Psychometrics, Assessment, and Diagnosis

In a series of papers, we demonstrated the validity and reliability of psychological scales that are not commonly applied to the eating disorders: the Eating Disorders Inventory as applied to binge eating disorder; the Personality Assessment Inventory; and a novel measure of eating disorder urges. Each were shown to be valid and reliable by our research. Currently, we are collaborating with colleagues at other centres to assess the validity of a group therapeutic factors scale, a measure of eating disorders knowledge, and a procedure to assess health care utilization. We also published an extensive review of assessment methods for binge eating and purging. Below is a list of publications for the Psychometrics and Assessment research theme.

Flament, M. F., Hill, E. M., Buchholz, A., Henderson, K., Tasca, G. A., & Goldfield, G. (in press). Internalization of the thin and muscular body ideal and disordered eating in adolescence: The mediation effects of body esteem. Body Image.

Abstract: This study investigates body esteem factors (weight-esteem and appearance-esteem) as mediators of the relationship between 'internalization of the ideal body figure' and disordered eating behaviors (restrained, emotional and external eating) in a community sample of adolescent males (n = 810) and females (n = 1137) who participated in the Ontario Research on Eating and Adolescent Lifestyles (REAL) study. Mediation models were tested using both the bootstrap approach and the Sobel test. In males, appearance-esteem completely mediated the relationship between internalization of the muscular ideal and restrained eating; appearance-esteem partially mediated effects in the emotional and external eating regressions. In females, appearance-esteem partially mediated the relationship between internalization of the thin ideal and all three forms of disordered eating. In both genders, weight-esteem partially mediated in the restrained and emotional eating regressions. We discuss the unique influence of each body esteem variable and the gender-specific qualities of the mediated relationships.
Joyce, A. S., MacNair-Semands, R., Tasca, G. A., & Ogrodniczuk, J. S. (in press). Factor structure and validity of the Therapeutic Factors Inventory- Short Form. Group Dynamics.

Abstract: Clinical group practice is commonly guided by the assumption that 11 distinct therapeutic factors operate in psychotherapy groups. This assumption should be examined closely as evidence for fewer, more global therapeutic factors emerges. The present study was conducted to continue the development of the Therapeutic Factors Inventory – Short Form (TFI-S), and focused on confirming its factor structure and evaluating aspects of its concurrent, discriminant, and predictive validity. A sample comprising a total of 435 clients from 52 distinct therapy groups at eight different clinical sites in the US and Canada was obtained. Structural equation modeling analyses indicated that a four-factor model had reasonably good fit to the TFI-S rating data, and provided indications for further refinement of the constituent items. The analyses resulted in a revised, 19-item measure, the TFI-19. Three-level HLM analyses demonstrated that the four factors were sensitive to change in member perceptions over time. Significant and meaningful relationships between the TFI-19 factors and subscales of the Group Climate Questionnaire-S provided support for the concurrent validity of the TFI-19. Two-level HLM analyses identified significant relationships between the TFI-19 factors and post-treatment status on symptomatic and interpersonal distress, thereby supporting the predictive validity of the TFI-19. The findings suggest that the TFI-19 may be a useful instrument to assess group members' perceptions of more global therapeutic factors.
Tasca, G. A., Keating, L., Maxwell, H., Hares, S., Trinneer, A., Barber, A., Bradwejn, J., & Bissada, H. (2011). Predicting acceptance of intensive treatment and participation in a randomized control trial of medication among women with anorexia nervosa. European Eating Disorders Review. Published online July, 2011.

Abstract: Objective: This study aimed to identify predictors of acceptance of intensive treatment and of participation in a randomized controlled trial (RCT) among women with anorexia nervosa (AN). Method: Participant data were drawn from a tertiary care intensive treatment programme including a previously published RCT. Women with AN (N = 106) were offered intensive treatment, and 69 were approached to participate in an RCT of olanzapine's efficacy as an adjunctive treatment for AN. AN subtype and pretreatment psychological variables were used to predict acceptance of intensive treatment and RCT participation. Results: AN binge purge subtype and higher depression and body dissatisfaction predicted intensive treatment acceptance. No variable predicted RCT participation among treatment acceptors. Discussion: Clinicians may focus on enhancing motivation or use a stepped care approach to increase intensive treatment acceptance especially among women with AN-restricting type and among all those with AN who have lower levels of distress. Copyright © 2011 John Wiley & Sons, Ltd and Eating Disorders Association.
Tasca, G. A., Maxwell, H., Bone, M., Trinneer, A., Balfour, L., & Bissada, H. (2011) Purging disorder: Psychopathology and treatment outcomes. International Journal of Eating Disorders. Published online, February, 2011.

Abstract: Objective: To characterize a tertiary care treatment-seeking sample and assess post-treatment remission and completion rates for purging disorder (PD). Method: Consecutively admitted women with PD (n 5 122), anorexia nervosa (AN) restricting subtype (AN-R; n 5 146), AN binge-purge subtype (AN-B; n 5 154), and bulimia nervosa-purging subtype BN-P; n 5 415) were compared on general and eating disorder specific psychopathology. A subsample (n 5 256) attended a day treatment program and were assessed for post-treatment remission and completion rates. Results: PD occurred in 17.3% of eating disorders not otherwise specified and 6.7% of total eating disorder referrals. PD patients were similar to AN-B and BN-P, but had higher levels than AN-R, on general and eating disorder psychopathology. PD individuals did not have different post-treatment remission or completion rates compared to other eating disorders. Discussion: The results add to a growing literature on the utility of PD as a diagnosis.
Tasca, G. A., Ritchie, K., & Balfour, L. (2011). Practice review: Implications of attachment theory and research for the assessment and treatment of eating disorders. Psychotherapy. Published online, March, 2011

Abstract: In this paper, we review the research literature on attachment and eating disorders and suggest a framework for assessing and treating attachment functioning in patients with an eating disorder. Treatment outcomes for individuals with eating disorders tend to be moderate. Those with attachment associated insecurities are likely to be the least to benefit from current symptom-focused therapies. We describe the common attachment categories (secure, avoidant, anxious), and then describe domains of attachment functioning within each category: affect regulation, interpersonal style, coherence of mind, and reflective functioning. We also note the impact of disorganized mental states related to loss or trauma. Assessing these domains of attachment functioning can guide focused interventions in the psychotherapy of eating disorders. Case examples are presented to illustrate assessment, case formulation, and group psychotherapy of eating disorders that are informed by attachment theory. Tailoring treatments to improve attachment functioning for patients with an eating disorder will likely result in better outcomes for those suffering from these particularly burdensome disorders.
Grenon, R., Tasca, G., Cwinn, E., Coyle, D., Sumner, A., Gick, M., & Bissada, H. (2010). Depressive symptoms are associated with medication use and lower health related quality of life in overweight women with binge eating disorder. Women's Health Issues, 20, 434-440

Abstract: Background: Depression is a most burdensome illness, with personal and societal costs surpassing those of any other illness. Furthermore, depression affects women at a much higher rate than men. The most prevalent eating disorder among adult women is binge eating disorder (BED). Depression and obesity are common in women with BED, most of whom seek treatment later in life. Depression, obesity, and age are associated with greater health care use and lower health-related quality of life (HRQOL). Hence, for women with BED estimating the effects of depression can be confounded by both age and body mass index (BMI). The current study examined the relationships between depression, HRQOL, and health care utilization among treatment seeking women with BED. Methods: Participants (n ¼ 105) completed the Structured Clinical Interview for DSM-IV, a health care utilization and cost survey, the Personality Assessment Inventory depression scale, and the EQ-5D to measure HRQOL. Findings: On average, participants were severely obese with a mean BMI of 38.20 (SD ¼ 6.80); 67.27% had a lifetime history of depression. Participants had higher health care costs and lower HRQOL than published age- and gender-matched norms. After controlling for age and BMI, depressive symptoms were significantly related to greater medication use (excluding antidepressants), and lower HRQOL. Conclusion: Results suggest that targeting depressive symptoms may reduce the economic and personal burden of BED for women.
Tasca, G.A., Demidenko, N., Krysanski, V., Bissada, H., Illing, V., Gick, M., Weekes, K., & Balfour, L. (2009). Personality dimensions among women with an eating disorder: Towards reconceptualizing DSM. European Eating Disorders Review, 17, 281-289.

Article Title: Personality dimensions among women with an eating disorder: Towards reconceptualizing DSM
Authors: Tasca GA; Demidenko N;Krysanski V; Bissada H;Illing V; Gick M; Weekes K; Balfour L;
Journal: Eur Eat Disord Rev Volume, Issue
Date of Publication: 2009 May 6
Abstract:
Objective: To evaluate the incremental validity of a dimensional assessment of personality, after controlling for diagnostic category, in accounting for meaningful variation in eating disorder attitudes and behaviours and in current affective distress among a clinical sample of eating disordered women. METHODS: 244 treatment seeking eating disordered women and 116 non-eating disordered women were assessed with the NEO five factor inventory (NEO-FFI), and with measures of eating disorder attitudes and of affective distress using a cross sectional design. RESULTS: As predicted, differences were found between eating disordered and non-eating disordered women on several NEO-FFI scales, which provided a context for subsequent analyses. NEO-FFI scales accounted for meaningful variation in eating disordered attitudes and behaviours as well as in levels of current affective distress over and above DSM-IV diagnostic category. Conclusion: A flexible approach to diagnosis, which includes personality dimensions along with a description of eating disorder symptoms, may result in a more inclusive and useful diagnostic scheme for treating women with eating disorders. Copyright (c) 2009 John Wiley & Sons, Ltd and Eating Disorders Association.
Tasca, G.A., Illing, V., Balfour, L., Krysanski, V., Demidenko, N., Nowakowski, J., & Bissada, H. (2009). Psychometric properties of self-monitoring of eating disorder urges among treatment seeking women: Ecological momentary assessment using a daily diary method. Eating Behaviors, 10, 59-61.

Article Title: Psychometric properties of self-monitoring of eating disorder urges among treatment seeking women: ecological momentary assessment using a daily diary method
Authors: Tasca GA; Illing V; Balfour L; Krysanski V; Demidenko N; Nowakowski J; Bissada H;
Journal: Eat Behav Volume 10 , Issue1
Date of Publication: 2009 Jan
Abstract:
Objective: The psychometric properties of an ecological momentary assessment (EMA) method used to evaluate eating disorder (ED) urges were examined. METHODS: Participants, 139 women who sought treatment for an ED, completed a daily diary measuring ED urges after each meal for a three week period at the start of treatment, and a measure of ED attitudes and behaviors pre- and post-treatment. RESULTS: The construct validity of this method was indicated by a two factor solution representing binge eating urges and ED compensatory behavior urges, and by significant differences between ED diagnostic groups on ED urge type. Correlations of ED urges early in therapy with outcomes provided evidence for predictive validity. Correlations at pre-treatment between ED urges and ED attitudes and behaviors supported criterion validity. EMA of ED urges showed good retest stability. CONCLUSIONS: EMA of ED urges is a valid and reliable approach that is associated with ED symptom severity, and predictive of treatment outcome.
Tasca, G.A., Krysanski, V., Demidenko, N., & Bissada, H. (2009). Binge eating and purging. In D.B. Allison and M.L. Baskin (Eds.), Handbook of assessment methods for obesity and eating behaviors, related problems, and weight: Measures, theory, and research (2nd ed., pp .347-396). Thousand Oaks, CA: Sage Publications.

Title: Binge Eating and Purging.
Authors: Tasca, GA; Krysanski, V; Demidenko, N; Bissada, H.
Book Chapter: D.B. Allison and M.L. Baskin (Eds.), Handbook of assessment methods for obesity and eating behaviors, related problems, and weight: Measures, theory, and research (2nd ed., pp .347-396). Thousand Oaks, CA: Sage Publications.
Date of Publication: 2009
Abstract: In this chapter, we define the parameters that must be addressed to reliably measure binge eating and purging. In doing so, we discuss some common difficulties encountered in attaining the desired uniformity in the definition of these parameters. Next, we review the methods available for assessing these behaviors. Assessment methods are organized into 2 categories: 1) self-report questionnaires and 2) structured clinical interviews. We will discuss the most commonly used specific instruments, highlighting the psychometric properties, research evidence, and relative advantages and disadvantages they offer the researcher and the clinician.
Tasca, G.A., Balfour, L., Kurichh, K., Potvin-Kent, M., Bissada, H. (2006). Actual-desired BMI discrepancy, body dissatisfaction, and self concept in women with Bulimia Nervosa and Binge Eating Disorder. In P. Swain (Ed.) Anorexia Nervosa and Bulimia Nervosa: New Research (pp.145-158). New York: Nova Science Publishers.

Title: Actual - desired BMI discrepancy, body dissatisfaction and self concept in women with bulimia nervosa and binge eating disorder. Anorexia nervosa and bulimia: New research.
Authors: Tasca, Giorgio A.; Balfour, Louise; Kurichh, Kanchan; Potvin-Kent, Monique; Bissada, Hany
Book: Swain, Pamela I. (2006). Anorexia nervosa and bulimia: New research. (pp. 145-158). Hauppauge, NY, US: Nova Science Publishers. xi, 219 pp.
Date of Publication: 2006
Abstract: It was hypothesized that higher levels of actual-desired BMI discrepancy (ADBD) would be associated with higher binge eating symptoms, lower self concept, and higher body dissatisfaction for a clinical sample of women with Bulimia Nervosa (BN) and Binge Eating Disorder (BED). A theoretical frame for ADBD was put forward based on self discrepancy theory. Also hypothesized was that women with BN would have greater eating and self related pathology than those with BED. Fifty-one participants diagnosed with BN and 41 with BED drawn from a clinical sample completed questionnaires assessing eating disorder and general psychopathology. Greater ADBD was related to greater body dissatisfaction for the BN and BED groups, and lower self concept for the BN group. Those with BN had more self-related pathology. ADBD can be easily assessed by clinicians and may be used as an index of body dissatisfaction and overall self concept. ADBD may be a vulnerability factor for developing an eating, disorder for women. (PsycINFO Database Record (c) 2009 APA, all rights reserved).
Tasca, G.A., Illing, V., Lybanon-Daigle, V., Bissada, H., & Balfour, L. (2003). Psychometric properties of the Eating Disorders Inventory - 2 among women seeking treatment for Binge Eating Disorder. Assessment, 10, 228-236.

Article Title: Psychometric properties of the eating disorders inventory-2 among women seeking treatment for binge eating disorder
Authors: Tasca GA; Illing V; Lybanon-Daigle V; Bissada H; Balfour L;
Journal: Assessment Volume 10, Issue 3
Date of Publication: 2003 Sep
Abstract: Psychometric properties of the Eating Disorders Inventory-2 (EDI-2) when used for women with Binge Eating Disorder (BED) are assessed. The EDI-2 was administered to 144 outpatients seeking treatment for BED and 152 outpatients seeking treatment for Bulimia Nervosa (BN). Most EDI-2 scales had acceptable internal consistence for both the BED and BN samples. EDI-2 scales demonstrated adequate stability within a subsample of those with BED who were retested. Confirmatory factor analyses revealed a hypothesized second-order two-factor structure for the original EDI scales for the BED group but not for those with BN. When the provisional EDI-2 scales were included, a two-factor structure was not supported for any group. Some scales differentiated the BED from the BN sample, and the second-order factors correlated with measures of similar constructs. The original EDI scales can be used reliably for those with BED.
Antoniou, M., Tasca, G.A., Wood, J., Bissada, H. (2003). Binge Eating Disorder versus overeating: A failure to replicate and common factors in severely obese treatment seeking women. Eating and Weight Disorders, 8, 145-149.

Article Title: Binge eating disorder versus overeating: a failure to replicate and common factors in severely obese treatment seeking women
Authors: Antoniou M; Tasca GA; Wood J; Bissada H;
Journal: Eating & Weight Disorders: EWD Volume 8, Issue 2
Date of Publication: 2003 June
Abstract: The disordered eating symptoms, general psychopathology and dieting history among obese women diagnosed with Binge Eating Disorder (BED) and obese women who overeat (OE) are examined. One hundred and thirty women (n=83 with BED and n=47 who overeat) seeking treatment for an eating disorder were diagnosed using the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) criteria and the Eating Disorders Examination (EDE). They also completed a battery of psychometric tests. Despite adequate statistical power to detect differences, MANOVAs revealed very few significant differences between the groups. Loss of control of eating does not adequately differentiate these two groups within an eating disorders treatment-seeking context. It is likely that only the most acutely distressed from each group is seeking treatment, so that differences found in a community sample would not be found. Using a principal components analysis, factors within each group that may underlie the common psychopathology of both groups (i.e. disordered eating symptoms, general psychopathology and dieting history) were found.
Tasca, G.A., Wood, J., Demidenko, N., & Bissada, H. (2002). Using the PAI in an eating disordered population: Scale characteristics and differences among diagnostic groups. Journal of Personality Assessment, 79, 337-356.

Article Title: Using the PAI with an eating disordered population: scale characteristics, factor structure, and differences among diagnostic groups. Clinical guidelines for the treatment of depressive disorders. VII. Comorbidity.[comment]. [Review] [230 refs]
Authors: Tasca GA; Wood J; Demidenko N; Bissada H;Enns MW; Swenson JR; McIntyre RS; Swinson RP; Kennedy SH; CANMAT Depression Work Group;
Journal: Journal of Personality Assessment Volume 79, Issue 2
Date of Publication: 2002 Oct
Abstract: Psychometric properties of the Personality Assessment Inventory (PAI; Morey 1991) within an eating disordered sample seeking treatment (N = 238) and differences among eating disorder diagnostic groups on the PAI were examined. The PAI showed acceptable alpha coefficients, item-total correlations, and interitem correlations. The factor structure was similar to that reported by Morey (1991), with the addition of another factor related to interpersonal coolness and distance. Those with binge eating disorder (BED) reported fewer problems and less distress in general compared to other eating disordered groups. The BED and bulimia nervosa groups were different from the anorexia nervosa groups in frequency of matching on two PAI clusters. Use of the PAI with an eating disordered population and its utility in understanding eating disorder diagnostic groups is supported.