European Eating Disorders Review

Eur. Eat. Disorders Rev. 16, 155–159 (2008)

Emotional Awareness and CoreBeliefs Among Women withEating Disorders

*Correspondence to: Glenn Waller, VinceCNWL NHS Foundation Trust, Osbert Stre2QU, UK. Tel: þ44-20-8237-2104. Fax: þ44-E-mail: glenn.waller@iop.kcl.ac.uk

Copyright # 2007 John Wiley & Sons, Ltd a

Published online 6 December 2007 in Wiley In

Rachel Lawson1,2, Francesca Emanuelli3,Jennie Sines4 and Glenn Waller1,5*1Institute of Psychiatry, King’s College London, UK2South Island Eating Disorders Service, Christchurch, New Zealand3Department of Clinical Psychology, University of East London, London, UK4Department of Human Sciences, University of Loughborough,Loughborough, UK5Central and North-West London NHS Foundation Trust, UK

Patients with eating disorders have been shown to experience theemotional components of alexithymia—difficulties in identifyingand describing emotions. In keeping with cognitive theories,which stress the role of schema-level beliefs in understandingemotions, this study examined the core beliefs that are associatedwith this difficulty in women with eating disorders. Seventy eat-ing-disordered women completed standardised measures of corebeliefs and alexithymia. There were no differences in alexithymiabetween diagnostic groups, so the women were treated as a single,transdiagnostic group. Multiple regression analyses showedspecific patterns of association between the core beliefs and theemotional elements of alexithymia. Difficulties in identifyingemotions were associated with entitlement beliefs, while difficul-ties in describing emotions were associated with both abandon-ment and emotional inhibition beliefs. These findings suggest thatit may be necessary to work with core beliefs in order to reducelevels of alexithymia, prior to addressing the emotions that driveand maintain pathological eating behaviours. Copyright # 2007John Wiley & Sons, Ltd and Eating Disorders Association.

Keywords: eating disorders; emotions; alexithymia; core beliefs

INTRODUCTION

Emotional factors have been identified as ante-cedents and triggers of a significant number ofeating-disordered behaviours (e.g. Arnow, Agras, &Kenardy, 1995; Fairburn, Cooper, & Shafran, 2003;Meyer, Waller, & Waters, 1998). However, working

nt Square Clinic,et, London SW1P20-8237-2280.

nd Eating Disorders

terScience (www.inte

with the relevant emotions is often very difficultin clinical practice, as many women with eatingdisorders appear to have deficits in processingaffective states (e.g. Bydlowski et al., 2005), oftenrelated to personality-level difficulties (e.g. Sureda,Valdes, Jodar, & de Pablo, 1999). Substantial research(Cochrane, Brewerton, Wilson, & Hodges, 1993;Corcos et al., 2000; Taylor, Parker, Bagby, & Bourke,1996; Troop, Schmidt, & Treasure, 1995) has shownthat this difficulty manifests as the affectivecomponents of alexithymia (problems in identifyingand expressing emotional states), although the

Association.

rscience.wiley.com) DOI: 10.1002/erv.848

156 R. Lawson et al.

cognitive element of alexithymia (poor externallyoriented thinking) does not appear to be relevant towomen with eating disorders.There is only limited evidence that diagnostic

groups differ in levels of the elements of alexithymia(e.g. Montebarocci, Codispoti, Surcinelli, Franzoni,Baldaro, & Rossi, 2006; Sexton, Sunday, Hurt, &Halmi, 1998), suggesting that this construct is besttreated as one that is applicable broadly to the eatingdisorders, rather than being specific to anorexic orbulimic manifestations. The importance of alex-ithymia is shown by the fact that it is predictiveof poor outcome from the eating disorders (e.g.Speranza, Loas, Wallier, & Corcos, 2007), althoughthis is not a universal finding (e.g. Becker-Stoll &Gerlinghoff, 2004). It is also important to note thattreatment of the eating disorders does not havean impact on the level of alexithymia (e.g. Iancu,Cohen, Yehuda, & Kotler, 2006).In order to address these difficulties in emotional

processing clinically, it is important to understandthe cognitions that underlie alexithymia in individ-ualswith eating disorders. In linewith Beck’smodelof cognitions and emotions (e.g. Beck, Rush, Shaw,& Emery, 1979), emotional states and processing arelikely to be linked to core beliefs (unconditional,schema-level cognitions). This study will examinelinks between core beliefs and alexithymia inwomen with eating disorders. In particular, it canbe hypothesised that difficulties in identifyingand describing emotions will be associated withemotional inhibition beliefs (the belief that it isnot acceptable or safe to experience or expressemotions).

METHOD

Participants

The participants were 70 eating-disordered women,with DSM-IV diagnoses (American PsychiatricAssociation, 1994) of anorexia nervosa (N¼ 11),bulimia nervosa (N¼ 21) or Eating Disorder NotOtherwise Specified (EDNOS; N¼ 38). Each patientwas diagnosed by clinicians using strict DSM-IVcriteria. Their mean age was 28.5 years (SD¼ 8.66),and their mean BMI was 22.1 (SD¼ 6.51). All thesepatients were recruited from a series of consecutivereferrals to a specialist eating disorders clinicbetween January and July 2004. A further 28patients failed to complete the measures, eitherfully or in part. The inclusion criteria were beingfemale and having a DSM-IV diagnosis of one of the

Copyright # 2007 John Wiley & Sons, Ltd and Eating Disorders A

three eating disorders detailed above. However,patients were excluded if they had a primarylearning difficulty, extreme emaciation (interferingwith capacity to consent and answer meaningfully)or a psychotic disorder. The proportion of patientsfrom each diagnostic group in the referrals was notequal, accounting for the unbalanced number ofparticipants in each group.

Measures and Procedure

Each patient was given the following two self-reportmeasures as part of their assessment for entry to aspecialist eating disorders service. They returnedthe measures in person or by post, using a pre-paid envelope. Their weight and height weremeasured objectively, to yield an accurate bodymass index (BMI¼weight [kg]/height [m]2).

Toronto alexithymia scale—20-item version(TAS-20; Bagby, Parker & Taylor, 1994)The TAS-20 is a 20-item measure of the three core

elements of alexithymia—difficulty in identifyingfeelings and distinguishing them from bodily sensa-tions; difficulty describing feelings to others; andexternally oriented thinking. It is widely used formeasuring the construct of alexithymia. The TAS-20has good psychometric properties, with acceptableinternal consistencies (alpha of 0.81–0.64) and a clearfactor structure in clinical and non-clinical groups(Bagby, Parker & Taylor, 1994). It has substantialconcurrent and convergent validity (Bagby, Parker& Taylor, 1994), in a wide range of cultural andlinguistic settings (Taylor, Bagby, & Parker, 2003).Fifteen of the items are positively scored, and fiveare reverse-scored. Higher scores reflect greaterlevels of alexithymia.

Young schema questionnaire-short form(YSQ-S; Young, 1998)The YSQ-S is a 75-item self-report questionnaire,

which measures 15 core beliefs (unconditional,schema-level cognitions about oneself, others andthe world). It has been validated among individualswith eating disorders (Waller, Ohanian, Meyer, &Osman, 2000), with Cronbach’s alpha of greaterthan .80 for all 15 scales in this clinical groupand associations between core beliefs and eatingbehaviours. However, there is little evidence ofdifferences in YSQ-S scores between eating disorderdiagnostic groups (Leung,Waller, & Thomas, 1999).The YSQ-S scales are: emotional deprivation,abandonment, mistrust/abuse, social isolation,dependence/incompetence, vulnerability to harm,

ssociation. Eur. Eat. Disorders Rev. 16, 155–159 (2008)

DOI: 10.1002/erv

ups,

compared

usingMANCOVA

entvariable

MANCOVA

algroup

nervosa

¼21)

EDNOS

(N¼38)

GroupF

(2,65)

AgeF

(1,65)

BMIF

(1,65)

;18.4–25.3)

21.6

(6.81;

19.7–24.7)

0.91

0.52

3.38

;12.7–17.5)

15.9

(4.99;

14.3–17.8)

0.49

0.09

0.82

;19.5–28.2)

19.9

(5.09;

16.6–22.9)

1.15

0.02

0.30

Emotional Awareness and Core Beliefs 157

enmeshment, defectiveness/shame, failure toachieve, subjugation, emotional inhibition, self-sacrifice, unrelenting standards, entitlement, andinsufficient self-control. The score for each scale isthe mean of its five items. A higher score reflects amore maladaptive, unhealthy core belief.

Data Analysis

Initially, levels of alexithymia were comparedacross diagnostic groups, using multivariateanalysis of covariance (controlling for differencesin age and BMI), with pairwise post hoc tests used todetermine the source of any overall group effects.Levene’s test showed that the data were normallydistributed for each of the three TAS scales(difficulty identifying feelings and distinguishingthem from bodily sensations—F¼ 0.97, p¼ .39;difficulty in describing feelings to others—F¼0.05, p¼ .95; externally oriented thinking—F¼2.34, p¼ .11). Thereafter, multiple regressionanalyses (simultaneous entry method) were usedto predict the levels of the different facets ofalexithymia (TAS-20 scores) from levels of corebeliefs (YSQ-S scores).

Tab

le1.

Meanlevelsofalexithymia

(SD;95%

CI)in

theeating-disord

ered

gro

Indep

end

Clinic

Dep

enden

tvariablesTAS-20scales

Anorexia

nervosa

(N¼11)

Bulimia

(N

Difficu

ltyiden

tifyingfeelings/

distinguishingem

otionsfrom

bodilysensations

20.0

(8.24;

13.1–23.5)

22.2

(7.81

Difficu

ltydescribingfeelingsto

others

14.9

(5.44;

10.5–17.8)

15.1

(4.96

Externally

orien

tedthinking

19.8

(4.60;

14.1–27.1)

23.9

(14.6

Fourteennodifferencesbetweengroupsorcovariate

effectswerestatisticallysignificant.

RESULTS

Table 1 shows the alexithymia levels (TAS-20 scores)of the three diagnostic groups. Those scores weresimilar to those of previous samples of eating-disordered women, and higher than those of non-clinical women (e.g. Taylor et al., 1996). In keepingwith much previous research (Troop et al., 1995),there were no statistically significant differencesbetween the groups, so they were treated as a singletransdiagnostic group for the purposes of sub-sequent analyses. To ensure replicability, it isnecessary to include the YSQ-S scores for thiseating-disordered group. The women’s YSQ-Sscores were: emotional deprivation¼ 3.33 (SD¼1.58), abandonment¼ 3.32 (SD¼ 1.61), mistrust/abuse¼ 3.25 (SD¼ 1.51), social isolation¼ 3.35(SD¼ 1.71), defectiveness/shame¼ 3.47 (SD¼ 1.68),failure to achieve¼ 3.28 (SD¼ 1.63), dependence/incompetence¼ 2.78 (SD¼ 1.32), vulnerability toharm¼ 2.78 (SD¼ 1.40); enmeshment¼ 2.16 (SD¼1.39), subjugation¼ 3.01 (SD¼ 1.41) self-sacrifice¼3.84 (SD¼ 1.31), emotional inhibition¼ 2.97 (SD¼1.49), unrelenting standards¼ 4.42 (SD¼ 1.36), enti-tlement¼ 2.31 (SD¼ 0.91) and insufficient self-control¼ 3.07 (SD¼ 1.20).

Copyright # 2007 John Wiley & Sons, Ltd and Eating Disorders Association. Eur. Eat. Disorders Rev. 16, 155–159 (2008

DOI: 10.1002/erv

)

Table 2. Multiple regression analyses, showing the association of core beliefs (YSQ-S scores) with alexithymia (TAS-20scores) among eating-disordered women (N¼ 70)

Dependent variable (TAS-20 scale) Overall effect Individual effects

Overall F p Adjusted R2 Significantindependent variables

(OMNI scales)

T p B (95%CI)

Difficulty identifying feelings/distinguishing emotions frombodily sensations

8.01 .001 .648 Entitlement 2.04 .05 1.89 (0.02–3.76)

Difficulty describing feelingsto others

5.93 .001 .569 Vulnerability to harm 2.26 .03 1.62 (0.17–3.07)

Emotional inhibition 2.69 .01 1.36 (0.34–2.39)Externally oriented thinking 1.25 .28 .066 — — — —

158 R. Lawson et al.

Table 2 shows the results of multiple regressionanalyses used to predict the levels of alexithymiafrom the women’s core beliefs. The core beliefs werenot associated with the women’s levels of externallyoriented thinking. Difficulty in identifying feelingsand in distinguishing those feelings from somaticexperiences was associated with high levels ofentitlement beliefs, suggesting that narcissisticthinking styles are linked to a lack of introspectiveability regarding emotional states. In contrast,difficulty in describing feelings was associatedwith two core beliefs–higher levels of vulnerabilityto harm and emotional inhibition. This patternsuggests that women with eating disorders whohave high levels of anxiety-related cognitions andwho tend to avoid the experience of emotions areless able to externalise their emotions so that otherscan understand them.

DISCUSSION

The levels of alexithymia and core beliefs in thisstudy were comparable with those of previousresearch using the same measures (e.g. Tayloret al., 1996; Waller et al., 2000). As hypothesised,there were significant and specific associationsbetween core beliefs and the affective componentsof alexithymia among this group of eating-disordered women, which is an extension of theconclusions of Taylor et al. (1996). In contrast, therewere no links to the cognitive element (difficultiesin externally oriented thinking). In other words,schema-level beliefs are associated specificallywith difficulties in identifying and describingemotions. However, the patterns of associationwere different across the two affective componentsof alexithymia. Problems in identifying one’semotional states and differentiating them fromsomatic states were linked with higher levels of

Copyright # 2007 John Wiley & Sons, Ltd and Eating Disorders A

entitlement belief (the assumption that one’sown wishes are paramount, and that one can actwithout considering others). Difficulty in describ-ing one’s feelings to others was associated withhigh levels of vulnerability cognitions (beliefs thatunderpin anxiety) and emotional inhibition cogni-tions (belief that emotional expression or experi-ence will have aversive consequences). Thesefindings are in need of extension with larger,more evenly matched clinical samples, includingcomparison with both clinical and non-clinicalcontrols. The demographic data of such samplesshould be considered to determine whether suchfeatures are related to these psychological vari-ables.As emotional factors are important for many

individuals with eating disorders (e.g. Fairburnet al., 2003;Meyer et al., 1998), it will be important toaddress emotional triggers and maintaining factors(e.g. Corstorphine, 2006) where those factors arepresent. However, such emotions and their under-lying cognitions are not always readily accessiblefor report and modification, and this difficultymanifests as alexithymia. The present findingssuggest that working with the emotional elementsof alexithymia in women with eating disordersrequires the addressing of specific schema-levelcognitions (entitlement, emotional inhibition, aban-donment). Addressing such beliefs in individualswith eating disorders requires a range of techniques(Waller, Kennerley, & Ohanian, 2007), centringon Socratic methods for cognitive restructuringand behavioural experimentation. For example, theindividual might need to be encouraged to considerthe origins of her emotional inhibition belief, and toconsider the evidence that the belief (which wasoriginally adaptive to the individual’s environment)has become maladaptive as the environment haschanged, leading to the possibility of different ways

ssociation. Eur. Eat. Disorders Rev. 16, 155–159 (2008)

DOI: 10.1002/erv

Emotional Awareness and Core Beliefs 159

of coping with emotions in the here and now (testedout using behavioural experiments). Further researchis needed to test the feasibility of this approach.The reduction of alexithymia will give access to

emotions, without necessarily modifying thoseaffective states. Therefore, it will also be importantto enhance the individual’s capacity for distresstolerance (e.g. Linehan, 1993) and to develop morepositive emotion coping skills (e.g. Corstorphine,2006). Finally, it is also necessary to address theunderlying concerns regarding eating, weight andshape (e.g. Fairburn et al., 2003) and to reduce theimpact of starvation, which can impair emotionalstability (e.g. Keys, Brozek, Henschel, Mickelsen, &Taylor, 1950). Addressing these eating and starva-tion-related issues will usually need to take placefrom early on in the treatment of individuals witheating disorders, but their effectiveness is likely tobe limited if the emotional issues are not addressed.

REFERENCES

American Psychiatric Association. (1994). Diagnostic andstatistical manual of mental disorders (4th ed.). Washing-ton, DC: American Psychiatric Association.

Arnow, B., Kenardy, J., & Agras, W. S. (1995). TheEmotional Eating Scale: The development of ameasureto assess coping with negative affect by eating. Inter-national Journal of Eating Disorders, 18, 79–90.

Bagby, R. M., Parker, J. D. A., & Taylor, G. J. (1994). Thetwenty-Item Toronto Alexithymia Scale—I. Item selec-tion and cross-validation of the factor structure. Journalof Psychosomatic Research, 38, 23–32.

Bagby, R. M., Parker, J. D. A., & Taylor, G. J. (1994). Thetwenty-item Toronto Alexithymia Scale-II: Conver-gent, discriminant, and concurrent validity. Journal ofPsychosomatic Research, 38, 33–40.

Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979).Cognitive therapy of depression. New York, NY: Guilford.

Becker-Stoll, F., & Gerlinghoff, M. (2004). The impact of afour-month day treatment programme on alexithymiain eating disorders. European Eating Disorders Review,12, 159–163.

Bydlowski, S., Corcos, M., Jeammet, P., Paterniti, S.,Berthoz, S., Laurier, C., Chambry, J., & Consoli, S. M.(2005). Emotion-processing deficits in eating disorders.International Journal of Eating Disorders, 37, 321–329.

Cochrane, C. E., Brewerton, T. D., Wilson, D. B., &Hodges, E. L. (1993). Alexithymia in eating disorders.International Journal of Eating Disorders, 14, 219–222.

Corcos, M., Guilbaud, O., Speranza, M., Paterniti, S., Loas,G., Stephan, P., & Jeammet, P. (2000). Alexithymia anddepression in eating disorders. Psychiatric Research, 93,263–266.

Corstorphine, E. (2006). Cognitive-emotional-behaviouraltherapy for the eating disorders: Working with beliefsabout emotions. European Eating Disorders Review, 14,448–461.

Copyright # 2007 John Wiley & Sons, Ltd and Eating Disorders A

Fairburn, C. G., Cooper, Z., & Shafran, R. (2003). Cognitivebehaviour therapy for eating disorders: A ‘transdiag-nostic’ theory and treatment. Behaviour Research andTherapy, 41, 509–528.

Iancu, I., Cohen, E., Yehuda, Y. B., & Kotler, M. (2006).Treatment of eating disorders improves eating symp-toms but not alexithymia and dissociation proneness.Comprehensive Psychiatry, 47, 189–193.

Keys, A., Brozek, J., Henschel, A., Mickelsen, O., & Taylor,H. L. (1950). The biology of human starvation. Minnea-polis: University of Minnesota Press.

Leung, N., Waller, G., & Thomas, G. (1999). Core beliefs inanorexic and bulimic women. Journal of Nervous andMental Disease, 187, 736–741.

Linehan, M. (1993). Cognitive-behavioural treatment of bor-derline personality disorders. New York, NY: Guilford.

Meyer, C., Waller, G., & Waters, A. (1998). Emotionalstates and bulimic psychopathology. In H. W. Hoek,J. L. Treasure, & M. A. Katzman (Eds.), Neurobiologyin the treatment of eating disorders. (pp. 271–289).Chichester, UK: Wiley.

Montebarocci, O., Codispoti, M., Surcinelli, P., Franzoni,E., Baldaro, B., & Rossi, N. (2006). Alexithymia infemale patients with eating disorders. Eating andWeight Disorders, 11, 14–21.

Sexton, M. C., Sunday, S. R., Hurt, S., & Halmi, K. A. (1998).The relationship between alexithymia, depression, andaxis II psychopathology in eating disorder inpatients.International Journal of Eating Disorders, 23, 277–286.

Speranza, M., Loas, G., Wallier, J., & Corcos, M. (2007).Predictive value of alexithymia in patients with eatingdisorders: A 3-year prospective study. Journal of Psy-chosomatic Research, 63, 365–371.

Sureda, B., Valdes, M., Jodar, I., & de Pablo, J. (1999).Alexithymia, type A behaviour and bulimia nervosa.European Eating Disorders Review, 7, 286–292.

Taylor, G. J., Bagby, R. M., & Parker, J. D. A. (2003). The20-Item Toronto Alexithymia Scale: IV. Reliability andfactorial validity in different languages and cultures.Journal of Psychosomatic Research, 55, 277–283.

Taylor, G. J., Parker, J. D. A., Bagby, M., & Bourke, M. P.(1996). Relationships between alexithymia and psycho-logical characteristics associated with eating disorders.Journal of Psychosomatic Research, 41, 561–568.

Troop, N. A., Schmidt, U. H., & Treasure, J. L. (1995).Feelings and fantasy in eating disorders: A factoranalysis of the Toronto alexithymia scale. InternationalJournal of Eating Disorders, 18, 151–157.

Waller, G., Kennerley, H., & Ohanian, V. (2007). Schema-focused cognitive behaviour therapy with eatingdisorders. In L. P. Riso, P. T. du Toit, & J. E. Young(Eds.), Cognitive schemas and core beliefs in psychiatricdisorders: A scientist-practitioner guide (pp. 139–175).New York, NY: American Psychological Association.

Waller, G., Ohanian, V., Meyer, C., & Osman, S. (2000).Cognitive content among bulimic women: The role ofcore beliefs. International Journal of Eating Disorders, 28,235–241.

Young, J. E. (1998). Young Schema Questionnaire—shortform (YSQ-S) (on-line). New York: Cognitive TherapyCentre. (Available: http://www.schematherapy. com).

ssociation. Eur. Eat. Disorders Rev. 16, 155–159 (2008)

DOI: 10.1002/erv