- Email: [email protected]
Perceived expressed emotion in anorexia nervosa, bulimia nervosa, and binge-eating disorder
Available online at www.sciencedirect.com
Comprehensive Psychiatry 51 (2010) 401 – 405 www.elsevier.com/locate/comppsych
Perceived expressed emotion in anorexia nervosa, bulimia nervosa, and binge-eating disorder Fiammetta Di Paolaa , Carlo Faravellia , Valdo Riccab,⁎ a Department of Psychology, University of Florence, Italy Department of Neurological and Psychiatric Sciences, University of Florence, Italy
b
Abstract The aim of this study was to verify the level of expressed emotion (EE) as perceived from patients with an eating disorder (ED). The Italian translation of the Level of Expressed Emotion Scale was administered to 63 female patients with Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, diagnosis of anorexia nervosa, bulimia nervosa, and binge-eating disorder and 63 control subjects, according to a case-control procedure. Patients with ED showed higher level of perceived EE than controls, whereas no significant differences were observed when comparing the 3 patient subgroups. The level of perceived EE was found to be independent of age, person who has been most influential in the patient's life, amount of contacts, and duration of treatment. Different associations between eating disorder psychopathology and EE were found, suggesting a close relationship between the emotional response and tolerance of influential person and the dysfunctional attitudes regarding eating, weight, and body shape. © 2010 Elsevier Inc. All rights reserved.
1. Introduction The concept of expressed emotion (EE) was developed in the 1960s by Rutter and Brown [1,2] to assess some aspects of family life associated with relapse in patients with schizophrenia. Later, EE has been shown to be a good predictor of relapse also for patients having other psychopathologic conditions, including Alzheimer disease, anxiety disorders, depression, substance abuse, and eating disorders [3-6]. Leff and Vaughn [4] reported that high EE is characterized by 4 attitudes or response styles: (1) high level of intrusiveness (ie, making repeated attempts to establish contact or to offer unsolicited and frequently critical advice), (2) highly emotional response to the patient's illness (ie, responding with anger, acute distress, reactions that tend to upset the patient), (3) negative attitude toward the patient's illness (ie, doubting that he/she has no control over symptoms; blaming or holding the patient
⁎ Corresponding author. Psychiatry Unit, Department of Neuropsychiatric Sciences, Florence University School of Medicine, 50134 Firenze, Italy. E-mail address: [email protected] (V. Ricca). 0010-440X/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.comppsych.2009.10.002
responsible for his/her condition), and (4) low level of tolerance and high expectations of the patient (ie, relatives are not convinced that the patient is really ill, they are intolerant of the patient's behaviors and social impairments). As far as eating disorders (EDs) are concerned, some studies described an interaction between EE and abnormal eating behaviors, and between EE and the development and maintenance of anorexia nervosa (AN) and bulimia nervosa (BN) [5,7]. On the other hand, binge-eating disorder (BED) has never been explored from this point of view. The Camberwell Family Interview (CFI) [8] was the first psychometric instrument devoted to the assessment of the family emotional climate and is considered the gold standard for it. However, CFI requires time for its training, administration, and coding and also requires the availability of a key relative. The Level of Expressed Emotion (LEE) Scale [9] examines the EE perceived from the patient's perspective and has been constructed from the conceptual framework described by Vaughn and Leff [3]. The present study is aimed at the evaluation of the level of EE in families with a member having AN, BN, and BED, considering the perceived EE from the patient's perspective. In addition, we investigated the possible effect on the LEE scores of the following factors: age, duration of treatment,
402
F. Di Paola et al. / Comprehensive Psychiatry 51 (2010) 401–405
Table 1 Patients (N = 63)—main clinical features Diagnosis
AN, n = 20 BN, n = 20 BED, n = 23
Age (y)
24.40 ± 7.03 25.25 ± 4.17 54.43 ± 11.28
Body mass index (kg/m2)
17.2 ± 2 25.1 ± 2 34.4 ± 2
Age at onset (y)
Duration of treatment (y)
15.75 ± 3.60 17.00 ± 1.60 34.52 ± 12.34
3.15 ± 6.57 1.51 ± 1.81 3.31 ± 3.54
Hospitalization, n (%) Without, n = 40 (63.5%)
With, n = 23 (36.5%)
8 (40%) 10 (50%) 22 (96%)
12 (60%) 10 (50%) 1 (4%)
Data are expressed as mean ± SD.
and amount of contact with the influential person as indicated by the patient.
ually and in reserved settings. Table 2 shows the main features of the control group. 2.2. Instruments
2. Methods 2.1. Participants A consecutive series of 63 patients with Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, diagnoses of EDs (20 AN, 20 BN, 23 BED) were studied. All the subjects were Italian white women with age ranging from 17 to 72 years old. The study was approved by the Human Subjects Review Committee of the Florence University, and all subjects participated with informed, voluntary, written consent. The patients were recruited at the Eating Disorders Clinic of the Department of Neurology and Psychiatry, University of Florence (Italy), and were diagnosed by means of the Structured Clinical Interview for DSM I [10]. Forty-two were single (n = 42; 66.67%), 17 were married (26.98%), 1 divorced (1.59%), and 3 were widowers (4.76%). Considering the occupation of the patients, 21 (33.33%) were students, 13 (20.63%) were housewives, 11 (17.46%) were office workers, 9 (14.29%) were workers, 6 (9.52%) were workers in services, and 3 (4.76%) were unemployed. Educational level was high school for 44 patients (69.84%) and secondary school for 19 patients (30.16%). Table 1 shows the major clinical features of the patients. Subjects completed, individually, the 60-item LEE Scale [13] and reported the most influential person in their life for the past 3 months and the amount of contacts per week with him or her. These 63 patients were also assessed by means of the Eating Disorder Examination Questionnaire (EDE-Q) [11], the Emotional Eating Scale (EES) [12], and the Binge Eating Scale (BES) [13]. These patients were compared with a sample of healthy controls drawn from general population. Of the 237 female subjects comprising the group used for the Italian validation of the LEE [17], 63 were randomly selected to match the patients sample for age and education, according to a case-control method. These subjects were originally recruited by opportunity sampling, on the basis of geographical area, sex, nationality, age, and employment status. Participants were not offered any incentives for participation. The questionnaire has been fulfilled individ-
2.2.1. Level of expressed emotion scale The Level of Expressed Emotion Scale is a 60-item selfreport rating the perceived EE from the patient's perspective. The item selection was based on the theory of Vaughn and Leff [14] that suggested 4 dimensions that could discriminate between high and low EE: intrusiveness, emotional response, attitude toward the illness, and tolerance/expectations concerning the patient. There are 15 true or false items for each component, generating a total score of 60. Higher scores indicate higher EE levels. Patient is classified as high EE when his or her score lies above the median [15]. The scale has good internal consistency, good test-retest reliability, and good temporal stability. Intrusiveness and tolerance/expectation subscale are significantly correlated with the critical comment scale of the CFI (r = 0.40) [16]. For the present study, we used the Italian version of the LEE Scale [17]. It has sound psychometric properties of construct validity, internal consistency, and reliability (the internal consistency reliability [KR-20] coefficient for the overall scale was 0.95; the subscales Table 2 Mean and SD EE scores in the 3 patient groups Age (y) Body mass index (kg/m2) Ethnicity Marital status Single Married Divorced Widowed Occupation Students Office workers Housewives Workers Workers in services Unemployed Educational level High school Secondary school a
Data are expressed as mean ± SD.
34.7 (± 15.6)a 24.8 (± 4.0)a White 69.84% (n = 44) 25.4% (n = 16) 3.17% (n = 2) 1.59% (n = 1) 36.51% (n = 23) 19.05% (n = 12) 20.63% (n = 13) 7.94% (n = 5) 11.11% (n = 7) 4.76% (n = 3) 69.84% (n = 44) 30.16% (n = 19)
F. Di Paola et al. / Comprehensive Psychiatry 51 (2010) 401–405 Table 3 Level of Expressed Emotion Scale t test between patients (N = 63) and an age/education-matched control sample (N = 63) Patients
Control group
Intrusiveness 6.43 ± 3.96 Emotional response 7.24 ± 4.49 Attitude toward illness 4.75 ± 4.02 Tolerance/expectations 6.57 ± 3.99 Total 24.98 ± 13.90
Levene test F
2.29 ± 1.96 38.35⁎ 1.67 ± 2.04 50.50⁎ 0.92 ± 1.27 94.19⁎ 1.78 ± 1.10 92.08⁎ 6.65 ± 3.65 103.74⁎
t test t⁎⁎ 7.44⁎ 8.97⁎ 7.21⁎ 9.18⁎ 10.13⁎
Data are expressed as mean ± SD. ⁎ P b .05. ⁎⁎ df = 124.
intrusiveness, emotional response, attitude toward illness and tolerance/expectations had these KR-20 values, respectively, 0.94, 0.81, 0.89, and 0.84); its sensitivity permits to clearly distinguish patients from healthy subjects. The patients also reported the number of waking hours they had spent per week with the person considered most influential in their life, during the previous 3 months. 2.2.2. Eating Disorder Examination-Questionnaire The Eating Disorder Examination-Questionnaire consists of 38 items, assessing the core ED psychopathologic features, and contains 4 subscales as follows: dietary restraint, eating concern, weight concern, and shape concern. The dietary restraint subscale is an admixture of cognitions and behaviors pertaining to dietary restriction. The 3 other subscales evaluate the dysfunctional attitudes regarding eating and overvalued thoughts regarding weight and shape. The global score represents the mean of the 4 subscale scores. Different studies showed that EDE-Q has an adequate test-retest reliability [11], good convergence with the EDE interview [18-22], and both discriminant and concurrent validity [23,24]. 2.2.3. Emotional eating scale Emotional eating was assessed by means of the EES, a 25-item self-report questionnaire. Each item consists of an emotion term (eg, jittery, angry, helpless), and the individual is asked to indicate the extent to which experiencing that emotion makes her or him likely to eat (no desire, small desire, moderate desire, strong urge, overwhelming urge) [25]. The 25 items form 3 subscales, reflecting eating in response to anger (anger/frustration),
403
anxiety (anxiety), and depressed mood (depression). These 3 subscales reflect the emotional antecedents of binge eating, and on each scale, higher scores reflect a greater tendency to eat in response to emotional state. The Emotional Eating Scale has demonstrated good internal consistency, construct validity, discriminant validity, and criterion-related validity [26]. 2.2.4. Binge eating scale To measure the severity of binge eating, the BES was applied [13]. The Binge Eating Scale has been proposed, with a threshold score of 17, as a rapid screening instrument for BED in obese patients, and it examines both behavioral signs (eating large amounts of food) and feeling or cognition during a binge episode (loss of control, guilt, fear of being unable to stop eating) through 16 items. 2.3. Statistical analysis Statistical analysis was performed using analysis of variance to determine the possible effect of others factors (diagnosis, duration of treatment, age, most influential person, number of hours contact with this person) on the LEE Scale scores. Statistical significance was tested using t test. Pearson correlations were conducted between the LEE Scale and the EDE-Q, the EES, and the BES. 3. Results The patients obtained significantly higher mean scores than controls in all the 4 subscales of the LEE (Table 3). Using the median obtained from the general population as cutoff, 56 patients (88.9%) of 63 had a high level of perceived expressed emotion. There were no significant differences in EE levels between the AN, BN, and BED groups (Table 4). Apart from the diagnoses, the LEE total and subscale scores were independent of age, duration of treatment, amount of contact (≤35 hours contact with the influential person vs ≥35 hours), and type of influential person. Similarly, the influence of the most important person, as indicated by the patient, on LEE Scale scores was no significant (F6,56 = 1.67; P b .05 for total EE). The EE scores about the influential person are shown in Table 5. It must be considered that some groups were too small to reach any statistical significance.
Table 4 Mean and SD EE scores in the 3 patient groups Diagnosis
Intrusiveness
Emotional response
Attitude toward illness
Tolerance/expectations
Total
AN BN BED Levene test (F)⁎ Analysis of variance (F)⁎
7.75 ± 3.68 6.20 ± 4.24 5.48 ± 3.79 0.79 1.86
8.05 ± 4.88 6.10 ± 4.18 7.52 ± 4.38 0.88 1.02
4.80 ± 4.47 3.60 ± 3.33 5.70 ± 4.06 1.92 1.48
7.15 ± 4.49 5.45 ± 3.38 7.04 ± 4.01 2.17 1.16
27.75 ± 15.04 21.35 ± 12.12 25.74 ± 14.23 1.57 1.12
Data are expressed as mean ± SD. ⁎ df1 = 2; df2 = 60.
404
F. Di Paola et al. / Comprehensive Psychiatry 51 (2010) 401–405
Table 5 Mean, SD, and analysis of variance of EE scores considering the influential person Influential person
Intrusiveness
Emotional response
Attitude toward illness
Tolerance/expectations
Total
Sister (n = 3) Father (n = 11) Mother (n = 19) Son (n = 2) Boyfriend (n = 5) Spouse (n = 15) Friend (n = 8) Total (N = 63) Levene test (F)⁎ Analysis of variance (F)⁎
8.33 ± 3.51 6.45 ± 4.32 8.16 ± 2.85 2.50 ± 2.12 2.60 ± 3.71 5.67 ± 3.79 6.38 ± 4.96 6.43 ± 3.96 1.18 2.13
5.67 ± 3.51 8.73 ± 4.96 7.26 ± 4.48 2.00 ± 1.41 5.40 ± 4.10 9.27 ± 3.81 4.38 ± 4.00 7.24 ± 4.49 .91 2.12
5.67 ± 6.66 5.73 ± 4.34 4.53 ± 3.92 1.50 ± 0.71 1.40 ± 1.95 6.13 ± 3.60 3.88 ± 4.19 4.75 ± 4.02 1.60 1.35
8.33 ± 5.13 7.27 ± 4.86 6.63 ± 3.70 3.00 ± 1.41 4.20 ± 2.39 7.80 ± 3.30 4.88 ± 4.82 6.57 ± 3.99 1.65 1.22
28.00 ± 14.00 28.18 ± 15.89 26.58 ± 12.46 9.00 ± 4.24 13.60 ± 11.26 28.87 ± 12.06 19.50 ± 16.41 24.98 ± 13.90 .83 1.67
Data are expressed as mean ± SD. ⁎ df1 = 6; df2 = 56.
3.1. Relationships between LEE Scale and measures of eating psychopathology There were several significant correlations between the LEE total and subscale scores and the EDE-Q total and subscale scores. In particular, the total EDE-Q score showed significant correlations with intrusiveness (r = 0.53), emotional response (r = 0.53), attitude toward illness (r = 0.60), tolerance/expectations (r = 0.57), and total EE (r = 0.65) scores. Intrusiveness subscale on the LEE significantly correlated with weight concern (r = 0.55) and eating concern (r = 0.71); emotional response correlated with weight concern (r = 0.53); attitude toward illness was positively and significantly correlated with weight concern (r = 0.57); and tolerance/expectations ratings correlated with eating concern (r = 0.57) and weight concern (r = 0.54). The total EE scores were correlated with weight concern (r = 0.64), eating concern (r = 0.64), and shape concern (r = 0.55). Considering the EES, the item “guilt” significantly correlated with the LEE subscales emotional response (r = 0.57), attitude toward illness (r = 0.49), and tolerance/ expectation (r = 0.62), whereas the item “excited” significantly correlated with the emotional response subscale (r = 0.54).
4. Discussion The findings of this preliminary research are of some interest because they represent the first study exploring the perceived expressed emotion in patients with AN and BN and evaluating the expressed emotion construct in patients with BED. The results of this study indicate that the expressed emotion is a factor that is strongly present in eating disorders, supporting some specific clinical features of ED patient's family, such as intrusiveness and criticism [27]. It must be stressed, however, that the method we used explores the subjectively perceived family climate rather than the actual expressed emotion. It is conceivable that, like in other kinds of self-assessment, the present psychopathology influences
the evaluation. The causal link expressed emotion → pathology could be at least partially reverted. All the interpretations of the present results must be kept under this frame, which is, however, common in a variety of studies in psychopathology. In this specific regard, however, it must be stressed that the relationship between illness and expressed emotion is likely to be bidirectional anyway, given that the patients' symptoms and behaviors are expected to elicit higher expressed emotion [28,29]. Our results seem to point out expressed emotion as a common factor in all the patients having ED, as patients with AN, BN, and BED show similar levels of EE, all well above the figures reported by the subjects drawn by the general population. As far as the key relative is concerned, different studies reported that mothers are significantly more involved than fathers [30-33], whereas we did not find significant differences between parents. Moreover, we did not find significant differences between the expressed emotion levels of potentially influential persons (as parents, boyfriend, or friend). Previous researches also reported that the relapse risk is higher when the amount of contact with an influential person is 35 hours or more per week [3,34], whereas in our study this criterion did not differentiate between patients with low or high expressed emotion level. The differences in the methods applied could account for these discrepancies. Finally, the correlations between the expressed emotions and the scales that explore the current psychopathology (LEE and BES, EES, EDE-Q) suggest that there is a relationship between the emotional response and tolerance of influential person on one hand and the dysfunctional attitudes toward food, weight, and body shape on the other. Expressed emotion also seems to correlate with guilt, a specific feeling relating to eating behavior as valuated by EES. In our study, no factor (diagnosis, age, onset, duration of treatment, influential person, or number of hours contact with this person) was found to affect EE. As the other studies used more sophisticated techniques to assess EE (CFI), it is conceivable that the subjective perception of the family expressed emotion as with the LEE reduces the sensitivity of the instrument.
F. Di Paola et al. / Comprehensive Psychiatry 51 (2010) 401–405
Considering the high levels of perceived expressed emotion observed in patients with BED, it is of note that different studies focused on the role that emotional states can play in the onset and maintenance of binge-eating symptoms [35,36]. In particular, some studies evaluated the possible role of emotional eating in subjects having BED, suggesting that episodes of binge eating are often precipitated by stress and negative affect [12,37] and that binge eating appears to be associated with a subsequent decrease in negative affect [38]. Our results indicate that patients with perceive high levels of expressed emotion in their families, suggest that such levels could favor the onset and maintenance of the binge behaviors, and can stress a possible relational domain of therapeutic intervention. As already pointed out, the perception of expressed emotion is not the same as the evaluation of the actual family expressed emotion. In this respect, our study could be seen as an exploration of the patient's subjective point of view on the family communication and atmosphere. On the other hand, as previously reported, the patient's behavior is expected to influence the family climate in any case. Understanding the relationships of perceived expressed emotion to other psychopathology within these patients may be useful for family therapy techniques. References [1] Brown GW, Rutter ML. The measurement of family activities and relationships. Hum Relat 1966;19:241. [2] Rutter ML, Brown GW. The reliability and validity of measures of family life and relationships in families containing a psychiatric patient. Soc Psychiatry 1966;1:38. [3] Vaughn CE, Leff JP. The influence of family and social factors in the course of psychiatric illness: a comparison of schizophrenic and depressed neurotic patients. Br J Psychiatry 1976;129:125-37. [4] Leff JP, Vaughn C. Expressed emotion in families. London: Guilford; 1985. [5] Butzlaff RL, Hooley JM. Expressed emotion and psychiatric relapse: a meta-analysis. Arch Gen Psychiatry 1998;55:547-52. [6] Hooley JM, Parker HA. Measuring expressed emotion: an evaluation of the shortcuts. J Fam Psychol 2006;20:386-96. [7] Bebbington P, Kuipers L. The predictive utility of expressed emotion in schizophrenia: an aggregate analysis. Psychol Med 1994;24:707-18. [8] Vaughn CE, Leff JP. The measurement of expressed emotion in the families of psychiatric patients. Br J Soc Clin Psychol 1976;15: 157-65. [9] Cole JD, Kazarian SS. The Level of Expressed Emotion Scale: a new measure of expressed emotion. J Clin Psychol 1988;44:392-7. [10] First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I), clinician version. Washington, DC: American Psychiatric Publishing Inc; 1995. [11] Luce KH, Crowther JH. The reliability of the Eating Disorder Examination-Self-Report Questionnaire Version (EDE-Q). Int J Eat Disord 1999;25:349-51. [12] Masheb RM, Grilo CM. Emotional overeating and its associations with eating disorder psychopathology among overweight patients with binge-eating disorder. Int J Eat Disord 2006;39:141-6. [13] Gormally J, Block S, Daston S, Rardin D. The assessment of binge eating severity among obese persons. Addict Behav 1982;7:47-55. [14] Vaughn CE, Leff JP. Patterns of emotional response in relatives of schizophrenic patients. Schzophrenia Bull 1981;7:43-4.
405
[15] Cole JD, Kazarian SS. Predictive validity of the Level of Expressed Emotion (LEE) Scale: readmission follow-up data for 1, 2, and 5-year periods. J Clin Psychol 1993;49:216-8. [16] Kazarian SS, Malla AK, Baker B, Cole JD. Comparisons of two expressed emotion scales with the Camberwell Family Interview. J Clin Psychol 1990;46:306-9. [17] Di Paola F, Faravelli C, Ricca V. The Level of Expressed Emotion Scale: validazione Italiana preliminare. Ital J Psychopathol 2008;14: 258-68. [18] Fairburn CG, Beglin SJ. Assessment of eating disorders: interview or self-report questionnaire? Int J Eat Disord 1994;16:363-70. [19] Black CMD, Wilson GT. Assessment of eating disorders: interview versus questionnaire. Int J Eat Disord 1996;20:43-50. [20] Wilfley DE, Schwartz MB, Spurrell EB, Fairburn CG. Assessing the specific psychopathology of binge-eating disorder patients: interview or self-report. Behav Res Ther 1997;35:1151-9. [21] Kalarchian MA, Wilson GT, Brolin RE, Bradley L. Assessment of eating disorders in bariatric surgery candidates: self-report questionnaire versus interview. Int J Eat Disord 2000;28:465-9. [22] Mond JM, Hay PJ, Rodgers B, Owen C, Beumont PJ. Temporal stability of the Eating Disorder Examination Questionnaire. Int J Eat Disord 2004;36:195-203. [23] Fairburn CG, Cooper Z. The Eating Disorder Examination. In: Fairburn CG, Wilson GT, editors. Binge eating: nature, assessment and treatment, 12th ed.New York: Guilford Press; 1993. p. 317-60. [24] Williamson DA, Anderson DA, Jackman LP, Jackson SR. Assessment of eating disordered thoughts, feelings, and behaviors. In: Allison DB, editor. Handbook of assessment methods for eating behaviours, and weight related problems: measures, theory and research. Thousand Oaks (Calif): Sage; 1995. p. 347-86. [25] Arnow B, Kenardy J, Agras WS. Binge eating among the obese: a descriptive study. J Behav Med 1992 Apr;15:155-70. [26] Arnow B, Kenardy J, Agras WS. The Emotional Eating Scale: the development of a measure to assess coping with negative affect by eating. Int J Eat Disord 1995;18:79-90. [27] Doane JA, Diamond D. Affect and Attachment in the Family. New York: Basic Books; 1994. [28] Shisslak CM, Crago M, Estes LS. The spectrum of eating disturbances. Int J Eat Disord 1995;18:209-19 Review. [29] Fairburn CG, Bohn K. Eating disorder NOS (EDNOS): an example of the troublesome ‘‘not otherwise specified” (NOS) category in DSM-IV. Behav Res Ther 2004;43:691-701. [30] Dare C, Le Grange D, Eisler I, Rutherford J. Redefining the psychosomatic family: family process of 26 eating disorder families. Int J Eat Disord 1994;16:211-26. [31] Besharat A, Eisler I. Family expressed emotion in eating disorders patients. Eu Psychiatr 1996;11:362s-3s. [32] Szmukler GI, Eisler I, Russell GFM, Dare C. Anorexia nervosa, parental expressed emotion, and dropping out of treatment. Br J Psychiatry 1985;147:265-71. [33] Van Furth EF, Van Strien DC, Martina LML, Van Son MJM, Hendrickx JJ, Van Engeland H. Expressed emotion and the prediction of outcome in adolescent eating disorders. Int J Eat Disord 1996;20:19-31. [34] Brown GW, Birley JLT, Wing JK. Influence of family life on the course of schizophrenic disorders: a replication. Br J Psychiatry 1972;121:241-58. [35] Engelberg MJ, Steiger H, Gauvin L, Wonderlich SA. Binge antecedents in bulimic syndromes: an examination of dissociation and negative affect. Int J Eat Disord 2007;40:531-6. [36] Stein RI, Kenardy J, Wiseman CV, Dounchis JZ, Arnow BA, Wilfley DE. What's driving the binge in binge-eating disorder?: a prospective examination of precursors and consequences. Int J Eat Disord 2007;40:195-203. [37] Telch CF, Agras WS. Do emotional states influence binge eating in the obese. Int J Eat Disord 1996;20:271-9. [38] Deaver CM, Miltenberger RG, Smyth J, Meidinger A, Crosby R. An evaluation of affect and binge eating. Behav Modif 2003;27: 578-99.
Comprehensive Psychiatry 51 (2010) 401 – 405 www.elsevier.com/locate/comppsych
Perceived expressed emotion in anorexia nervosa, bulimia nervosa, and binge-eating disorder Fiammetta Di Paolaa , Carlo Faravellia , Valdo Riccab,⁎ a Department of Psychology, University of Florence, Italy Department of Neurological and Psychiatric Sciences, University of Florence, Italy
b
Abstract The aim of this study was to verify the level of expressed emotion (EE) as perceived from patients with an eating disorder (ED). The Italian translation of the Level of Expressed Emotion Scale was administered to 63 female patients with Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, diagnosis of anorexia nervosa, bulimia nervosa, and binge-eating disorder and 63 control subjects, according to a case-control procedure. Patients with ED showed higher level of perceived EE than controls, whereas no significant differences were observed when comparing the 3 patient subgroups. The level of perceived EE was found to be independent of age, person who has been most influential in the patient's life, amount of contacts, and duration of treatment. Different associations between eating disorder psychopathology and EE were found, suggesting a close relationship between the emotional response and tolerance of influential person and the dysfunctional attitudes regarding eating, weight, and body shape. © 2010 Elsevier Inc. All rights reserved.
1. Introduction The concept of expressed emotion (EE) was developed in the 1960s by Rutter and Brown [1,2] to assess some aspects of family life associated with relapse in patients with schizophrenia. Later, EE has been shown to be a good predictor of relapse also for patients having other psychopathologic conditions, including Alzheimer disease, anxiety disorders, depression, substance abuse, and eating disorders [3-6]. Leff and Vaughn [4] reported that high EE is characterized by 4 attitudes or response styles: (1) high level of intrusiveness (ie, making repeated attempts to establish contact or to offer unsolicited and frequently critical advice), (2) highly emotional response to the patient's illness (ie, responding with anger, acute distress, reactions that tend to upset the patient), (3) negative attitude toward the patient's illness (ie, doubting that he/she has no control over symptoms; blaming or holding the patient
⁎ Corresponding author. Psychiatry Unit, Department of Neuropsychiatric Sciences, Florence University School of Medicine, 50134 Firenze, Italy. E-mail address: [email protected] (V. Ricca). 0010-440X/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.comppsych.2009.10.002
responsible for his/her condition), and (4) low level of tolerance and high expectations of the patient (ie, relatives are not convinced that the patient is really ill, they are intolerant of the patient's behaviors and social impairments). As far as eating disorders (EDs) are concerned, some studies described an interaction between EE and abnormal eating behaviors, and between EE and the development and maintenance of anorexia nervosa (AN) and bulimia nervosa (BN) [5,7]. On the other hand, binge-eating disorder (BED) has never been explored from this point of view. The Camberwell Family Interview (CFI) [8] was the first psychometric instrument devoted to the assessment of the family emotional climate and is considered the gold standard for it. However, CFI requires time for its training, administration, and coding and also requires the availability of a key relative. The Level of Expressed Emotion (LEE) Scale [9] examines the EE perceived from the patient's perspective and has been constructed from the conceptual framework described by Vaughn and Leff [3]. The present study is aimed at the evaluation of the level of EE in families with a member having AN, BN, and BED, considering the perceived EE from the patient's perspective. In addition, we investigated the possible effect on the LEE scores of the following factors: age, duration of treatment,
402
F. Di Paola et al. / Comprehensive Psychiatry 51 (2010) 401–405
Table 1 Patients (N = 63)—main clinical features Diagnosis
AN, n = 20 BN, n = 20 BED, n = 23
Age (y)
24.40 ± 7.03 25.25 ± 4.17 54.43 ± 11.28
Body mass index (kg/m2)
17.2 ± 2 25.1 ± 2 34.4 ± 2
Age at onset (y)
Duration of treatment (y)
15.75 ± 3.60 17.00 ± 1.60 34.52 ± 12.34
3.15 ± 6.57 1.51 ± 1.81 3.31 ± 3.54
Hospitalization, n (%) Without, n = 40 (63.5%)
With, n = 23 (36.5%)
8 (40%) 10 (50%) 22 (96%)
12 (60%) 10 (50%) 1 (4%)
Data are expressed as mean ± SD.
and amount of contact with the influential person as indicated by the patient.
ually and in reserved settings. Table 2 shows the main features of the control group. 2.2. Instruments
2. Methods 2.1. Participants A consecutive series of 63 patients with Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, diagnoses of EDs (20 AN, 20 BN, 23 BED) were studied. All the subjects were Italian white women with age ranging from 17 to 72 years old. The study was approved by the Human Subjects Review Committee of the Florence University, and all subjects participated with informed, voluntary, written consent. The patients were recruited at the Eating Disorders Clinic of the Department of Neurology and Psychiatry, University of Florence (Italy), and were diagnosed by means of the Structured Clinical Interview for DSM I [10]. Forty-two were single (n = 42; 66.67%), 17 were married (26.98%), 1 divorced (1.59%), and 3 were widowers (4.76%). Considering the occupation of the patients, 21 (33.33%) were students, 13 (20.63%) were housewives, 11 (17.46%) were office workers, 9 (14.29%) were workers, 6 (9.52%) were workers in services, and 3 (4.76%) were unemployed. Educational level was high school for 44 patients (69.84%) and secondary school for 19 patients (30.16%). Table 1 shows the major clinical features of the patients. Subjects completed, individually, the 60-item LEE Scale [13] and reported the most influential person in their life for the past 3 months and the amount of contacts per week with him or her. These 63 patients were also assessed by means of the Eating Disorder Examination Questionnaire (EDE-Q) [11], the Emotional Eating Scale (EES) [12], and the Binge Eating Scale (BES) [13]. These patients were compared with a sample of healthy controls drawn from general population. Of the 237 female subjects comprising the group used for the Italian validation of the LEE [17], 63 were randomly selected to match the patients sample for age and education, according to a case-control method. These subjects were originally recruited by opportunity sampling, on the basis of geographical area, sex, nationality, age, and employment status. Participants were not offered any incentives for participation. The questionnaire has been fulfilled individ-
2.2.1. Level of expressed emotion scale The Level of Expressed Emotion Scale is a 60-item selfreport rating the perceived EE from the patient's perspective. The item selection was based on the theory of Vaughn and Leff [14] that suggested 4 dimensions that could discriminate between high and low EE: intrusiveness, emotional response, attitude toward the illness, and tolerance/expectations concerning the patient. There are 15 true or false items for each component, generating a total score of 60. Higher scores indicate higher EE levels. Patient is classified as high EE when his or her score lies above the median [15]. The scale has good internal consistency, good test-retest reliability, and good temporal stability. Intrusiveness and tolerance/expectation subscale are significantly correlated with the critical comment scale of the CFI (r = 0.40) [16]. For the present study, we used the Italian version of the LEE Scale [17]. It has sound psychometric properties of construct validity, internal consistency, and reliability (the internal consistency reliability [KR-20] coefficient for the overall scale was 0.95; the subscales Table 2 Mean and SD EE scores in the 3 patient groups Age (y) Body mass index (kg/m2) Ethnicity Marital status Single Married Divorced Widowed Occupation Students Office workers Housewives Workers Workers in services Unemployed Educational level High school Secondary school a
Data are expressed as mean ± SD.
34.7 (± 15.6)a 24.8 (± 4.0)a White 69.84% (n = 44) 25.4% (n = 16) 3.17% (n = 2) 1.59% (n = 1) 36.51% (n = 23) 19.05% (n = 12) 20.63% (n = 13) 7.94% (n = 5) 11.11% (n = 7) 4.76% (n = 3) 69.84% (n = 44) 30.16% (n = 19)
F. Di Paola et al. / Comprehensive Psychiatry 51 (2010) 401–405 Table 3 Level of Expressed Emotion Scale t test between patients (N = 63) and an age/education-matched control sample (N = 63) Patients
Control group
Intrusiveness 6.43 ± 3.96 Emotional response 7.24 ± 4.49 Attitude toward illness 4.75 ± 4.02 Tolerance/expectations 6.57 ± 3.99 Total 24.98 ± 13.90
Levene test F
2.29 ± 1.96 38.35⁎ 1.67 ± 2.04 50.50⁎ 0.92 ± 1.27 94.19⁎ 1.78 ± 1.10 92.08⁎ 6.65 ± 3.65 103.74⁎
t test t⁎⁎ 7.44⁎ 8.97⁎ 7.21⁎ 9.18⁎ 10.13⁎
Data are expressed as mean ± SD. ⁎ P b .05. ⁎⁎ df = 124.
intrusiveness, emotional response, attitude toward illness and tolerance/expectations had these KR-20 values, respectively, 0.94, 0.81, 0.89, and 0.84); its sensitivity permits to clearly distinguish patients from healthy subjects. The patients also reported the number of waking hours they had spent per week with the person considered most influential in their life, during the previous 3 months. 2.2.2. Eating Disorder Examination-Questionnaire The Eating Disorder Examination-Questionnaire consists of 38 items, assessing the core ED psychopathologic features, and contains 4 subscales as follows: dietary restraint, eating concern, weight concern, and shape concern. The dietary restraint subscale is an admixture of cognitions and behaviors pertaining to dietary restriction. The 3 other subscales evaluate the dysfunctional attitudes regarding eating and overvalued thoughts regarding weight and shape. The global score represents the mean of the 4 subscale scores. Different studies showed that EDE-Q has an adequate test-retest reliability [11], good convergence with the EDE interview [18-22], and both discriminant and concurrent validity [23,24]. 2.2.3. Emotional eating scale Emotional eating was assessed by means of the EES, a 25-item self-report questionnaire. Each item consists of an emotion term (eg, jittery, angry, helpless), and the individual is asked to indicate the extent to which experiencing that emotion makes her or him likely to eat (no desire, small desire, moderate desire, strong urge, overwhelming urge) [25]. The 25 items form 3 subscales, reflecting eating in response to anger (anger/frustration),
403
anxiety (anxiety), and depressed mood (depression). These 3 subscales reflect the emotional antecedents of binge eating, and on each scale, higher scores reflect a greater tendency to eat in response to emotional state. The Emotional Eating Scale has demonstrated good internal consistency, construct validity, discriminant validity, and criterion-related validity [26]. 2.2.4. Binge eating scale To measure the severity of binge eating, the BES was applied [13]. The Binge Eating Scale has been proposed, with a threshold score of 17, as a rapid screening instrument for BED in obese patients, and it examines both behavioral signs (eating large amounts of food) and feeling or cognition during a binge episode (loss of control, guilt, fear of being unable to stop eating) through 16 items. 2.3. Statistical analysis Statistical analysis was performed using analysis of variance to determine the possible effect of others factors (diagnosis, duration of treatment, age, most influential person, number of hours contact with this person) on the LEE Scale scores. Statistical significance was tested using t test. Pearson correlations were conducted between the LEE Scale and the EDE-Q, the EES, and the BES. 3. Results The patients obtained significantly higher mean scores than controls in all the 4 subscales of the LEE (Table 3). Using the median obtained from the general population as cutoff, 56 patients (88.9%) of 63 had a high level of perceived expressed emotion. There were no significant differences in EE levels between the AN, BN, and BED groups (Table 4). Apart from the diagnoses, the LEE total and subscale scores were independent of age, duration of treatment, amount of contact (≤35 hours contact with the influential person vs ≥35 hours), and type of influential person. Similarly, the influence of the most important person, as indicated by the patient, on LEE Scale scores was no significant (F6,56 = 1.67; P b .05 for total EE). The EE scores about the influential person are shown in Table 5. It must be considered that some groups were too small to reach any statistical significance.
Table 4 Mean and SD EE scores in the 3 patient groups Diagnosis
Intrusiveness
Emotional response
Attitude toward illness
Tolerance/expectations
Total
AN BN BED Levene test (F)⁎ Analysis of variance (F)⁎
7.75 ± 3.68 6.20 ± 4.24 5.48 ± 3.79 0.79 1.86
8.05 ± 4.88 6.10 ± 4.18 7.52 ± 4.38 0.88 1.02
4.80 ± 4.47 3.60 ± 3.33 5.70 ± 4.06 1.92 1.48
7.15 ± 4.49 5.45 ± 3.38 7.04 ± 4.01 2.17 1.16
27.75 ± 15.04 21.35 ± 12.12 25.74 ± 14.23 1.57 1.12
Data are expressed as mean ± SD. ⁎ df1 = 2; df2 = 60.
404
F. Di Paola et al. / Comprehensive Psychiatry 51 (2010) 401–405
Table 5 Mean, SD, and analysis of variance of EE scores considering the influential person Influential person
Intrusiveness
Emotional response
Attitude toward illness
Tolerance/expectations
Total
Sister (n = 3) Father (n = 11) Mother (n = 19) Son (n = 2) Boyfriend (n = 5) Spouse (n = 15) Friend (n = 8) Total (N = 63) Levene test (F)⁎ Analysis of variance (F)⁎
8.33 ± 3.51 6.45 ± 4.32 8.16 ± 2.85 2.50 ± 2.12 2.60 ± 3.71 5.67 ± 3.79 6.38 ± 4.96 6.43 ± 3.96 1.18 2.13
5.67 ± 3.51 8.73 ± 4.96 7.26 ± 4.48 2.00 ± 1.41 5.40 ± 4.10 9.27 ± 3.81 4.38 ± 4.00 7.24 ± 4.49 .91 2.12
5.67 ± 6.66 5.73 ± 4.34 4.53 ± 3.92 1.50 ± 0.71 1.40 ± 1.95 6.13 ± 3.60 3.88 ± 4.19 4.75 ± 4.02 1.60 1.35
8.33 ± 5.13 7.27 ± 4.86 6.63 ± 3.70 3.00 ± 1.41 4.20 ± 2.39 7.80 ± 3.30 4.88 ± 4.82 6.57 ± 3.99 1.65 1.22
28.00 ± 14.00 28.18 ± 15.89 26.58 ± 12.46 9.00 ± 4.24 13.60 ± 11.26 28.87 ± 12.06 19.50 ± 16.41 24.98 ± 13.90 .83 1.67
Data are expressed as mean ± SD. ⁎ df1 = 6; df2 = 56.
3.1. Relationships between LEE Scale and measures of eating psychopathology There were several significant correlations between the LEE total and subscale scores and the EDE-Q total and subscale scores. In particular, the total EDE-Q score showed significant correlations with intrusiveness (r = 0.53), emotional response (r = 0.53), attitude toward illness (r = 0.60), tolerance/expectations (r = 0.57), and total EE (r = 0.65) scores. Intrusiveness subscale on the LEE significantly correlated with weight concern (r = 0.55) and eating concern (r = 0.71); emotional response correlated with weight concern (r = 0.53); attitude toward illness was positively and significantly correlated with weight concern (r = 0.57); and tolerance/expectations ratings correlated with eating concern (r = 0.57) and weight concern (r = 0.54). The total EE scores were correlated with weight concern (r = 0.64), eating concern (r = 0.64), and shape concern (r = 0.55). Considering the EES, the item “guilt” significantly correlated with the LEE subscales emotional response (r = 0.57), attitude toward illness (r = 0.49), and tolerance/ expectation (r = 0.62), whereas the item “excited” significantly correlated with the emotional response subscale (r = 0.54).
4. Discussion The findings of this preliminary research are of some interest because they represent the first study exploring the perceived expressed emotion in patients with AN and BN and evaluating the expressed emotion construct in patients with BED. The results of this study indicate that the expressed emotion is a factor that is strongly present in eating disorders, supporting some specific clinical features of ED patient's family, such as intrusiveness and criticism [27]. It must be stressed, however, that the method we used explores the subjectively perceived family climate rather than the actual expressed emotion. It is conceivable that, like in other kinds of self-assessment, the present psychopathology influences
the evaluation. The causal link expressed emotion → pathology could be at least partially reverted. All the interpretations of the present results must be kept under this frame, which is, however, common in a variety of studies in psychopathology. In this specific regard, however, it must be stressed that the relationship between illness and expressed emotion is likely to be bidirectional anyway, given that the patients' symptoms and behaviors are expected to elicit higher expressed emotion [28,29]. Our results seem to point out expressed emotion as a common factor in all the patients having ED, as patients with AN, BN, and BED show similar levels of EE, all well above the figures reported by the subjects drawn by the general population. As far as the key relative is concerned, different studies reported that mothers are significantly more involved than fathers [30-33], whereas we did not find significant differences between parents. Moreover, we did not find significant differences between the expressed emotion levels of potentially influential persons (as parents, boyfriend, or friend). Previous researches also reported that the relapse risk is higher when the amount of contact with an influential person is 35 hours or more per week [3,34], whereas in our study this criterion did not differentiate between patients with low or high expressed emotion level. The differences in the methods applied could account for these discrepancies. Finally, the correlations between the expressed emotions and the scales that explore the current psychopathology (LEE and BES, EES, EDE-Q) suggest that there is a relationship between the emotional response and tolerance of influential person on one hand and the dysfunctional attitudes toward food, weight, and body shape on the other. Expressed emotion also seems to correlate with guilt, a specific feeling relating to eating behavior as valuated by EES. In our study, no factor (diagnosis, age, onset, duration of treatment, influential person, or number of hours contact with this person) was found to affect EE. As the other studies used more sophisticated techniques to assess EE (CFI), it is conceivable that the subjective perception of the family expressed emotion as with the LEE reduces the sensitivity of the instrument.
F. Di Paola et al. / Comprehensive Psychiatry 51 (2010) 401–405
Considering the high levels of perceived expressed emotion observed in patients with BED, it is of note that different studies focused on the role that emotional states can play in the onset and maintenance of binge-eating symptoms [35,36]. In particular, some studies evaluated the possible role of emotional eating in subjects having BED, suggesting that episodes of binge eating are often precipitated by stress and negative affect [12,37] and that binge eating appears to be associated with a subsequent decrease in negative affect [38]. Our results indicate that patients with perceive high levels of expressed emotion in their families, suggest that such levels could favor the onset and maintenance of the binge behaviors, and can stress a possible relational domain of therapeutic intervention. As already pointed out, the perception of expressed emotion is not the same as the evaluation of the actual family expressed emotion. In this respect, our study could be seen as an exploration of the patient's subjective point of view on the family communication and atmosphere. On the other hand, as previously reported, the patient's behavior is expected to influence the family climate in any case. Understanding the relationships of perceived expressed emotion to other psychopathology within these patients may be useful for family therapy techniques. References [1] Brown GW, Rutter ML. The measurement of family activities and relationships. Hum Relat 1966;19:241. [2] Rutter ML, Brown GW. The reliability and validity of measures of family life and relationships in families containing a psychiatric patient. Soc Psychiatry 1966;1:38. [3] Vaughn CE, Leff JP. The influence of family and social factors in the course of psychiatric illness: a comparison of schizophrenic and depressed neurotic patients. Br J Psychiatry 1976;129:125-37. [4] Leff JP, Vaughn C. Expressed emotion in families. London: Guilford; 1985. [5] Butzlaff RL, Hooley JM. Expressed emotion and psychiatric relapse: a meta-analysis. Arch Gen Psychiatry 1998;55:547-52. [6] Hooley JM, Parker HA. Measuring expressed emotion: an evaluation of the shortcuts. J Fam Psychol 2006;20:386-96. [7] Bebbington P, Kuipers L. The predictive utility of expressed emotion in schizophrenia: an aggregate analysis. Psychol Med 1994;24:707-18. [8] Vaughn CE, Leff JP. The measurement of expressed emotion in the families of psychiatric patients. Br J Soc Clin Psychol 1976;15: 157-65. [9] Cole JD, Kazarian SS. The Level of Expressed Emotion Scale: a new measure of expressed emotion. J Clin Psychol 1988;44:392-7. [10] First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I), clinician version. Washington, DC: American Psychiatric Publishing Inc; 1995. [11] Luce KH, Crowther JH. The reliability of the Eating Disorder Examination-Self-Report Questionnaire Version (EDE-Q). Int J Eat Disord 1999;25:349-51. [12] Masheb RM, Grilo CM. Emotional overeating and its associations with eating disorder psychopathology among overweight patients with binge-eating disorder. Int J Eat Disord 2006;39:141-6. [13] Gormally J, Block S, Daston S, Rardin D. The assessment of binge eating severity among obese persons. Addict Behav 1982;7:47-55. [14] Vaughn CE, Leff JP. Patterns of emotional response in relatives of schizophrenic patients. Schzophrenia Bull 1981;7:43-4.
405
[15] Cole JD, Kazarian SS. Predictive validity of the Level of Expressed Emotion (LEE) Scale: readmission follow-up data for 1, 2, and 5-year periods. J Clin Psychol 1993;49:216-8. [16] Kazarian SS, Malla AK, Baker B, Cole JD. Comparisons of two expressed emotion scales with the Camberwell Family Interview. J Clin Psychol 1990;46:306-9. [17] Di Paola F, Faravelli C, Ricca V. The Level of Expressed Emotion Scale: validazione Italiana preliminare. Ital J Psychopathol 2008;14: 258-68. [18] Fairburn CG, Beglin SJ. Assessment of eating disorders: interview or self-report questionnaire? Int J Eat Disord 1994;16:363-70. [19] Black CMD, Wilson GT. Assessment of eating disorders: interview versus questionnaire. Int J Eat Disord 1996;20:43-50. [20] Wilfley DE, Schwartz MB, Spurrell EB, Fairburn CG. Assessing the specific psychopathology of binge-eating disorder patients: interview or self-report. Behav Res Ther 1997;35:1151-9. [21] Kalarchian MA, Wilson GT, Brolin RE, Bradley L. Assessment of eating disorders in bariatric surgery candidates: self-report questionnaire versus interview. Int J Eat Disord 2000;28:465-9. [22] Mond JM, Hay PJ, Rodgers B, Owen C, Beumont PJ. Temporal stability of the Eating Disorder Examination Questionnaire. Int J Eat Disord 2004;36:195-203. [23] Fairburn CG, Cooper Z. The Eating Disorder Examination. In: Fairburn CG, Wilson GT, editors. Binge eating: nature, assessment and treatment, 12th ed.New York: Guilford Press; 1993. p. 317-60. [24] Williamson DA, Anderson DA, Jackman LP, Jackson SR. Assessment of eating disordered thoughts, feelings, and behaviors. In: Allison DB, editor. Handbook of assessment methods for eating behaviours, and weight related problems: measures, theory and research. Thousand Oaks (Calif): Sage; 1995. p. 347-86. [25] Arnow B, Kenardy J, Agras WS. Binge eating among the obese: a descriptive study. J Behav Med 1992 Apr;15:155-70. [26] Arnow B, Kenardy J, Agras WS. The Emotional Eating Scale: the development of a measure to assess coping with negative affect by eating. Int J Eat Disord 1995;18:79-90. [27] Doane JA, Diamond D. Affect and Attachment in the Family. New York: Basic Books; 1994. [28] Shisslak CM, Crago M, Estes LS. The spectrum of eating disturbances. Int J Eat Disord 1995;18:209-19 Review. [29] Fairburn CG, Bohn K. Eating disorder NOS (EDNOS): an example of the troublesome ‘‘not otherwise specified” (NOS) category in DSM-IV. Behav Res Ther 2004;43:691-701. [30] Dare C, Le Grange D, Eisler I, Rutherford J. Redefining the psychosomatic family: family process of 26 eating disorder families. Int J Eat Disord 1994;16:211-26. [31] Besharat A, Eisler I. Family expressed emotion in eating disorders patients. Eu Psychiatr 1996;11:362s-3s. [32] Szmukler GI, Eisler I, Russell GFM, Dare C. Anorexia nervosa, parental expressed emotion, and dropping out of treatment. Br J Psychiatry 1985;147:265-71. [33] Van Furth EF, Van Strien DC, Martina LML, Van Son MJM, Hendrickx JJ, Van Engeland H. Expressed emotion and the prediction of outcome in adolescent eating disorders. Int J Eat Disord 1996;20:19-31. [34] Brown GW, Birley JLT, Wing JK. Influence of family life on the course of schizophrenic disorders: a replication. Br J Psychiatry 1972;121:241-58. [35] Engelberg MJ, Steiger H, Gauvin L, Wonderlich SA. Binge antecedents in bulimic syndromes: an examination of dissociation and negative affect. Int J Eat Disord 2007;40:531-6. [36] Stein RI, Kenardy J, Wiseman CV, Dounchis JZ, Arnow BA, Wilfley DE. What's driving the binge in binge-eating disorder?: a prospective examination of precursors and consequences. Int J Eat Disord 2007;40:195-203. [37] Telch CF, Agras WS. Do emotional states influence binge eating in the obese. Int J Eat Disord 1996;20:271-9. [38] Deaver CM, Miltenberger RG, Smyth J, Meidinger A, Crosby R. An evaluation of affect and binge eating. Behav Modif 2003;27: 578-99.