistakes
10
11
0.37***
0.21
0.52
3
3
0.92*
0.02
1.00
FMPS-Doubts about Actions
9
10
0.54***
0.30
0.71
3
3
0.89
-0.10
0.99
HMPS-Socially Prescribed
Perfectionism
1
1
0.26*
0.07
0.43
1
1
0.78***
0.66
0.86
FMPS-Personal Standards
10
11
0.10*
0.01
0.19
3
3
0.44**
0.15
0.66
HMPS-Self-Oriented
Perfectionism
1
1
0.16
-0.04
0.34
1
1
0.83***
0.74
0.89
FMPS-Organisation
8
9
0.08*
0.02
0.14
3
3
0.41**
0.13
0.64
FMPS-Parental Expectations
and Criticism
10
21
0.16*
0.06
0.25
3
5
0.36
-0.35
0.81
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.
PERFECTIONISM AND PSYCHOPATHOLOGY 49
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)
Variable
Model
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-Organisation
-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, that 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 AND PSYCHOPATHOLOGY 50
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 vs. Non-Clinical Samples
Variable
Model 1
B
SE
β
Clinical
samples
(k = 42,
n = 233)
Perfectionistic strivings
-0.72
0.40
-0.19
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-Organisation
-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
Non-
clinical
samples
(k = 216,
n = 1,332)
Perfectionistic strivings
-0.21**
0.06
-0.16
FMPS-Concern over Mistakes
-0.02
0.14
-0.01
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-Organisation
-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. 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 AND PSYCHOPATHOLOGY 51
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
Variable
Model
B
SE
β
Anxiety disorders
(k = 20, n = 127)
Perfectionistic strivings
-0.60*
0.24
-0.26
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-Organisation
-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, n = 51)
Perfectionistic strivings
1.49
0.55
0.41
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-Organisation
-4.77*
1.57
-1.02
Age
-0.10*
0.03
-2.46
Gender (% female)
-0.02
0.01
-1.26
Eating disorders
(k = 12, n = 46)
Perfectionistic strivings
-1.11
3.40
-0.17
FMPS-Concern over Mistakes
0.26
2.20
0.04
FMPS-Organisation
-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 regression
coefficient, Outcome time = months between baseline assessment of perfectionism and assessment of
outcome.
***p < 0.001.
PERFECTIONISM AND PSYCHOPATHOLOGY 52
Figure 1. Categories and subcategories of outcomes that were analysed in the meta-analysis.
PERFECTIONISM AND PSYCHOPATHOLOGY 53
Figure 2. Process of study selection.
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)
Identified in systematic search (n = 2,497)
PERFECTIONISM AND PSYCHOPATHOLOGY 54
Figure 3. Funnel Plot of the Standard Error by the Effect Size Fisher's Z.
PERFECTIONISM AND PSYCHOPATHOLOGY 55
Figure 4. Diagram depicting generalized meta-analytic path model for effects of
perfectionism dimensions on psychopathological outcomes.
Perfectionistic
Strivings
Perfectionistic
Concerns
Psychopathology
Outcome
