ArticlePDF Available

The Costs of Repression: A Meta-Analysis on the Relation Between Repressive Coping and Somatic Diseases


Abstract and Figures

Objective: When Freud introduced the term repression, he stated its pathogenic potential. Since then, this notion was adapted and continued to date. Surprisingly, there is no attempt to synthesize research on the effect of repressive coping on somatic diseases quantitatively. The current study closes this gap and examines the association between repressive coping as operationalized by Weinberger, Schwartz, and Davidson (1979) and the incidence of cancer, cardiovascular diseases, asthma, and diabetes by applying meta-analytic techniques. Method: An extensive literature search identified 22 studies that met all inclusion criteria and assessed a total of 6,775 participants. The study set was analyzed both as a whole sample (k = 22 studies) and disease-specific (k = 2-10 studies; no study found for diabetes). Results: The results of the meta-analysis indicate a higher risk for repressive copers to suffer from one of the investigated diseases, especially cancer (logOdds after diagnosis: 0.41; 95% confidence interval, 0.09-0.73, p = .012) and hypertension (logOdds: 0.59; 95% confidence interval, 0.32-0.86, p < .0001). After application of a continuity correction due to a missing control group the results for coronary heart disease imply an increased risk for nonrepressors to be affected. Results for cardiovascular diseases in general, heart attack, and asthma did not reach significance. Conclusions: The current meta-analysis revealed significant associations between repressive coping, cancer, and cardiovascular diseases, especially hypertension. These results add to the notion of repressive coping as a consequence of cancer as well as to its important role for the issue of hypertension.
Content may be subject to copyright.
The Costs of Repression: A Meta-Analysis on the Relation between Repressive Coping and
Somatic Diseases
Marcus Mund & Kristin Mitte
Friedrich-Schiller-Universität Jena
Health Psychology, 2012, Volume 31, No. 5, 640–649
American Psychological Association
This article may not exactly replicate the final version published in the APA journal.
It is not the copy of record.
Author Note
This article was published Online First November 14, 2011. DOI: 10.1037/a0026257
We are indebted to André Kretzschmar and Dörte Mund for their time and effort at coding,
as well as for their helpful comments on earlier versions of this paper. We also thank Leila Al-Kuwaiti,
Christiane Gentzel and Sarah Neubauer for their valuable comments on the revision of this manuscript.
Kristin Mitte is now at the Department of Psychology, University of Kassel, Kassel, Germany.
Correspondence concerning this article should be addressed to Marcus Mund, Friedrich-Schiller-Universität,
Institut für Psychologie, Humboldstraße 11, 07743 Jena, Germany. E-mail:
Objective: When Freud introduced the term repression, he stated its pathogenic potential.
Since then, this notion was adapted and continued to date. Surprisingly, there is no attempt
to synthesize research on the effect of repressive coping on somatic diseases quantitatively.
The current study closes this gap and examines the association between repressive coping
as operationalized by Weinberger et al. (1979) and the incidence of cancer, cardiovascular
diseases, asthma, and diabetes by applying meta-analytic techniques. Methods: An extensive
literature search identified 22 studies that met all inclusion criteria and assessed a total of
6 775 participants. The study set was analyzed both as a whole sample (k=22 studies) and
disease-specific (k=210 studies; no study found for diabetes). Results: The results of the
meta-analysis indicate a higher risk for repressive copers to suffer from one of the investigated
diseases, especially cancer (logOdd after diagnosis: 0.41; 95% confidence interval, 0.09–0.73,
p=.012) and hypertension (logOdd: 0.59; 95% confidence interval, 0.32–0.86, p< .0001).
After application of a continuity correction due to a missing control group the results for
coronary heart disease imply an increased risk for non-repressors to be affected. Results
for cardiovascular diseases in general, heart attack and asthma did not reach significance.
Conclusions: The current meta-analysis revealed significant associations between repressive
coping, cancer, and cardiovascular diseases, especially hypertension. These results add to the
notion of repressive coping as a consequence of cancer as well as to its important role for the
issue of hypertension.
Keywords: repressive coping; meta-analysis; cancer; hypertension; cardiovascular
The Costs of Repression: A Meta-Analysis on the Relation
between Repressive Coping and Somatic Diseases
As early as 1894, Freud stated that inhibited negative feelings can lead to severe diseases
like phobias, neuroses, or conversion hysteria (Freud, 1925/1999). While the former two affect
the psyche of a person, conversion hysteria is conceptualized as the transformation of negative
affect into bodily symptoms (Alexander, 1943). The mechanism behind this pathological
process was introduced as ‘repression’ and conceptualized as a way to keep unwanted feelings
unconscious (Freud, 1915/1999). Over the years, Freud’s and Alexander’s original concepts
were rejected (Erdelyi, 2006; Holmes, 1990) and therefore underwent numerous revisions and
reformulations. This led to a considerable amount of constructs labeled ‘repression’ meaning
something different, and of constructs labeled differently but meaning essentially the same
(Garssen, 2007).
Nevertheless, the concept of repression remained popular and scientists became more
interested in it as a dimension of personality. One of the first attempts to identify people high
in this dimension was the Repression-Sensitization scale (R-S scale; Byrne, 1961), which was
a derivative of the MMPI. Repressors classified in this way are said to have a higher threshold
for the perception of negative stimuli whereas sensitizers would approach to those stimuli
cognitively. However, there are some serious problems with the R-S scale concerning its validity
(Slough, Kleinknecht, & Thorndike, 1984) and rationale (Weinberger, 1990). Weinberger,
Schwartz, and Davidson (1979) tried to address these shortcomings by using a measure of
defensiveness (i. e. the tendency trying to be rational, self-protecting and to dismiss feelings)
and a standard trait anxiety measure simultaneously, which enabled them to distinguish between
people who are really not anxious and those who only pretend not to be anxious. Additionally,
they used several physiological measures to test whether those who scored low on anxiety, but
high on defensiveness differ from the other three groups derived by splitting both instruments.
The first group is referred to as repressors or repressive copers1, which are defined as people
who try to avoid the conscious perception of negative feelings (Asendorpf & Scherer, 1983;
Weinberger, 1990). The three other groups—throughout the remainder of this paper referred to
as non-repressors—are named true low anxious, high anxious and defensive high anxious.
Despite stating not to be stressed and to be even less anxious, the repressive copers’ heart
rate and skin conductance increased even above the respective levels of the other three groups
when confronted with a stressful task. This finding was reproduced by other authors that used
different tasks and different physiological measures (Asendorpf & Scherer, 1983).
The dissociation between reported and objectively existing physiological and
psychological arousal might be the effect of the repressors’ tendency to deceive others (Furnham,
Petrides, Sisterson, & Baluch, 2003). However, it seems more plausible to assume a tendency
for self-deception, due to converging evidence concerning repressors’ self-image, coping
styles and memory (for review see Myers, 2010). Additionally, when distinguishing self- from
other-deceivers, most of the repressors are found to belong to the former group (e.g. Derakshan
& Eysenck, 1999).
Despite its lasting popularity and its function as a starting for very fruitful research
(Myers, 2010), the Weinberger et al. approach was criticized for its descriptive character that
does not explain the basal mechanisms of repressive coping. Current theories (Derakshan,
Eysenck, & Myers, 2007; Hock & Krohne, 2004; Mendolia & Baker, 2008) addressed these
shortcomings and shed light on the underlying processes.
Of course, research on repression as a defense mechanism also advanced. The notion
that defense mechanisms work in hierarchically organized clusters is now well established
(Cramer, 1998; Soldz & Vaillant, 1998; Vaillant, Bond, & Vaillant, 1986), with repression located
at an intermediate level of maturity as part of the cluster of neurotic defenses (Vaillant et al.,
1986). Along with the theoretical improvements, new instruments for the assessment of defense
1The terms repressor and repressive coper are used synonymously throughout this paper, albeit there is some
debate whether the repression of negative feelings is a conscious or an unconscious process (Cramer, 1998; Erdelyi,
2006). We use both terms to refer to people applying repressive coping.
mechanisms as conceptualized from a psychodynamic point of view have been introduced, with
the Life Style Index (Conte & Apter, 1995) and the Defense Style Questionnaire (Bond, Gardner,
Christian, & Sigal, 1983) as the most popular. Recently, a Defensive Functioning Scale was also
included in the DSM-IV, which locates repression at the level of mental inhibition (Cramer,
2000). However, beside the fact that repressive coping and the defense of repression both serve
as adaptational processes in response to adversities, their further particulars are very different
(Cramer, 1998).
Repressive Coping and Health
One of the most consistent findings of the research on repressive coping as conceptualized
by Weinberger et al. is the already mentioned discrepancy between reported and objectively
existing response to emotional stimuli or stressful tasks, as long as they are sufficiently
self-relevant (Houtveen, Rietveld, Schoutrop, Spiering, & Brosschot, 2001; Weinberger et
al., 1979). Confronted with such stimuli, repressors typically react with increased heart rate
and decreased skin conductance resistance. People who are told to repress their emotions in
experimental studies show exactly the same features and differ from those who are allowed
to express their feelings as they like (Nyklíˇ
cek, Vingerhoets, & Denollet, 2002). If repressive
copers repressed unwanted feelings permanently (Weinberger, 1990), and if repression of feelings
was associated with the mentioned physiological features, it is close at hand to infer that a high
proportion of repressors should be affected by pathologically high blood pressure or associated
diseases like coronary heart disease (CHD). Indeed, there are numerous studies linking both
(Gleiberman, 2007; Myers et al., 2008) and showing serologically an increased risk for severe
cardiovascular diseases (CVD; Barger, Marsland, Bachen, & Manuck, 2000; Niaura, Herbert,
McMahon, & Sommerville, 1992). It was also shown that their heart rate variability, which is
the time between two subsequent heartbeats and depends on sympathetic and parasympathetic
activity (Acharya, Joseph, Kannathal, Lim, & Suri, 2006), differs from those of non-repressors in
a pathological direction (Fuller, 1992) and that this is predictive for cardiac problems (Acharya et
al., 2006).
Additionally, repressive coping is assumed to be associated with the development of
cancer (Myers et al., 2008). It was always suspected to be part of the ‘cancer personality’, a
configuration of personality traits that differentiates premorbidly between people who will suffer
from cancer once in their life and those who will not. The empirical evidence for this construct is
mixed, with both supporting (Dattore, Shontz, & Coyne, 1980; Kune, Kune, Watson, & Bahnson,
1991; McKenna, Zevon, Corn, & Rounds, 1999) and rejecting (Bleiker & van der Ploeg, 1999;
C. G. Watson & Schuld, 1977) findings. However, most studies are based on the use of the
MMPI or its derivatives like the R-S scale (Dattore et al., 1980; C. G. Watson & Schuld, 1977),
very broad conceptualizations of the repressive coping style (Bleiker & van der Ploeg, 1999;
McKenna et al., 1999) or self-made instruments with unknown psychometric quality (Kune et al.,
1991). Furthermore, most of the self-report measures were administered only once following the
diagnosis. Hence, it is possible that the ‘cancer personality’ is rather a difference in coping with
the diagnosis than a predisposing factor. Nevertheless, concerning the immunological markers
it was shown that repressors are more cancer-prone than non-repressors (Barger et al., 2000;
Jamner, Schwartz, & Leigh, 1988).
The same is true for asthma and diabetes, which both can be linked to several immune
features. The number of eosinophile, for instance, is increased in repressive copers (Jamner et al.,
1988) and in case of an existing asthma can lead to an aggravation of the symptoms. Research on
diabetes shows an increased level of both insulin and glucose in repressors (Barger et al., 2000;
Jamner et al., 1988). Both markers can signal a beginning insulin resistance which is the most
popular feature of type 2 diabetes.
In summary, repressive copers seem to be more predisposed to certain diseases than
non-repressors. But surprisingly, they seem to feel quite comfortable and report only few
symptoms (Boden, Hyland, & Dale, 2005). Until now, it is unclear whether this a) originates
from their high level of defensiveness and the associated tendency to protect themselves from
feeling weak, vulnerable and out of control (Boden et al., 2005), b) it is just an artifact of the
instruments used (Furnham et al., 2003), c) they do not notice their symptoms, or d) they are
just not sick.
Current Study
Given the proposed increased vulnerability, it is all the more surprising that to our best
knowledge there is no meta-analytic attempt to assess whether repressors as conceptualized
by Weinberger et al. (1979) really do suffer more often from several physical diseases than
non-repressors and, if so, to quantify the difference. The current study tries to close this gap,
thereby allowing generalizable evidence.
Inclusion Criteria
Studies had to meet the following criteria to be included in the meta-analysis: a) The
disease under investigation must have been cancer, CVD (i. e. hypertension, CHD or heart
attack), asthma or diabetes. These had been selected because they all have theoretical as well as
physiological associations with repressive coping. b) Because the theoretical framework used is
the repression-approach established by Weinberger et al. (1979), studies must either have utilized
the Weinberger Adjustment Inventory (WAI; Weinberger & Schwartz, 1990) or a combination
of trait anxiety and defensiveness. The WAI was primarily designed to measure an individual’s
socio-emotional adjustment using the two scales distress (i. e. the affective component of
adjustment like depression, low self-esteem, low subjective well-being, proneness to anxiety) and
self-restraint (i. e. ability to suppress egoistic interests in favor of long-term goals and the relation
to others like impulse control, consideration of others, suppression of aggression, responsibility).
For inclusion, at least the proneness to anxiety subscale in combination with the self-restraint
scale must have been administered. As trait anxiety measure, the TMAS (Taylor, 1953), its short
form (Bendig, 1956), the STAI (Spielberger, Gorsuch, & Lushene, 1970) and the R-S scale as
well as their modified forms were accepted. Due to its high correlation with anxiety (e. g. Slough
et al., 1984), the R-S scale was only accepted, if it was combined with a defensiveness scale.
Studies were also included, if they aggregated the anxiety and the defensiveness measure to one
single continuous score.
Concerning defensiveness, the Marlowe-Crowne Social Desirability Scale (MC-SDS;
Crowne & Marlowe, 1960) and the Social-Desirability-Scale-17 (SDS-17; Stöber, 2001) as well
as their modified forms were accepted. The MC-SDS was also used by Weinberger et al. and
in most other studies on repressive coping which adopted their approach. The SDS-17 tries to
update the MC-SDS by including contemporaryly formulated items. The correlation between
both instruments is appropriate (Stöber, 2001).
Recent research attempted to identify repressors using different instruments or different
emotions from the spectrum of negative affectivity (D. Watson & Clark, 1984), which inflamed
a debate about the equivalence of different operationalizations (Mendolia, 1999; Egloff &
Hock, 1999; Furnham & Traynar, 1999). Because the rates of concordance with the original
operationalization were too low in all cases, studies adopting such alternative strategies have
been excluded. Applying such strict criteria ensures to analyze the effects of repressive coping
according to Weinberger et al., and excludes other ‘repression-like’ constructs (Garssen, 2007).
As an additional criterion, c) only studies whose participants were 18 years of age and
older were considered.
Search Strategies
To find as many relevant studies as possible, an extensive literature search was conducted
during May and June 2010. In a first step, the full-text of the databases Psyndex, PubMed,
ScienceDirect, Web of Knowledge, Psycinfo, PsycArticles and the Psychology and Behavioral
Sciences Collection was scrutinized. Additionally, ProQuest and OpenSigle were browsed to
include relevant gray literature. As search terms we applied (repression OR repressive coping
OR defensive coping OR repressor OR represser) AND (cancer OR asthma OR coronary heart
disease OR chd OR blood pressure OR hypertens* OR cardiovascular disease OR diabet*) NOT
(child* OR adolesc* OR juvenil*). In a second step, all possibly relevant studies were searched
thoroughly for more relevant articles that could meet the inclusion criteria as evaluated by title
and context of citation. As third step, a forward search was carried out starting from the articles
of which at least one had to be cited by relevant studies (i. e. Crowne & Marlowe, 1960; Stöber,
2001; Weinberger et al., 1979). Forward search was conducted using PubMed and the Web of
Knowledge. Besides the full-text databases, the International Journal of Behavioral Medicine,
Psychological Inquiry, British Medical Journal, Health Psychology, British Journal of Health
Psychology, Psychology and Health, Psycho-Oncology, Psychosomatic Medicine and Cancer
were separately searched, also manually, if possible. As a last step, authors known to have done
research on the repressive coping style and several diseases were asked for unpublished data.
A flow chart of the conducted literature search is shown in figure 1.
Data were extracted using a coding form implemented in a database due to error
reduction. Two independent coders repeated the procedure with nine randomly selected studies.
κamounted 0.84 for the variables relevant to the analysis, thus the agreement is considered very
Effect Size and Data Analysis
Because we were particularly interested in whether repressors suffer more often from
certain diseases than non-repressors, odds ratios (OR) were used as effect size, hence they are
the appropriate coefficient for fourfold tables (Haddock, Rindskopf, & Shadish, 1998). Because
of better mathematical properties and similarity to more popular coefficients, we log-transformed
every study-specific OR (logOdds; Haddock et al., 1998). Fourfold tables were created so that
logOdds greater than 0 indicate a higher risk for repressors to suffer from the analyzed disease.
When the selected studies did not include a healthy control group2, a continuity correction was
applied that used the pooled data of the studies with control group. In case of studies using
dimensional measures of repressive coping, the effect sizes were transformed into logOdds.
When multiple effects could be computed (e. g., because multiple variations of splitting the
instruments were used, in order to identify repressors and non-repressors), parameters were
averaged to avoid an overrepresentation of the respective study.
2In the sense of absence of the disease under investigation.
For combining individual effect sizes, random-effects models (RE) should have been
calculated. The RE differs from the fixed-effect model (FE) primarily with respect to the
generalizability of results. While the FE allows inferences only about the included sample of
studies, RE allows drawing unconditional inferences about a complete universe of studies.
Furthermore, the FE assumes a common population effect size and explains variance in
individual effect sizes by sampling error (within-study variance) alone. On the contrary, the RE
takes into account the sampling error and additionally an error resulting from randomly selecting
studies out of a universe of studies (between-study variance). The effect in RE is thought of as
the mean of a distribution of effect sizes (Hedges & Vevea, 1998; Viechtbauer, 2010). Since the
results of RE are only reliable with k>5 studies (Hedges & Vevea, 1998), it was not always
possible to use RE in the disease-specific analyses. All RE-analyses were conducted using the
Restricted Maximum-Likelihood estimator, for it is superior to the DerSimonian-Laird estimator
in most cases and addresses some of the latter’s shortcomings concerning moderate number of
studies (Viechtbauer, 2005).
As measure of heterogeneity, the Q-statistic was used, thereby taking into account its low
power by applying a probability level of .10.
Publication bias was addressed by means of trim-and-fill corrected funnel plots (Duval &
Tweedie, 2000) and normal-quantil plots (Wang & Bushman, 1998). To evaluate the robustness
of the results, fail-safe numbers were calculated using the formulas proposed by Rosenberg
All analyses were conducted using the metafor-package for R (R Development Core
Team, 2010; Viechtbauer, 2010).
Study Sample
Table 1 presents the characteristics of all studies that matched the inclusion criteria.
Data of 6 775 participants were integrated in the analysis, 3 883 of them female and
2 843 male; 49 did not indicate gender. Most studies investigated cancer (45%), followed by
hypertension and other CVD. Only two studies dealing with asthma could be identified and
none studying diabetes. Furthermore, it is evident that all studies using a categorical approach
of repressive coping utilized the MC-SDS. This trend towards established instruments is also
evident regarding measures of trait anxiety.
Repressive Coping and Somatic Diseases
The first analysis addressed the question whether the risk of suffering from one of the
investigated diseases is increased for repressors. Across all k=22 studies, an effect size of
0.27 was found (OR =1.31; see table 2 for more details). This value is significant and implies
that the repressors’ risk for suffering from one of the diseases is increased by 31%. This result
was not contradicted by the additionally conducted sensitivity analyses, which used different
estimators or left single studies out. According to the fail-safe-N-statistic, 19 studies would be
necessary to nullify the effect. Applying the criterion proposed by Rosenthal (1979), the mean
effect size can not be considered robust. However, taking into account that only 22 studies met all
inclusion criteria at all, it seems not feasible to assume that another 19 studies are unpublished or
not retrievable.
Because the applied continuity correction might have biased the results, an additional
analysis was conducted that contained only the studies with a healthy control group (Ginzburg,
Solomon, & Bleich, 2002; Jensen, 1987; Kraft, 1998; Kreitler, Chaitchik, & Kreitler, 1993; Perini
et al., 1994; Servaes, Vingerhoets, Vreugdenhil, Keuning, & Broekhuijsen, 1999; Zachariae et al.,
2004) or operationalized repressive coping as a dimensional construct (Nyklíˇ
cek, Vingerhoets,
van Heck, & van Limpt, 1998; Nyklíˇ
cek, Vingerhoets, & van Heck, 1999). This analysis yielded
an effect size of 0.53 (OR =1.70). Again, this result proved to be robust in the sensitivity
analyses and with regard to the fail-safe number (see table 2).
With regard to potential moderators of the effect size, regression analyses were
conducted. Note that it was only possible to calculate regressions with not more than two
predictors due to the small number of studies. Among the studies with a control group, those
conducted in the USA produced greater effect sizes; no other significant moderators were
Repressive Coping and Cancer
10 studies focused on cancer (2 015 participants, 95.83% females).
Across all 10 studies, a significant effect size of 0.41 (OR =1.51) resulted. Thus,
for repressors the risk of a cancer diagnosis is increased by 51%. Again, this result proved
to be robust in the conducted sensitivity analyses. With regard to the fail-safe number, 44
studies would be necessary to nullify the effect, which is not a robust number according to the
Rosenthal-criterion. Again, given the fact that only 10 studies met all inclusion criteria at all, it
seems not plausible to assume that another 44 studies are unpublished or not retrievable. Only
two studies (Kreitler et al., 1993; Zachariae et al., 2004) assessed the repressive coping style
before diagnosis. These two studies were analyzed using FE, yielding an overall effect size of
0.20 (OR =1.22), which was not significant (p=.4; see table 2 for more details).
Repressive Coping and CVD
10 studies with 4 664 (40.46% females) participants in total were included that focused
either on hypertension, CHD, or heart attack. The mean effect size of 0.21 (OR =1.23; see table
2) was not significant. This result was confirmed through all sensitivity analyses.
Considering every single disease (see table 2), only the FE-analysis for hypertension
yielded a significant and robust result. The analysis suggests an increased risk of suffering from
higher blood pressure in the repressor group. Nevertheless, due to the small study sample and
the significant heterogeneity of the studies this result can not be generalized (Hedges & Vevea,
1998; Viechtbauer, 2010).
Although the result for the FE-analysis concerning CHD was also significant, thereby
implying an increased risk for non-repressors, it should be interpreted very cautiously for two
reasons. First, all four studies had to be corrected due to a missing control group. Second, the
effect depends on one single study (Denollet, Martens, Nyklíˇ
cek, Conraads, & de Gelder, 2008);
3Other moderators examined were the measure of anxiety, year of publication, sample size, sample age, sample
sex (only female, only male, mixed), disease, number of groups assessed, way of categorization (splitting at median,
mean, a specific reference value or a certain quantile) and categorical vs. dimensional operationalization.
leaving it out led to non-significance.
The analysis for repressive coping and heart attack yielded no significant result. Note that
the continuity correction had to be applied for both studies, too.
Repressive Coping and Asthma
As for heart attack, the two asthma studies had to be corrected due to a missing control
group. Again, the result was not significant (logOdds =0.15,OR =0.86; see table 2).
Although current reviews conclude that there is much evidence for the association
between repressive coping and somatic diseases (Myers, 2010; Myers et al., 2008), there was
not yet a summarizing meta-analysis to quantify this effect. The current paper tried to close this
gap. The main question investigated was whether repressors, according to the conceptualization
established by Weinberger et al. (1979), do have an increased risk for suffering from cancer,
CVD, or asthma.
The first main finding of this study is that the Weinberger et al. approach seems to be
very popular in research, but is rarely adopted for use with patients. Therefore, only 22 studies
could have been identified. In sum, the results imply a significantly increased risk for repressors
to suffer from one of the investigated diseases, especially cancer and elevated blood pressure
or even hypertension. Analyses for the continuity corrected CHD and asthma studies yielded a
significantly increased risk for non-repressors or null effects, respectively.
Stating that repressors have an increased risk of suffering from several diseases provokes
subsequent questions. First, does repressive coping cause the diseases and, if so, what are the
underlying processes? Second, what consequences does a repressive coping style have with
regard to the progression of the disease?
For cancer, the present results imply that repressive coping does not precede the
diagnosis, but is rather a consequence of it. However, only two studies assessed repressive
coping before the diagnosis, hence this interpretation is fragile. Nevertheless, neither Kreitler
et al. (1993) nor Zachariae et al. (2004) were able to predict the later patients by repressive
coping. In contrast, after the diagnosis the risk of being repressive increased. Both research
groups interpreted their results in terms of repressive coping as a consequence of cancer.
Similarly, Drageset and Lindstrøm (2003) could show that among women awaiting the results
of a biopsy those who applied cognitive defenses were less anxious, and they hypothesized
that this mechanism is especially useful in situations without the possibility for active coping.
Indeed, Zachariae et al. (2004) found a decrease of the proportion of repressors 12 weeks after
the diagnosis, which supports the notion of repressive coping as a first self-protective mechanism.
This line of interpretation is also concordant with other current reviews (e. g. Bleiker & van
der Ploeg, 1999). Extending the focus beyond repressive coping leads to similar results with
no single personality trait being found to be carcinogenic (e. g. Hansen, Floderus, Frederiksen,
& Johansen, 2005). Nevertheless, some explication is needed for the point estimate of 0.22
on the logarithmic odds ratio scale resulting from the analysis of cancer and repressive coping
before the diagnosis. If personality was unimportant, a logOdds of 0 would have been expected.
The deviation found (although non-significant) could be ascribed to the high selectivity of the
primary samples. Both studies assessed women undergoing biopsy. Hence, the setting itself—a
hospital, incalculable result—could have fostered repressive tendencies. Furthermore it can
be assumed that the patients belonged to at-risk groups (e. g. genetic predisposition, smokers,
older age) and therefore had a premonition. The repressors in the Zachariae et al. (2004) study
were slightly more convinced of having cancer which again could have activated the repressive
defense. In sum, the slightly increased risk of repressors to be diagnosed with cancer found in the
meta-analysis could be explained by processes that had already started before the primary studies
were even conducted. Future research could benefit from the distinction between state- and
trait-repressive coping (Mendolia, Moore, & Tesser, 1996) to isolate the effects of dispositions
on the development and the course of cancer (and other relevant diseases) from the effects
situations. However, this would necessitate much longer-lasting studies to identify the point in
time when situational effects begin to mix up with dispositional effects.
If repressive coping was a consequence of the diagnosis—what does it mean for the
course of the disease? The preliminary answers are somewhat ambiguous. On the one hand,
Giese-Davis and Spiegel (2002) conclude that repressors do not regularly participate in and
terminate therapy earlier. Jensen (1987) found shorter duration of remission and more recurrence
among the repressive patients. Other studies report higher mortality for repressive or similarly
configured patients (Weihs, Enright, Simmens, & Reiss, 2000). On the other hand, there is
evidence that repressors suffer less under the chemotherapy and ask for support more often
(Ward, Leventhal, & Love, 1988), exhibit less symptoms of acute stress (Fischer Pedersen
& Zachariae, 2010), and experience more time without relapse than non-repressors (Dean &
Surtees, 1989). Finally, other authors could not find any associations between coping styles and
disease course at all (Petticrew, Bell, & Hunter, 2002). Because of these contradicting results it
is nearly impossible to draw any conclusions about repressive coping and disease progression, all
the more because a wide range of conceptualizations and instruments has been used in the cited
studies, which might also have caused the ambiguity of the findings.
Concerning CVD, the meta-analysis showed that repressors’ risk of suffering from at
least elevated blood pressure is increased by 80% compared to non-repressors. Three out of the
four analyzed studies assessed participants without abnormal hypertension, but it is known that
blood pressure increases with age which in turn increases the probability of an early diagnosed
pathological hypertension among the repressors, and consequently the risk for consecutive
illnesses like CHD or heart attack. Furthermore, the time during which anti-hypertensive drugs
have to be taken is prolonged. The processes behind this association are not yet clear. According
to the discontinuity hypothesis (Hock & Krohne, 2004), repressors experience more arousal when
confronted with ambiguous situations which could increase blood pressure via the release of
cortisol (Brown et al., 1996). Another explanation stems from the opioid-hypothesis of repressive
coping (e. g. Jamner et al., 1988). According to this approach, the hypothalamic-pituitary-adrenal
axis (HPA) is supposed to be hyperactive in repressors. Consequently, more opioid is released
which leads to a permanently euphoric state. This would explain why repressors’ in fact higher
trait anxiety does not manifest in questionnaires. Additionally, the HPA-hyperactivity leads to an
increased release of cortisol and in turn to an increased blood pressure. Hence, according to the
opioid-hypothesis, repressive coping and hypertension are two facets of one single process. It
should be noted, however, that although the present results for the other CVD imply an increased
risk for non-repressors to suffer from CHD, they are determined by one single study and were
all continuity corrected. Thus, it is possible that future studies including a healthy control
group could yield even conversed results. However, either outcome is possible and plausible.
First, due to the potentially earlier diagnosis of hypertension in repressors it is conceivable that
preventive actions are taken on time to avoid CHD. This would lead to an increased proportion
of non-repressors among the CHD patients, since they have not had so much time to adapt
their lifestyle to their cardiovascular conditions. Second, an increased risk for repressors could
be a consequence of the prolonged duration of hypertension which leads to an accelerated
development of CHD. However, both assumptions are speculative yet and should be subject to
further research.
With regard to asthma, the main finding is an urgent need for more research. The analysis
showed a slightly increased risk for non-repressors to be affected. Note that the continuity
correction had to be applied for both included studies. Nevertheless, with regard to disease
progression, repressors could be expected to have a better prognosis because many tasks of
disease management support their wish for control (e. g. Weinberger, 1990; Weinberger et
al., 1979), for instance the monitoring of lung volume and the avoidance of asthma triggers.
Indeed, González-Freire, Vázquez-Rodríguez, Marcos-Velázquez, and González de la Cuesta
(2010) noticed that the repressors report less severe asthmatic symptoms than the affected
non-repressors, and Cooke, Myers, and Derakshan (2003) observed a better therapy adherence.
However, both studies admonish caution, because lung volume as an objective marker of disease
severity was more decreased in the repressor groups. González-Freire et al. (2010) also note that
repressors could have applied an optimistic bias concerning their symptoms, which is in line with
earlier research (for review see Myers, 2010).
The current paper was strictly guided by the repressive coping concept established by
Weinberger et al. (1979). This line of action seemed appropriate considering the heterogeneity
of related constructs (Garssen, 2007). Nevertheless, future research should try to illuminate
the convergence between several ways of identifying repressors to provide a variety of tools
to bypass repressors’ response styles. This would allow to extend the current analysis as well
as previous work done, which used other constructs, in terms of interpretation. Furthermore, it
could help clearing the contradictions in current research if it could be shown that some of the
operationalizations measure in fact different constructs.
To our surprise, no study focusing repressive coping and diabetes could be found despite
the strong psychophysiological component linking both. Indeed, some authors report about such
studies (Myers et al., 2008), but data, however, were not available.
Moreover, it is still possible that age confounds the results. Erskine, Kvavilashvili,
Conway, and Myers (2007) report a higher prevalence of repressors among elderly. Due to the
lack of group-specific means it was impossible to control for that variable in the current study.
If the results were strongly confounded with age effects, the interpretation would have to be
changed and would then suggest that there is no increased risk for repressors to suffer from
cancer or hypertension, but for older people. Educational status could be another potentially
confounding factor. However, data about coping style and education are reported in two studies
only. While there is evidence for an association in the work of Hunt (2001), there is none in that
of Zachariae et al. (2004). Future research should pay more attention to age and education in
order to partial out their effects and to facilitate conclusions about the main effects of repressive
From a psychodynamic point of view, the study could be criticized for taking only one
single style of ego defense into account. Current theories postulate that defense mechanisms are
employed in clusters and do not work independently (Cramer, 1998; Vaillant et al., 1986). Hence,
Paika et al. (2010) demonstrated that, for cancer, mechanisms from the same cluster can have
very different influences on the same outcome variable. However, this again leads to the necessity
of establishing a more sophisticated nomological network around the construct of repressive
coping to enable the consideration of several processes simultaneously.
Even though the meta-analysis provides some ideas about the associations between
repressive coping and health it is necessary to conduct more longitudinal and prospective studies
to draw reliable conclusions about antecedent and consecutive effects.
The current meta-analysis closes the gap between the often stated notion of an association
between repressive coping and somatic diseases and the lack of integrated empirical data.
It was shown that repressors according to the definition of Weinberger et al. (1979) have an
increased risk for suffering from a somatic disease, with the most increased risk for hypertension.
Concerning cancer, an association was evident only following the diagnosis.
References marked with an asterisk indicate studies included in the meta-analysis.
Acharya, U. R., Joseph, K. P., Kannathal, N., Lim, C. M., & Suri, J. S. (2006). Heart
rate variability: A review. Medical and Biological Engineering and Computing,44,
Alexander, F. (1943). Fundamental concepts of psychosomatic research: Psychogenesis,
conversion, specificity. Psychosomatic Medicine,5, 205–210.
Asendorpf, J. B., & Scherer, K. R. (1983). The discrepant repressor: Differentiation between
low anxiety, high anxiety, and repression of anxiety by autonomic-facial-verbal patterns of
behavior. Journal of Personality and Social Psychology,45, 1334–1346.
Barger, S. D., Marsland, A. L., Bachen, E. A., & Manuck, S. B. (2000). Repressive coping and
blood measures of disease risk: Lipids and endocrine and immunological responses to a
laboratory stressor. Journal of Applied Social Psychology,30, 1619–1638.
Bendig, A. W. (1956). The development of a short form of the Manifest Anxiety Scale. Journal
of Consulting Psychology,20, 384.
Bleiker, E. M. A., & van der Ploeg, H. M. (1999). Psychosocial factors in the etiology of breast
cancer: Review of a popular link. Patient Education and Counseling,37, 201–214.
Boden, J. M., Hyland, M. E., & Dale, K. L. (2005). Defensiveness and symptom reporting.
Current Psychology,24, 153–170.
Bond, M., Gardner, S. T., Christian, J., & Sigal, J. J. (1983). Empirical study of self-rated defense
styles. Archives of General Psychiatry,40, 333–338.
Brown, L. L., Tomarken, A. J., North, D. N., Loosen, P. T., Kalin, N. H., & Davidson, R. J.
(1996). Individual differences in repressive-defensiveness predict basal salivary cortisol
levels. Journal of Personality and Social Psychology,70, 362–371.
Byrne, D. (1961). The repression-sensitization scale: Rationale, reliability, and validity. Journal
of Personality,29, 334–349.
Conte, H. R., & Apter, A. (1995). The Life Style Index: A self-report measure of ego
defenses. In H. R. Conte & R. Plutchik (Eds.), Ego defenses: Theory and measurement
(pp. 179–201). New York: Wiley.
Cooke, L., Myers, L. B., & Derakshan, N. (2003). Lung function, adherence and denial in
asthma patients who exhibit a repressive coping style. Psychology, Health & Medicine,8,
Cramer, P. (1998). Coping and defense mechanisms: What’s the difference? Journal of
Personality,66, 919–946.
Cramer, P. (2000). Defense mechanisms in psychology today: Further processes for adaptation.
American Psychologist,55, 637–646.
Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of
psychopathology. Journal of Consulting Psychology,24, 349–354.
Dattore, P. J., Shontz, F. C., & Coyne, L. (1980). Premorbid personality differentiation of cancer
and noncancer groups: A test of the hypothesis of cancer proneness. Journal of Consulting
and Clinical Psychology,48, 388–394.
Dean, C., & Surtees, P. G. (1989). Do psychological factors predict survival in breast cancer?
Journal of Psychosomatic Research,33, 561–569.
Denollet, J. (1991). Negative affectivity and repressive coping: Pervasive influence on
self-reported mood, health, and coronary-prone behavior. Psychosomatic Medicine,53,
Denollet, J. (1999). Repressive coping predicts mortality in patients with coronary heart disease.
Psychosomatic Medicine,61, 125.
Denollet, J., Martens, E. J., Nyklíˇ
cek, I., Conraads, V. M., & de Gelder, B. (2008). Clinical
events in coronary patients who report low distress: Adverse effect of repressive coping.
Health Psychology,27, 302–308.
Derakshan, N., & Eysenck, M. W. (1999). Are repressors self-deceivers or other-deceivers?
Cognition and Emotion,13, 1–17.
Derakshan, N., Eysenck, M. W., & Myers, L. B. (2007). Emotional information processing in
repressors: The vigilance-avoidance theory. Cognition and Emotion,21, 1585–1614.
Drageset, S., & Lindstrøm, T. C. (2003). The mental health of women with suspected
breast cancer: The relationship between social support, anxiety, coping and defence in
maintaining mental health. Journal of Psychiatric and Mental Health Nursing,10,
Duval, S., & Tweedie, R. (2000). Trim and Fill: A simple Funnel Plot based method of testing
and adjusting for publication bias in meta-analysis. Biometrics,56, 455–463.
Egloff, B., & Hock, M. (1999). When is an agreement between two methods for assessing
coping styles ‘good’? A reply to Mendolia. Personality and Individual Differences,27,
Erdelyi, M. H. (2006). The unified theory of repression. Behavioral and Brain Sciences,29,
Erskine, J. A. K., Kvavilashvili, L., Conway, M. A., & Myers, L. B. (2007). The effects of age on
psychopathology, well-being and repressive coping. Aging & Mental Health,11, 394–404.
Fischer Pedersen, A., & Zachariae, R. (2010). Cancer, acute stress disorder, and repressive
coping. Scandinavian Journal of Psychology,51, 84–91.
Frasure-Smith, N., Lespérance, F., Gravel, G., Masson, A., Juneau, M., & Bourassa, M. G.
(2002). Long-term survival differences among low-anxious, high-anxious and repressive
copers enrolled in the Montreal Heart Attack Readjustment Trial. Psychosomatic Medicine,
64, 571–579.
Frasure-Smith, N., Lespérance, F., Prince, R. H., Verrier, P., Garber, R. A., Juneau, M., . . .
Bourassa, M. G. (1997). Randomised trial of home-based psychosocial nursing
intervention for patients recovering from myocardial infarction. Lancet,350, 473–479.
Freud, S. (1915/1999). Die Verdrängung [Repression.]. In A. Freud (Ed.), Sigm. Freud:
Gesammelte Werke, chronologisch geordnet: Zehnter Band, Werke aus den Jahren
1913-1917 (pp. 247–261). Frankfurt am Main: Fischer.
Freud, S. (1925/1999). Studien über Hysterie [Studies in hysteria.]. In A. Freud (Ed.), Sigm.
Freud: Gesammelte Werke, chronologisch geordnet: Erster Band, Werke aus den Jahren
1892-1899 (pp. 75–312). Frankfurt am Main: Fischer.
Fuller, B. F. (1992). The effects of stress-anxiety and coping styles on heart rate variability.
International Journal of Psychophysiology,12, 91–96.
Furnham, A., Petrides, K. V., Sisterson, G., & Baluch, B. (2003). Repressive coping style and
positive self-presentation. British Journal of Health Psychology,8, 223–249.
Furnham, A., & Traynar, J. (1999). Repression and effective coping styles. European Journal of
Personality,13, 465–492.
Garssen, B. (2007). Repression: Finding our way in the maze of concepts. Journal of Behavioral
Medicine,30, 471–481.
Giese-Davis, J., & Spiegel, D. (2002). Emotion regulation and metastatic breast cancer.
International Congress Series,1241, 31–35.
Ginzburg, K., Solomon, Z., & Bleich, A. (2002). Repressive coping style, acute stress disorder,
and posttraumatic stress disorder after myocardial infarction. Psychosomatic Medicine,64,
Gleiberman, L. (2007). Repressive/defensive coping, blood pressure, and cardiovascular
rehabilitation. Current Hypertension Reports,9, 7–12.
González-Freire, B., Vázquez-Rodríguez, I., Marcos-Velázquez, P., & González de la Cuesta, C.
(2010). Repression and coping styles in asthmatic patients. Journal of Clinical Psychology
in Medical Settings,17, 220–229.
Haddock, C. K., Rindskopf, D., & Shadish, W. R. (1998). Using odds ratios as effect sizes
for meta-analysis of dichotomous data: A primer on methods and issues. Psychological
Methods,3, 339–353.
Hansen, P. E., Floderus, B., Frederiksen, K., & Johansen, C. (2005). Personality traits, health
behavior, and risk for cancer: A prospective study of a Swedish twin cohort. Cancer,103,
Hedges, L. V., & Vevea, J. L. (1998). Fixed- and random-effects models in meta-analysis.
Psychological Methods,3, 486–504.
Hock, M., & Krohne, H. W. (2004). Coping with threat and memory for ambiguous information:
Testing the repressive discontinuity hypothesis. Emotion,4, 65–86.
Holmes, D. S. (1990). The evidence for repression: An examination of sixty years of research.
In J. L. Singer (Ed.), Repression and dissociation: Implications for personality theory,
psychopathology, and health (pp. 85–102). Chicago: University of Chicago Press.
Houtveen, J. H., Rietveld, S., Schoutrop, M., Spiering, M., & Brosschot, J. F. (2001). A
repressive coping style and affective, facial and physiological responses to looking at
emotional pictures. International Journal of Psychophysiology,42, 265–277.
Hunt, M. E. (2001). Exploring the repressive coping styles of Caucasian and African-American
women with newly diagnosed stage II and recurrent breast cancer: The impact of
religiosity and under-reporting of stressful life events (Unpublished doctoral dissertation).
George Washington University, Washington, D. C..
Jamner, L. D., Schwartz, G. E., & Leigh, H. (1988). The relationship between repressive and
defensive coping styles and monocyte, eosinophile, and serum glucose levels: Support for
the opioid peptide hypothesis of repression. Psychosomatic Medicine,50, 567–575.
Jensen, M. R. (1987). Psychobiological factors predicting the course of breast cancer. Journal
of Personality,55, 317–342.
Jensen-Johansen, M. B., Christensen, S., Valdimarsdottir, H., Zakowski, S., Bovbjerg, D. H.,
Jensen, A. B., & Zachariae, R. (in preparation). The effect of an expressive writing
intervention on self-reported health among breast cancer patients: Results from a
nationwide randomized clinical trial.
King, A. C., Taylor, C. B., Albright, C. A., & Haskell, W. L. (1990). The relationship
between repressive and defensive coping styles and blood pressure responses in healthy,
middle-aged men and women. Journal of Psychosomatic Research,34, 461–471.
Kraft, M. E. (1998). The relationship between the repressive coping style and breast cancer
incidence (Unpublished doctoral dissertation). University of California, Irvine.
Kreitler, S., Chaitchik, S., & Kreitler, H. (1993). Repressiveness: Cause or result of cancer?
Psycho-Oncology,2, 43–54.
Kune, G. A., Kune, S., Watson, L. F., & Bahnson, C. B. (1991). Personality as a risk factor in
large bowel cancer: Data from the Melbourne Colorectal Cancer Study. Psychological
Medicine,21, 29–41.
McKenna, M. C., Zevon, M. A., Corn, B., & Rounds, J. (1999). Psychosocial factors and the
development of breast cancer: A meta-analysis. Health Psychology,18, 520–531.
Mendolia, M. (1999). Agreement between two methods for identifying dispositional repressors:
A comment on Egloff and Hock (1997). Personality and Individual Differences,27,
Mendolia, M., & Baker, G. A. (2008). Attentional mechanisms associated with repressive
distancing. Journal of Research in Personality,42, 546–543.
Mendolia, M., Moore, J., & Tesser, A. (1996). Dispositional and situational determinants of
repression. Journal of Personality and Social Psychology,70, 856–867.
Mizrahi, J. B. (2003). The impact of family process and repressive coping style on change in
distress after treatment for stage II or III breast cancer (Unpublished doctoral dissertation).
George Mason University, Fairfax.
Myers, L. B. (2010). The importance of the repressive coping style: Findings from 30 years of
research. Anxiety, Stress, & Coping,23, 3–17.
Myers, L. B., Burns, J. W., Derakshan, N., Elfant, E., Eysenck, M. W., & Phipps, S. (2008).
Current issues in repressive coping and health. In A. J. Vingerhoets, I. Nyklíˇ
cek, &
J. Denollet (Eds.), Emotion regulation: Conceptual and clinical issues (pp. 69–86). New
York: Springer.
Niaura, R., Herbert, P. N., McMahon, N., & Sommerville, L. (1992). Repressive coping and
blood lipids in men and women. Psychosomatic Medicine,54, 698–706.
cek, I., Vingerhoets, A., & Denollet, J. (2002). Emotional (non-)expression and health:
Data, questions and challenges. Psychology and Health,17, 517–528.
cek, I., Vingerhoets, A. J. J. M., & van Heck, G. L. (1999). Elevated blood pressure and
self-reported symptom complaints, daily hassles, and defensiveness. International Journal
of Behavioral Medicine,6, 177–189.
cek, I., Vingerhoets, A. J. J. M., van Heck, G. L., & van Limpt, M. C. A. M. (1998).
Defensive coping in relation to casual blood pressure and self-reported daily hassles and
life events. Journal of Behavioral Medicine,21, 145–161.
Paika, V., Almyroudi, A., Tomenson, B., Creed, F., Kampletsas, E. O., Siafaka, V., . . . Hyphantis,
T. (2010). Personality variables are associated with colorectal cancer patients’ quality
of life independent of psychological distress and disease severity. Psycho-Oncology,19,
Perini, C., Smith, D. H. G., Neutel, J. M., Smith, M. A., Henry, J. P., Bühler, F. R., . . . Weber,
M. A. (1994). A repressive coping style protecting from emotional distress in low-renin
essential hypertensives. Journal of Hypertension,12, 601–607.
Petticrew, M., Bell, R., & Hunter, D. (2002). Influence of psychological coping on survival
and recurrence in people with cancer: Systematic review. British Medical Journal,325,
R Development Core Team. (2010). R: A language and environment for statistical computing.
Vienna: R Foundation for Statistical Computing.
Rosenberg, M. S. (2005). The file-drawer problem revisited: A general weighted method for
calculating fail-safe numbers in meta-analysis. Evolution,59, 464–468.
Rosenthal, R. (1979). The ‘file drawer problem’ and tolerance for null results. Psychological
Bulletin,86, 638–641.
Servaes, P., Vingerhoets, A. J. J. M., Vreugdenhil, G., Keuning, J. J., & Broekhuijsen, A. M.
(1999). Inhibition of emotional expression in breast cancer patients. Behavioral Medicine,
25, 23–27.
Shaw, R. E., Cohen, F., Fishman-Rosen, J., Murphy, M. C., Stertzer, S. H., Clark, D. A., &
Myler, R. K. (1986). Psychologic predictors of psychosocial and medical outcomes in
patients undergoing coronary angioplasty. Psychosomatic Medicine,48, 582–597.
Slough, N., Kleinknecht, R. A., & Thorndike, R. M. (1984). Relationship of the
Repression-Sensitization Scales to anxiety. Journal of Personality Assessment,48,
Soldz, S., & Vaillant, G. E. (1998). A 50-year longitudinal study of defense use among inner city
men: A validation of the DSM-IV defense axis. The Journal of Nervous & Mental Disease,
186, 104–111.
Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E. (1970). Manual for the State-Trait Anxiety
Inventory. Palo Alto: Consulting Psychologists Press.
Stöber, J. (2001). The Social-Desirability-Scale-17 (SDS-17): Convergent validity, discriminant
validity, and relationship with age. European Journal of Psychological Assessment,17,
Taylor, J. A. (1953). A personality scale for manifest anxiety. The Journal of Abnormal and
Social Psychology,48, 285–290.
Vaillant, G. E., Bond, M., & Vaillant, C. O. (1986). An empirically validated hierarchy of defense
mechanisms. Archives of General Psychiatry,43, 786–794.
Viechtbauer, W. (2005). Bias and efficiency of meta-analytic variance estimators in the
random-effects model. Journal of Educational and Behavioral Statistics,30, 261–293.
Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. Journal of
Statistical Software,36, 1–48.
Wang, M. C., & Bushman, B. J. (1998). Using the Normal Quantile Plot to explore meta-analytic
data sets. Psychological Methods,3, 46–54.
Ward, S. E., Leventhal, H., & Love, R. (1988). Repression revisited: Tactics used in coping with
a severe health threat. Personality and Social Psychology Bulletin,14, 735–746.
Watson, C. G., & Schuld, D. (1977). Psychosomatic factors in the etiology of neoplasms. Journal
of Consulting and Clinical Psychology,45, 455–461.
Watson, D., & Clark, L. A. (1984). Negative affectivity: The disposition to experience aversive
emotional states. Psychological Bulletin,96, 465–490.
Weihs, K. L., Enright, T. M., Simmens, S. J., & Reiss, D. (2000). Negative affectivity,
restriction of emotions, and site of metastases predict mortality in recurrent breast cancer.
Journal of Psychosomatic Research,49, 59–68.
Weinberger, D. A. (1990). The construct validity of the repressive coping style. In J. L. Singer
(Ed.), Repression and dissociation: Implications for personality theory, psychopathology,
and health (pp. 337–386). Chicago: University of Chicago Press.
Weinberger, D. A., & Schwartz, G. E. (1990). Distress and restraint as superordinate dimensions
of self-reported adjustment: A typological perspective. Journal of Personality,58,
Weinberger, D. A., Schwartz, G. E., & Davidson, R. J. (1979). Low-anxious, high-anxious, and
repressive coping styles: Psychometric patterns and behavioral and physiological responses
to stress. Journal of Abnormal Psychology,88, 369–380.
Zachariae, R., Jensen, A. B., Pedersen, C., Jørgensen, M. M., Christensen, S., Lassesen, B.,
& Lehbrink, M. (2004). Repressive coping before and after diagnosis of breast cancer.
Psycho-Oncology,13, 547–561.
Table 1
Descriptive data of the selected studies (k = 22 studies).
Neffect sizef
Author(s), year countryasumbfemale male REPcdefensivenessdanxietyelogOdd0logOdd1
Fischer Pedersen &
Zachariae, 2010
DK 112 60 52 31 MC-SDS SF-MAS 0.07
Hunt, 2001 USA 127 127 0 39 MC-SDS SF-MAS 0.25
Jensen, 1987 86 86 0 42 MC-SDS SF-MAS 0.51
Jensen-Johansen et al., in
DK 507 507 0 160 MC-SDS TMAS 0.17
Kraft, 1998 USA 220 220 0 59 MC-SDS SF-MAS 1.42
Kreitler et al., 1993 ISR 98 98 0 26 MC-SDS TMAS 0.13 1.03
Mizrahi, 2003 USA 91 91 0 MC-SDS SF-MAS 0.79
Servaes et al., 1999 96 96 0 WAI 0.07
Table 1 — continued
Author(s), year countryasumbfemale male REPcdefensivenessdanxietyelogOdd0logOdd1
Weihs et al., 2000 USA 32 5 MC-SDS SF-MAS 0.81
Zachariae et al., 2004 DK 646 646 0 164 MC-SDS SF-MAS 0.21 0.69
King, Taylor, Albright, &
Haskell, 1990
USA 120 60 60 29 MC-SDS SF-MAS 1.70
cek et al., 1998 NL 417 90 310 MC-SDS-mod WAI-mod 0.36
cek et al., 1999 NL 1061 1061 g— MC-SDS WAI-mod 0.54
Perini et al., 1994 USA 87 23 64 13 MC-SDS SF-MAS 1.69
Coronary heart disease
Denollet, 1991 BEL 178 0 178 62 MC-SDS-mod STAI 0.29
Denollet, 1999 409 49 360 97 MC-SDS STAI 0.25
Table 1 — continued
Author(s), year countryasumbfemale male REPcdefensivenessdanxietyelogOdd0logOdd1
Denollet et al., 2008 BEL 731 75 656 159 MC-SDS STAI 0.36
Shaw et al., 1986 USA 97 14 83 25 MC-SDS-mod SF-MAS 0.078
Heart attack
Frasure-Smith et al., 2002hCND 1376 473 903 408 MC-SDS STAI 0.07
Ginzburg et al., 2002 ISR 188 43 145 49 MC-SDS TMAS 0.22
Cooke et al., 2003 GB 21 15 6 7 MC-SDS SF-MAS 0.13
González-Freire et al., 2010 ESP 75 49 26 18 MC-SDS STAI 0.25
Note. a: DK: Denmark; ISR: Israel; NL: Netherlands; BEL: Belgium; CND: Canada; GB: Great Britain; ESP: Spain; —: unknown;
b: not all participants indicated gender, thus data can differ; c: REP: repressors; d: MC-SDS: Marlowe-Crowne Social Desirability
Scale, MC-SDS-mod: modified version of MC-SDS; e: SF-MAS: Short Form Manifest Anxiety Scale, TMAS: Taylor Manifest
Anxiety Scale, WAI: Weinberger Adjustment Inventory, WAI-mod: modified version of WAI; f: 0 indicates effect size before
diagnosis, 1 indicates effect size after diagnosis of disease; g: data of 906 male participants had been requested, but were unavailable
until completion of the meta-analysis; h: descriptive data taken from Frasure-Smith et al. (1997).
Table 2
Statistical data of the analyses.
CI95% Heterogeneity
Model k M(E S)min max p Q pQFSN
RE complete 22 0.27 0.04 0.49 .0197 75.06 <.0001 19
RE complete CG 9 0.53 0.19 0.86 .002 16.17 .04 61
FE cancer t02 0.20 0.26 0.66 .4 0.02 .89
RE cancer t110 0.41 0.09 0.73 .012 22.29 .01 44
RE CVD 10 0.21 0.14 0.55 .24 42.81 <.0001 —
FE Hypertension 4 0.59 0.32 0.86 <.0001 10.25 .02 12
FE CHD 4 0.22 0.39 0.05 .0095 6.80 .08 10
FE Heart attack 2 0.05 0.11 0.21 .527 0.69 .40
FE Asthma 2 0.15 0.78 0.47 .63 0.26 .61
Note. k: Number of studies; RE: Random-Effects Model; FE: Fixed-Effect Model; M(ES): mean
effect size (logOdd s); FSN: fail-safe number according to Rosenberg (2005).
Figure 1: Flow chart of the conducted literature search.
... This assumption is supported by ample evidence indicating that repressors exhibit greater avoidance of internal arousal states, as indicated by a discrepancy between physiological and subjective stress responses: selfreported anxiety is low, whereas the physiological response is strong, compared to the other coping groups (Derakshan & Eysenck, 1997;Derakshan et al., 2007;Kohlmann et al., 1996;Newton & Contrada, 1992). This decreased awareness of their own physical condition is often taken as an explanation for the higher prevalence of stress-related illnesses among repressors (Denollet et al., 2008;Frasure-Smith et al., 2002;Giese-Davis et al., 2006;Mund & Mitte, 2012;Myers, 2010;Schwerdtfeger et al., 2006), which contrasts the tendency toward positive ratings in mental health questionnaires (e.g. coping with bereavement: Coifman et al., 2007;alexithymia: Myers & Derakshan, 2015; adjustment to severe illness: Phipps & Srivastava, 1997). ...
... In summary, the results of this study mildly suggest that the difference between repressors and sensitizers in their preference for certain stimuli is considerably more intense for positive stimuli than for negative stimuli. The higher prevalence of stress-related illnesses in repressors (Denollet et al., 2008;Frasure-Smith et al., 2002;Giese-Davis et al., 2006;Mund & Mitte, 2012;Myers, 2010), which is often attributed to their decreased awareness of their own physical states (e.g., Myers, 2010;Schwerdtfeger et al., 2006), could therefore also be related to this information processing bias towards positive stimuli. In recent years, a new branch of research has developed that not only measures cognitive biases, but seeks to modify them (e.g., Beard, 2011;MacLeod & Mathews, 2012). ...
Full-text available
Individual differences in emotional coping styles are likely to affect information processing on different stages. Repressive coping is assumed to be related to an attentional bias away from threatening information. Possible links to biases in later stages of information processing have not been investigated to date. In the current study, 82 participants completed the visual dot-probe task as a measure of attentional bias and the Approach-Avoidance Task (AAT) as a measure of approach/ avoidance bias and classified into coping groups via the Mainz Coping Inventory (MCI). Prevalence of attention bias and approach/avoidance bias were compared between groups. Main results revealed a strong approach tendency toward positive stimuli for repressors and a strong avoidance tendency for sensitizers. No group differences were found for approach bias to negative stimuli or for attention bias. The present findings of strong preferential processing of positive stimuli in repressors may be part of broader information processing alterations, which may also be linked to alterations in emotion processing.
... However, when feelings of anxiety become persistent and excessive, individuals may have a higher risk for cardiovascular diseases [4] or may even be diagnosed with an anxiety disorder when symptoms lead to functional impairments in daily life [5]. Coping strategies in response to stressful events have been implicated in the pathogenesis of mental and physical disorders (e.g., [6][7][8]). Therefore, coping skills are potential targets for prevention or intervention programs [9]. ...
Full-text available
Cognitive coping strategies to deal with anxiety-provoking events have an impact on mental and physical health. Dispositional vigilance is characterized by an increased analysis of the threatening environment, whereas cognitive avoidance comprises strategies to inhibit threat processing. To date, functional neuroimaging studies on the neural underpinnings of these coping styles are scarce and have revealed discrepant findings. In the present study, we examined automatic brain responsiveness as a function of coping styles using functional magnetic resonance imaging. We administered a perceptual load paradigm with contemptuous and fearful faces as distractor stimuli in a sample of N = 43 healthy participants. The Mainz Coping Inventory was used to assess cognitive avoidance and vigilance. An association of cognitive avoidance with reduced contempt and fear processing under high perceptual load was observed in a widespread network including the amygdala, thalamus, cingulate gyrus, insula, and frontal, parietal, temporal, and occipital areas. Our findings indicate that the dispositional tendency to divert one’s attention away from distressing stimuli is a valuable predictor of diminished automatic neural responses to threat in several cortical and subcortical areas. A reduced processing in brain regions involved in emotion perception and attention might indicate a potential threat resilience associated with cognitive avoidance.
... Previous research has shown that individuals who demonstrate a tendency to inhibit the experience and expression of feelings (repressive coping) show physiological reactions (e.g., cardiac reactivity, increased skin conductance) that are not compatible with subjective ratings of distress. As such, those repressing and unable to disclose and express their feelings are more likely to experience poor health and somatic illness such as cancer and coronary heart disease (Mund & Mitte, 2012). ...
Full-text available
To help organizations find and nurture leaders to be prepared for the complexity and multiplicity of challenges, Lombardo and Eichinger (2000) developed the framework characterized as learning agility. Their research, along with other colleagues (De Meuse, 2022; De Meuse et al., 2011) have typically identified five dimensions (observable behaviors) to learning agility: mental agility, people agility, change agility, results agility, and self-awareness. Although in the early stages of research, there is a promising body of evidence showing learning agility as a significant predictor of identifying potential leaders, leader performance, and leader success (Burke & Smith, 2019; De Meuse, 2019). This article summarizes the current conceptualization of learning agility, provides a neuroscience lens, and identifies associated key brain structures involved in enhancing learning agility. Specific evidence-based neuroscience hacks for practitioners are introduced to enhance each of those five more popularly cited dimensions of learning agility.
... Major reasons for choosing the Model of Coping Modes (Krohne, 1989), were health-related findings regarding repressive coping mechanisms. Research has shown that repressive coping (high CAV, low VIG coping pattern) is associated with poor physical health and diseases such as heart disease, hypertension, types of cancer, (e.g., Di Giuseppe et al., 2018;Gleiberman, 2007;Mund & Mitte, 2012;Myers, 2010). Therefore, the focus on CAV and VIG provides on the one hand a great opportunity to advance research on repressive coping in relation to health, especially to repressive behavior in medical settings. ...
Full-text available
Modern science understands stress as a complex process, and various aspects thereof have been associated with poor health. However, defining the term still is difficult and available stress assessment tools rely on different conceptualizations and target various aspects of stress. The aim of this dissertation is to clarify stress assessment components; thereby, the focus lies on three gaps in stress assessment. The empirical studies summarized in this thesis aim at closing these gaps by 1) Offering a new life stress assessment tool, 2) Discussing a new multi-modal approach to assess a laboratory stressful task, and 3) Presenting a new developed stimulus-response inventory to assess coping. Study one presents the German version of the ‘Stress and Adversity Inventory’ for adults to assess life stress exposure (Sturmbauer, Shields, et al., 2019). Study two presents an adapted real-life situation as an example for a new stress assessment protocol to experimentally induce arousal by investigating simulated medical consultations (Sturmbauer et al., 2022). Study three presents the ‘Coping Inventory for Medical Situations’ as a new tool to assess cognitive avoidant and vigilant coping addressed for researchers and clinicians working with patients in medical care (Sturmbauer, Hock, et al., 2019). Consequently, the three empirical studies are integrated into current research and matched with a published stress typology for precise stress assessment. Finally, future directions will emphasize the future applicability of the presented tools and conclude this thesis
... Clinical studies have found that adults with chronic pain who are identified as repressors are more likely to self-manage their pain and thus would not present to primary health care teams as often (Franklin et al., 2014). This is particularly concerning as repressive coping has been associated with an increased risk of somatic diseases, such as heart disease and cancer (Mund and Mitte, 2012). Given the risk of overlooking children with repression in primary health care teams, one could consider administering a brief social desirability scale as part of an overall evaluation to help identify children who respond in socially desirable ways. ...
Full-text available
Repression has been linked to greater illness, somatic symptoms, and poorer physical health, both in adult and pediatric populations. The current study examined psychological and pain profiles of children with chronic pain who may under-report levels of psychological distress at a first interdisciplinary chronic pain assessment. Children and their caregiver completed measures of psychopathology and pain intensity, while clinicians rated their levels of disability. Based on self-report measures, children were classified as “repressors” (low anxiety/high social desirability) or as “true low anxious” (low anxiety/low social desirability). Groups were then compared on psychological and pain characteristics. Compared to children with true low anxiety, repressors reported lower levels of depressive and somatic symptoms but provided higher ratings on pain intensity, pain–unpleasantness, and self-oriented perfectionism. Caregivers of repressors rated their children as having higher levels of adaptability compared to caregivers of children in the true low anxious group. Groups did not differ on clinician-rated level of disability. Children classified as repressors exhibited different profiles than children classified as having true low anxiety on both psychological outcomes and pain characteristics. Repression may be an important factor to consider for those assessing and treating children with chronic pain.
... Depression seriously damages the physical and mental health of the elderly in rural China (6,7). For instance, depressive symptoms increase the incidence rate of acute diseases (such as heart disease and cancer) and chronic diseases (such as high blood sugar, high cholesterol, high blood pressure, and arthritis) among the rural elderly (8). A study conducted in China found that the per annual capita medical cost of adults with depressive status in 2012 was estimated to be 268.8 ...
Full-text available
Objectives: This study was designed to explore the effect of the New Rural Pension Scheme on depressive symptoms or medical costs induced by depression. Methods: We used the Logit, OLS and 2SLS models to explore the impact of the pension on depression and medical costs. We also adopt the method of quantile regression and discontinuity regression to verify the causal relationship between the New Rural Pension Scheme and depression or medical costs induced by depression of the rural elderly. Results: We have found that the New Rural Pension Scheme decreases depressive symptoms of elderly in rural China (OR = 0.90), and the medical costs induced by depressive symptoms by 4.6%. Regression discontinuity results showed that pension significantly reduced the depressive symptoms (depression) and the medical expenditure caused by depressive symptoms (depression) by using parametric and non-parametric methods, and performing a placebo test. The mediating effect results showed that pension may improve mental health by increasing confidence about the future. Conclusion: We demonstrate that the pension significantly decreased both mental health problems and the medical expenses due to depressive symptoms and depression of elderly in rural China. Therefore, our results suggest that the Chinese government should perfect the New Rural Pension Scheme to eliminate barriers to mental health resources, especially for the rural elderly.
... Freud's work on hysteria focused attention on a person's unconscious reactions to unpleasant, anxiety-provoking thoughts and feelings. Mund and Mitte (34) noted that the defense operations used (Freud called the process repression) were described as pathogenic (conversion neurosis in hysteria). Nowadays, it is clear that regulatory defensive mechanisms have an essential function to which an individual can "automatically" fall back in stressful situations. ...
Full-text available
Objectives: The concept of defense mechanisms has undergone extensive revision and expansion since Freud first described these processes. Initially formulated as an unconscious repression of unpleasant memories, with further development focusing on the role of defense mechanisms in the regulation of internal conflicts, the concept shifted and evolved to incorporate the adaptation to external demands, including intrapsychic and interpersonal handling of burden of illness. In addition to defense mechanisms, coping provides another perspective on human adjustment to difficult life events. While there is substantial research on both coping and defense mechanisms in various psychiatric and somatic diseases, including cancer, little is known about defensive regulation, coping, and their interaction in male breast cancer patients. Methods: The present study is part of the N-Male project conducted between 2016 and 2018 in Germany (Male breast cancer: patients' needs in prevention, diagnosis, treatment, rehabilitation, and follow-up care). Semi-standardized interviews with 27 male breast cancer patients were analyzed with regard to defense mechanisms. In addition, fear of progression and repressive coping was assessed by self-report. Results: There was considerable variety in levels of defensive functioning as well as repressive coping in our sample. We found no difference in overall levels of defensive functioning between men with vs. without repressive coping. However, patients with repressive coping demonstrated a decopupled association between fear of progression and defensive functioning as compared to patients without repressive coping. Discussion: The study provides the first evidence of disease processing in male breast cancer patients Knowledge of patients' defense patterns and repressive coping seems promising for better planning targeted intervention strategies.
Duygu düzenleme; duyguların oluşumunu, çeşidini, yoğunluğunu, zaman akışındaki yerini ve ifadesini belirlediğimiz bilinçli ve bilinçdışı süreçleri içermektedir. Yapılan çalışmalar, duygu düzenlemenin ruh sağlığı üzerindeki kritik etkisinin yanı sıra, fiziksel sağlık üzerinde de önemli etkilerinin olduğuna işaret etmektedir. Özellikle kronik hastalığı olan bireylerde duyguların nasıl düzenlendiği öz-bakım, tedaviye uyum ve stres kaynaklarıyla baş etme gibi pek çok süreci etkileyerek hem fiziksel hem de psikolojik iyilik hali için belirleyici olmaktadır. Bunun yanında duygu düzenlemeyle ilgili sorunların bazı kronik hastalıkların etiyolojisinde ve seyrinde rol oynadığına dair kanıtlar mevcuttur. Bu derleme makalesinin amacı, kronik hastalıklarda duygu ve duygu düzenlemeye dair bilimsel çalışma bulgularını özetlemek ve mevcut çalışmalardaki eksiklere değinerek gelecek çalışmalar için öneriler sunmaktır. Bu amaçla, duygular ve sağlık ilişkisine dair mekanizmaların kısaca özetlenmesinin ardından; sırasıyla, kalp ve damar sistemi (KDS) hastalıkları, kanser, migren, fibromiyalji ve sedef hastalığında duygu düzenlemeyle ilgili bulgulara değinilmiştir. Özetle, KDS hastalıklarının etiyolojisinde duygu ve duygu düzenlemenin rolüne dair kanıtların güçlü olduğu görülmekteyken, kanser için etiyolojiye dair bulguların çelişkili olduğu söylenebilir. Migren, sedef hastalığı ve fibromiyalji için ise çalışma bulguları ilişkisel niteliktedir. Genel olarak çalışmalar kronik hastalığı olanlarda olmayanlara göre duygu düzenleme sürecinde aksaklıklar olduğunu göstermektedir. Hastalar arasında işlevsel duygu düzenleme stratejilerinin daha az kullanımı; psikopatoloji belirtileri, bedensel belirtiler ve düşük yaşam kalitesi gibi olumsuz sonuçlarla ilişkilendirilmiştir. Duygu düzenleme süreçlerinin kronik hastalıklardaki rolünün anlaşılması, önleyici çalışmaların planlanması ve kronik hastalığı olan bireylerin yaşam kalitesini arttıracak müdahalelerin geliştirilmesinde rehberlik etmesi açısından önemlidir. Makale, alanyazının genel bir değerlendirmesi ve öneriler ile sona ermektedir.
It is critical to investigate the role of psychological and psychosocial factors involved in cancer diagnosis, in order to improve prevention and treatment strategies. Two hundred and six Italian participants (healthy group, 118 adult volunteers, 53.4% female; target group, 88 patients with first-time cancer diagnosis, 55.7% female) completed questionnaires measuring maladaptive personality traits, alexithymia, and traumatic events. Controlling for gender, age, and education, results showed that higher age and lower education, together with higher exposure to traumatic events and increased levels of negative affectivity, were significant predictors of belonging to the target group. Our findings suggest the need of an early assessment of psychological functioning and traumatic history in cancer patients, in order to promote more effective prevention strategies and tailored treatment approaches.
Nowadays, the comedy industry is threatened by Cancel Culture and comedians are cautious about sensitive themes. Dave Chapelle and Ali Wong are standup comedians in American, their comedy performances The Closer and Baby Cobra released on Netflix were considered ‘insulting’ and ‘offensive’ by cancel culture. This article examines the relationships among cancel culture, comedians, and commercial institutions by the case study of Dave Chapelle and Ali Wong. Exploring whether comedy is limited by the censorship system and cancel culture. I argue that the cancel culture supporters will pressure on institutions that have real power on the cancelling of public figures. To prevent misjudging the intentions behinds the jokes, the cancel culture should examine the ‘offensive’ materials carefully. I will further explain the argument in detail by focusing on the rebellious nature of comedy, the multiple stakeholders involved in the censorship process, celebrity Schadenfreude culture, freedom of speech, and the price of repression, all of which are likely to make the censorship of comedy short-sighted.
Full-text available
The question of a cancer-prone personality has been extensively researched, yet few if any studies have employed a premorbid data base. Thus, nearly all evidence has been confounded, most importantly by alterations of psychological functioning due to the disease itself. The present study provides a more equitable test of the cancer proneness hypothesis. Premorbid MMPI records were collected from 75 cancer and 125 noncancer patients of a VA hospital. It was reasoned that the hypothesis of the existence of psychologically related cancer proneness would gain some support if stepwise discriminant function analysis of the MMPI scores yielded a significant discrimination between the cancer and noncancer groups. Results support this contention. The group of men with cancer (irrespective of site) was significantly separated from the noncancer group primarily on the basis of lower scores on the Repression-Sensitization Scale (i.e., greater repression) and on the Depression scale of the MMPI (i.e., less self-report of depression). Results are discussed in terms of their theoretical, heuristic, and clinical relevance. (34 ref)
Full-text available
A long-standing problem in stress research has been that individuals' reports of their tendencies to become anxious are often inconsistent with relevant behavioral and physiological indices. This study investigated the distinction between (a) truly low-anxious Ss, who report low trait anxiety on the Taylor Manifest Anxiety Scale and low defensiveness on the Marlowe-Crowne Social Desirability Scale, and (b) repressors, who report low anxiety but high defensiveness. These groups were compared with a moderately high-anxious one. Heart rate, spontaneous skin resistance responses, and forehead muscle tension were recorded from 40 male college students during a phrase association task. Significant differences in the 3 physiological measures as well as in 3 behavioral ones (reaction time, content avoidance, and verbal interference) all indicated that the repressors were more stressed than the low-anxious Ss despite their claims of lower trait anxiety. The high-anxious group exhibited a 3rd pattern suggesting an intermediate level of anxious responding. These data document the need to distinguish between repressors and truly low-anxious persons in research concerned with relations between self-reported anxiety and behavioral and physiological responses to stress. (42 ref)
Although the concept of the defense mechanism was rejected from academic psychology for a number of years, recent empirical studies show renewed interest in defenses. Cognitive psychologists have confirmed the existence of unconscious psychological processes, a requisite for defenses. Developmental, personality, and social psychologists have all found evidence for defense mechanisms that explicate psychological functioning. The relevance of this new information for clinical practice is discussed.
This study investigated the relationship between repressive coping style as defined by Weinberger (1990)-low imported anxiety and high reported defensiveness-and several individual difference measures related to general coping styles. Twenty-nine studies looking at cognitive, behavioural, and individual difference factors: associated with repressive coping were tabulated and reviewed. In the current study 116 women and 32 men were administered a questionnaire including measures of personality, ways of coping, self-monitoring, interpersonal influence, and locus of control to further explore the repressor coping style. Two measures, as opposed to the usual one, were used to categorize repressors the use of anxiety and social desirability measures of Weinberger, Schwartz and Davidson (1979) and the Gudjonsson (1981) method using Neuroticism and Lie scales from the EPQ(R). Analysis of variance showed that repressors reported using mole positive/healthy and less negative/unhealthy coping styles. The results did not change substantially when more extreme scoring groups were used. Factor analysis of the scales revealed six clear factors relating to ways of coping again, repressors scored highly on the factor denoting positive coping. It was also found that the Lie and Neuroticism scales of the revised EPQ personality questionnaire of Eysenck, Eysenck and Barrett (1985) could be used as a substitute for anxiety and defensiveness to predict repression with considerable success. However results front both one- and two-way analysis of variance indicated that the two different ways of classifying repressors was not totally inter-changeable as the overlap in results occurred only with highly significant effects. The results ave discussed in the light of previous investigations into repression which suggested that repressors report an overly optimistic way of coping in order to avoid negative affect. Copyright (C) 1999 John Wiley & Sons, Ltd.
To properly test the hypothesis that two independent methods for classifying individuals (e.g., according to their tendency to exhibit a particular trait) are associated and thus lead to similar assignments, it is necessary to demonstrate that (a) the methods assess similar constructs-that is, the scale dimensions are correlated-and (b) the interrater agreement between the methods is strong. Egloff and Hock (1997) failed to meet these criteria when testing the association between two measurement schemes for assessing dispositional repression-one based on anxiety and defensiveness (Weinberger, Schwartz, & Davidson, 1979), the other on vigilance and cognitive avoidance (Krohne, 1993). Additional analyses performed on the researchers' data demonstrate a weak association between the scale dimensions and, overall, poor agreement between these classification methods. Future research is needed before integrating data from the two models of repression.