78524110

Terapia psicológica

Sociedad Chilena de Psicología Clínica

sochpscl@entelchile.net

ISSN (Versión impresa): 0716-6184

CHILE

2006

Barbara César Machado / Óscar F. Gonçalves / Paulo P. P. Machado / Margarida R.

Henriques / António Roma Torres / Isabel Brandão

ANOREXIA NERVOSA: DIVERGENT VALIDITY OF A PROTOTYPE NARRATIVE

Terapia psicológica,

junio, año/vol. 24, número 001

Sociedad Chilena de Psicología Clínica

Santiago, Chile

pp. 99-104

Red de Revistas Científicas de América Latina y el Caribe, España y Portugal

Universidad Autónoma del Estado de México

http://redalyc.uaemex.mx

TERAPIA PSICOLÓGICA

2006, Vol. 24, Nº 1, 99–104

ISSN 0716-6184

Anorexia Nervosa: Divergent Validity of a Prototype Narrative

Anorexia Nerviosa:Validez Divergente de un Prototipo de Narrativa

Bárbara César Machado

Universidade do Minho

Óscar F. Gonçalves

Universidade do Minho

Paulo P. P. Machado

Universidade do Minho

Margarida R. Henriques

Universidade do Porto

António Roma-Torres

Hospital de São João, Porto

Isabel Brandão

Hospital de São João, Porto

(Rec: 03 abril 2006 – Acep: 04 mayo 2006)

Abstract

The objective of this paper was to test the divergent validity (degree of discrimination) of an anorectic

prototype narrative (i.e., communality of themes in the individuals’ core auto-biographical memories), as

well as explore different characteristics of the participants which may be associated with the degree of

prototype discrimination. Seventy participants diagnosed with anorexia nervosa participated in the study

and were asked to indicate their degree of identification with four different narrative prototypes (depressive,

agoraphobia, anorexic, alcoholic and drug addiction prototypes). Results did not confirm the divergent validity

of the anorexic prototype narrative. Participants tended to identify primarily with the depressive prototype

narrative. No significant differences were found between levels of identification with the anorexic prototype

and depression, or agoraphobia and alcoholism prototype. The only significant difference found was between

the anorexic and drug addiction prototype. However, severity and duration of the clinical condition were

found to be associated with the degree of identification of prototype narrative. Results are discussed in terms

of a transdiagnostic versus a prototype approach to the eating disorders psychopathology.

Keywords: Narratives, psychopathology, prototype narrative, anorexia, divergent validity

Resumen

El objetivo de éste artículo es tanto poner a prueba la validez divergente (grado de discrimnación) de un

prototipo de narrativa anoréxica (por ejemplo, conjunto de temáticas ligadas a la memoria que resultan centra-

les a la autobiografía de los individuos), como explorar las diferentes características de los participantes que

podrían estar asociadas al grado de discriminación prototípica. Participaron en el estudio setenta pacientes

diagnosticadas con anorexia nerviosa. Se les consultó que indicaran su grado de identificaciópn con cuatro

diferentes prototipos de narrativa (prototipos depresivo, agorafóbico, anoréxico y de adicción a drogas y alco-

hol). Los resultados no confirman la validez divergente de la narrativa prototípica anoréxica. Los participantes

tendieron a identificarse principalmente con la narrativa de tipo depresiva. No se hallaron diferencias significa-

tivas entre los niveles de identificación del prototipo anoréxico y depresivo, ni entre el prototipo agorafóbico y

alcohólico. La única diferencia significativa se encontró entre la narrativa de tipo anoréxica y la drogodependiente.

Sin embargo, se encontró asociación entre la severidad y duración de las conductas clínicas con el grado de

identificación a la narrativa prototípica. los resultados se discuten en términos de un transdiagnóstico versus un

enfoque de los prototipos de la psicopatología de los desórdenes alimentarios.

Palabras clave: Narrativas, psicopatología, prototipos de narrativa, anorexia, validez divergente

Correspondence to: Óscar F. Gonçalves, PhD, Universidade do Minho, Departamento de Psicologia., Campus de Gualtar, P-4710 BRAGA, PORTUGAL.

Tel #: +351-53-604240; Fax#: +351-53-678987; E-mail: goncalves@iep.uminho.pt

100

BÁRBARA CÉSAR MACHADO / ÓSCAR F. GONÇALVES / PAULO P. P. MACHADO /

MARGARIDA R. HENRIQUES / ANTÓNIO ROMA-TORRES / ISABEL BRANDÃO

TERAPIA PSICOLÓGICA 2006, Vol. 24, Nº1, 99–104

Introduction

Several authors have been claiming that psychological

disorders are characterized by specific meaning systems

and that this systems can be best captured in narrative

prototypes (e.g, Gonçalves et al., 2001; Herrmans &

Hermans-Jansen, 1995; Leahy, 1991).

According to Gonçalves et al. (2001), prototype

narratives refer to the communality of themes in the

individuals’ core auto-biographical memories. These

prototypes are hypothesized to differentiate meaning

organization of different psychological disorders.

Following a ground analytic method Gonçalves and

colleagues were able to construct narrative prototypes for

different psychological disorders, namely: Anorexia Nervosa,

Depression, Agoraphobia, Alcoholism and Drug addiction (c.

f., Gonçalves & Machado, 1999). Several studies brought

evidence for the convergent validity of these narrative

prototypes. That is, patients with different psychological

disorders were able to identify with their narrative prototype

(c. f. Gonçalves et al., 2000). In spite of the promising nature

on the convergent validity of the different prototype narratives,

it remains to be understood if individuals with a specific

psychological disorder are able to discriminate between

different narrative prototypes (i. e., divergent validity).

The objective of this study was to test the extent to which

individuals diagnosed with anorexia nervosa were able to

identify differentially with the anorectic narrative prototypes

(comparatively with the depression, agoraphobia,

alcoholism, and drug addiction). Additionally, we explored

different characteristics of the anorectic patients’ sample

that may contribute to differentiate the degree of

identification with the prototype narratives.

Method

Participants

Seventy patients, diagnosed with anorexia nervosa (AN)

according to DSM-IV participated in this study. Forty-eight

(68.6%) with restricting sub-type, and 22 (31.4%) with

binge-eating/purging sub-type. The majority of the

participants were female (68; 91.7%), only two (2.9%) were

male. Their age ranged between 12 to 37 years old (M =

20.29; SD = 5.93). Most were single (65; 92.9%); and

student (51; 72.9%) with an educational level ranging from

middle school to university degree (M = 10.73; SD = 1.48).

The inclusion criterion for this study was the existence

of a current diagnosis of anorexia nervosa according to

DSM-IV (APA; 1994). Any co-morbidity with other axis I

or II disorders constituted exclusion criteria.

The study procedures were approved by the relevant Ethical

Committees and all participants gave their written informed

consent after the procedures were fully explained to them.

Measures

A Clinical Diagnose Questionnaire was designed to assess

participants’ symptoms. This questionnaire was filled in by

the patient’s therapist and also included questions about the

presence of binge eating episodes and compensatory

behaviors (i.e., self-induced vomit, laxatives, diet/low calorie

food intake and excessive exercise) assessed in terms of

frequency and severity (i.e., “not existing"; “up to once a

week/mild"; “2 to 3 times a week/moderate"; “4 times a week

up to daily/severe"; “more that once a week/extreme").

Participants were asked to fill a questionnaire with

demographic and socio-economic information and

questions regarding their eating disorder and treatment

history (i.e., anorexia nervosa course and duration, therapy

sessions and inpatient care).

For the assessment of psychopathology and psychological

distress we used the Symptom Checklist 90-R (SCL 90-R;

Derrogatis, 1977). It is a self-report measure with 90 items

in a five point scale of distress: 9 primary symptom

dimensions (Somatization, Obsessive-Compulsive,

Interpersonal Sensitivity, Depression, Anxiety, Hostility,

Phobic Anxiety, Paranoid Ideation and Psychoticism) and 3

global indices of distress (Global Severity Index, Positive

Symptom Distress Index and Positive Symptom Total).

Eating related attitudes and behaviours were assessed

by the Eating Disorders Inventory (EDI; Garner, Olmsted

& Polivy, 1983). The EDI is a self report measure with 64

items. Participants are asked to answer a six-point forced

choice format. EDI has 8 sub-scales: 3 for attitudes and

behaviours concerning eating, weight and shape (Drive for

Thinness; Bulimia; Body Dissatisfaction) and 5 related to

more general organizing constructs or clinically relevant

psychological traits for eating disorders (Ineffectiveness;

Perfectionism; Interpersonal Distrust; Interoceptive

Awareness; Maturity Fears).

Finally, and to study the divergent validity in individuals

with anorexia nervosa, the Prototype Narratives Hierarchy

Questionnaire (Gonçalves & Henriques, 2000) was used. It

was conceived to elicit a comparative evaluation of the

participant’s identification with each of the five prototype

narratives. Participants’ task was to create a hierarchy,

organizing their identifications by assigning a ranking order

for the degree of identification with each prototype narrative.

Procedure

All participants were recruited in clinical settings by

their staff psychiatrist and referred to the research team

along with the Clinical Diagnose Questionnaire. Then, an

individual interview took place where the different

psychological measures were taken (Demographic and

Socio-Economic Questionnaire; Prototype Narratives

Hierarchy Questionnaire; SCL 90-R; EDI).

ANOREXIA NERVOSA: DIVERGENT VALIDITY OF A PROTOTYPE NARRATIVE

101

TERAPIA PSICOLÓGICA 2006, Vol. 24, Nº1, 99–104

Data Analysis

Nonparametric statistic analyses were used. Anorexia

prototype narrative divergent validity was computed using

Friedman test for the comparing order of identification for

each narrative prototype. A Multiple Comparison Formula

(MCF) was used to test the significance of differences

between each prototype narrative. Finally, a multiple

regression analysis was used to test the predictive power of

different participants’ demographic and clinical variables,

assuming the mean order given to the anorectic prototype

narrative as the dependent variable, and demographic,

diagnose, and psychopathology and eating disorders

symptoms as independent variables.

Results

According to the present symptomatology and clinical

history, the results found show that the majority of the

participants had restricting subtype diagnosis (48; 68.6%).

All were on psychotherapeutic process and almost half of

them had been in inpatient care, at least once (31; 44.3%).

A third (34.3%) of the participants had had anorexia nervosa

disorder for two years and a half or longer. The majority of

the participants (65; 92.9%) lost weight and/or maintained

it under normal range by dieting or fasting. Finally, most

of the means and standard deviation of SCL 90-R and EDI

subscales were (with the exception of EDI Bulimia

subscale) above the cutoff ’s score.

Results on the divergent validity (see Table 1), show

that participants identify in the following order with the

different narrative prototypes: first, depression; then ano-

rexia, agoraphobia, alcoholism, and drug addiction. Only

the comparison between the identification with the ano-

rexia prototype narrative and the drug addiction prototype

narrative (see Table 2) was found to be statistically

significant (

=50>40.6; p<.05)

. There were no significant

differences between the identification with the anorexia

prototype narrative and the other three narratives (i.e.,

depression prototype narrative, agoraphobia prototype

narrative and alcoholism prototype narrative).

Table 1. Mean order of identification with the prototype

narratives (PN) and Friedman’s

for the ordination

results.

Prototype Narratives Mean Order of Identifications

Depression PN

3.31

Anorexia PN

3.12

Agoraphobia PN

3.10

Alcoholism PN

3.06

Drug addiction PN

2.40

(4 g.I) = 13.66, p<.01

Note: The highest mean order value is in bold.

Ano PN

Dep PN

(219)

(233)

-14

-14<40.6; N.S.

Ano PN

Ago PN

(219)

(218)

1<40.6; N.S.

Ano PN

Alc PN

(219)

(215)

4<40.6; N.S.

Ano PN

DrAd PN

(219)

(169)

50>40.6; p<.05

Dep PN

Ago PN

(233)

(218)

15<40.6; N.S.

Dep PN

Alc PN

(233)

(215)

18<40.6; N.S.

Dep PN

DrAd PN

(233)

(169)

64>40.6; p<.05

Ago PN

Alc PN

(218)

(215)

3<40.6; N.S.

Ago PN

DrAd PN

(218)

(169)

49>40.6; p<.05

Alc PN

DrAd PN

(215)

(169)

46>40.6; p<.05

Table 2. Prototype narratives pair’s comparison and statistic significance of their differences using the Multiple Comparison

Formula (MCF) which were obtained, for the order results, the value of 40.6.

Prototype Narratives Pairs

Differences in the

Statistical significance

(order values sum)

order values sum

102

BÁRBARA CÉSAR MACHADO / ÓSCAR F. GONÇALVES / PAULO P. P. MACHADO /

MARGARIDA R. HENRIQUES / ANTÓNIO ROMA-TORRES / ISABEL BRANDÃO

TERAPIA PSICOLÓGICA 2006, Vol. 24, Nº1, 99–104

In addition we conducted exploratory analysis to ex-

plore how different variables might impact the degree of

identification with the anorectic prototype narrative. Several

interesting patterns arose. There was a tendency for highest

identification with the anorexia prototype narrative,

compared with the identification with the drug addiction

prototype narrative, when participants had a highest socio-

economic status (

= 31>20.4; p<.01), when they binged

and/or purged most severely (

= 19>15.22; p<.05); and,

when they had clinically significant scores on SCL 90

paranoid ideation subscale (

= 42>32.88; p<.05) or EDI

(

= 40>34.3; p<.05). Subjects that had the clinical

condition for at least two and a half years and a score above

the cutoff point on the Bulimia EDI subscale had a higher

identification with the anorexia prototype narrative

compared not only with the drug addiction prototype

narrative (

= 40>24.2; p<.01; X2 = 33>24.3; p<.01) but

also with the agoraphobia prototype narrative (

= 27>24.2;

p<.01; X2 = 28>24.3; p<.01) (see Tables 3 and 4).

An opposite pattern of results was found for other

clinical measures. Participants that had therapy sessions

for less than six months had significantly highest

identification with the depression prototype narrative than

with drug addiction prototype narrative (

= 45>23.59;

p<.01) as well as anorexia prototype narrative (

25>23.59; p<.01) (see Table 5). Similarly, participants who

scored bellow the cutoff value in SCL 90 Interpersonal

Sensitivity and Anxiety subscales and EDI Total and

Interoceptive Awareness subscales had a significantly

highest identification with agoraphobia prototype narrative

compared with the drug addiction (

= 28>19.95; p<.05;

= 41>19.95; p<.01;

= 36>19.95; p<.01;

= 43>19.95;

p<.01) and anorexia prototype narratives (

= 20>19.95;

p<.05;

= 31>19.95; p<.01;

= 26>19.95; p<.01;

24>19.95; p<.01) (see Table 6).

Table 4. Divergent validity results in Paranoid Ideation (PI) of SCL 90-R and Total (T) and Bulimia (B) subscales of EDI.

PI SCL 90-R above cut of point

T EDI above cut of point

B EDI above cut of point

n = 45

n = 49

n = 24

Ano PN (150)

Ano PN (164)

Ano PN (88)

Dep PN (146 / 4)

Dep PN (164 / 0)

Dep PN (83 / 5)

Ago PN (138 / 12)

Alc PN (146 / 18)

Alc PN (74 / 14)

Alc PN (133 / 17)

Ago PN (137 / 27)

Ago PN (60 / 28)

DrAd PN (108 / 42)

DrAd PN (124 / 40)

DrAd PN (55 / 33)

(4 g.I) = 9.67, p <.05

(4 g.I) = 9.86, p <.05

(4 g.I) = 13.57, p <.01

MCF = 32.88

MCF = 34.3

MCF = 24.3

Superior SES

BE/P = 2 times a day

AND = 2,5 years

n = 18

n = 9

n = 24

Ano PN (66)

Ano PN (35)

Ano PN (91)

Dep PN (61 / 5)

Dep PN (32 / 3)

Alc PN (80 / 11)

Alc PN (60 / 6)

Alc PN (30 / 5)

Dep PN (74 / 17)

Ago PN (52 / 14)

Ago PN (22 / 13)

Ago PN (64 / 27)

DrAd PN (35 / 31)

DrAd PN (16 / 19)

DrAd PN (51 / 40)

(4 g.I) = 13.54, p <.01

(4 g.I) = 10.84, p <.05

(4 g.I) = 15.57, p <.01

MCF = 20.4

MCF = 15.22

MCF = 24.2

Table 3. Divergent validity results in socio-economic status (SES), binge-eating/purge (BE/P) and anorexia nervosa

duration (AND).

ANOREXIA NERVOSA: DIVERGENT VALIDITY OF A PROTOTYPE NARRATIVE

103

TERAPIA PSICOLÓGICA 2006, Vol. 24, Nº1, 99–104

Finally, results of multiple regression analysis showed

that the duration of the clinical condition (anorexia nervosa)

was a significant predictor of the identification mean order

given to the anorexia prototype narrative (F(1,35) = 7.56,

p<.01). Post-hoc comparisons (Mann-Whitney) showed that

participants with anorexia nervosa for at least two and a

half years tended to identify most with the anorexia

prototype narrative (44.13>31; p<.01) (see Table 7).

Discussion

The pattern of results found in this study, revealed a

tendency for individuals diagnosed with anorexia nervosa

to identify themselves with the depression prototype

narrative. Participants only discriminated the anorexia

prototype narrative from the drug addiction prototype

narrative. There was also a tendency for highest

identification with the anorexia prototype narrative,

compared with the drug addiction and, less often, with

agoraphobia prototype narrative, in those from a highest

socio-economic status, or with most severe indices of

psychopathology. On the other hand, only the duration of

illness had a predictive power on the degree of identification

with the anorexic prototype narrative.

Altogether, the present study brings some evidence

against the existence of a prototype narrative of anorexia

nervosa. These data seems to be in accordance with the

most recent transdiagnostic approaches claiming that

multiple common processes are evident in different

psychological disorders (c. f. Fairburn et al., 2003; Fairburn

& Harrison, 2003; Russel, 2003) therefore invalidating the

possibility of specific prototypes for different disorders.

However, the results also point out that duration and

severity of the clinical condition increasead the possibility

of significantly discriminating the anorexic prototype.

TS < 6 months

n = 24

Dep PN (96 / -25)

Alc PN (73 / -2)

Ago PN (73 / -2)

Ano PN (71)

DrAd PN (51 / 20)

(4 g.I) = 16.96, p <.01

MCF = 23.59

Table 5. Divergent validity results according to therapy

sessions (TS)

Table 6. Divergent validity results in Interpersonal Sensitivity (IS) and Anxiety (A) of SCL 90-R subscales and Total (T)

and Interoceptive Awareness (IA) of EDI subscales.

IS of SCL 90-R

A of SCL 90-R

T of EDI

IA of EDI

below cut of point

n = 17

n = 25

n = 17

n = 21

Ago PN (66 / -20)

Ago PN (101 / -31)

Ago PN (70 / -26)

Ago PN (83 / -24)

Alc PN (56 / -10)

Dep PN (75 / -5)

Dep PN (55 / -11)

Alc PN (68 / -9)

Dep PN (53 / -7)

Alc PN (73 / -3)

Alc PN (52 / -8)

Dep PN (65 / -6)

Ano PN (46)

Ano PN (70)

Ano PN (44)

Ano PN (59)

DrAd PN (38 / 8)

DrAd PN (60 / 10)

DrAd PN (34 / 10)

DrAd PN (40 / 19)

(4 g.I) = 10.47, p <.05

(4 g.I) = 14.73, p <.01

(4 g.I) = 16.85, p <.01

(4 g.I) = 18.55, p <.01

MCF = 19.95

MCF = 24.23

MCF = 20.17

MCF = 22.42

Predictor

Final Model

AN Duration

.11

.33

2.75

F(1,35) = 7.56*

Table 7. Multiple Regression Analysis to predict the order of identification given to the anorexia prototype narrative.

*p < .01

104

BÁRBARA CÉSAR MACHADO / ÓSCAR F. GONÇALVES / PAULO P. P. MACHADO /

MARGARIDA R. HENRIQUES / ANTÓNIO ROMA-TORRES / ISABEL BRANDÃO

TERAPIA PSICOLÓGICA 2006, Vol. 24, Nº1, 99–104

Chronic anorexia nervosa is, usually, considered a bad

prognostic factor for the course and recovery of the disorder

(c. f. Fairburn & Harrison, 2003; Steinhausen, 1995;

Sullivan, 2002). This chronic condition may be associated

with the development of a more rigid narrative prototype

(Gonçalves et al., 2000).

If this is the case both, transdiagnostic and prototype

approaches may co-exist, being the less morbid condition

associated with transdiagnostic processes and more chronic

situations associated with narrative prototypes.

Acknowledgements

This study was partial funded by Fundação Ciência e

Tecnologia (FCT; POCTI/33252/99).

References

American Psychiatric Association (1996). Manual de Diagnóstico e

Estatística das Perturbações Mentais (4ªed.) (DSM-IV). Lisboa:

Climepsi Editores. (Publicação original 1994).

Cooper, Z. (1995). Eating disorders and their relationship to mood and

anxiety disorders. In: Brownell, K. D. & Fairburn, C. G. (Eds.). Eating

Disorders and Obesity (pp. 159-164). N.Y.: The Guilford Press.

Derrogatis, L. R. (1977). SCL – 90: Administration, scoring and procedures

manual – I for the r(evised) version. John Hopkins University School

of Medicine, Clinical Psychometrics Research Unit: Baltimore.

Fairburn, C. G. & Harrison, P. J. (2003). Eating disorders. The Lancet,

361 (1), 407-416.

Fairburn, C. G., Cooper, Z. & Shafran, R. (2003). Cognitive behaviour

therapy for eating disorders: a “transdiagnostic" theory and treatment.

Behaviour Research and Therapy, 41, 509-528.

Garner, D. M., Olmsted, M. P. & Polivy, J. (1983). Development and

validation of a multidimensional eating disorder inventory for anorexia

nervosa and bulimia nervosa. International Journal of Eating Disorders,

2, 15-34.

Gonçalves, O. F. & Henriques, M. R. (2000). Manual de avaliação do

conteúdo e multiplicidade narrativa. Braga: Universidade do Minho.

Gonçalves, O. F. & Machado, P. P. (1999). Cognitive narrative

psychotherapy: Research foudations. Journal of Clinical Psychology,

55, 1179-1192.

Gonçalves, O. F., Machado, P. P. Korman, Y., & Angus, L. (2001).

Assessing psychopathology: A narrative approach. In: Beutler, L. &

Malik, M. (eds.) Alternatives to the DSM—IV: Psychological

perspectives. Washington, DC: American Psychological Association.

Gonçalves, O. F., Korman, Y. & Angus, L. (2000). Constructing

psychopathology from a cognitive narrative perspective. In:. Neimeyer,

R. A & Raskin, J. D. (Eds.), Constructions of disorder: Meaning-making

frameworks for psychotherapy, pp. 265-284. Washington: American

Psychological Association.

Hermans, H. J. M. & Hermans-Jansen, E. (1995). Self-narratives: The

construction of meaning in psychotherapy. New-York: Guilford.

Leahy, R. L. (1991). Scripts in cognitive therapy: The systemic perspective.

Journal of Cognitive Psychotherapy, 5, 291-304.

Russel, G. F. M. (2003, October). Limitations in the classifications of

eating disorders: The gulf between the clinician and the researcher.

Paper presented at the annual meeting of the Eating Disorders Research

Society, Ravello, Italy.

Steinhausen, H. C. (1995). The course and outcome of anorexia nervosa.

In: Browell, K. D. & Fairburn, C. G. (Eds.). Eating disorders and obesity:

A comprehensive handbook, pp. 234-237. N. Y.: The Guilford Press.

Sullivan, P. F. (2002). Course and outcome of anorexia nervosa and buli-

mia nervosa. In: Fairburn, C. G. & Brownell, K. D. (Eds.). Eating

Disorders and Obesity: A Comprehensive Handbook, second edition

(pp. 226-232). N. Y.: The Guilford Press.