Binge Eating Scale (BES)
KeywordsEating Disorder Binge Eating Binge Eating Disorder Binge Eating Scale Binge Eating Episode
The binge eating scale (BES) is a 16-item self-report questionnaire designed to capture the behavioral (eight items, e.g., large amount of food consumed), as well as the cognitive and emotional (eight items, e.g., feeling out of control while eating, preoccupation with food and eating), features of objective binge eating (OBE) in overweight and obese adults (Gormally et al. 1982). For each item, respondents are asked to select one of three or four response options, coded zero to two or three, respectively. Individuals’ scores are summed and range from 0 to 46, with higher scores indicating more severe binge eating problems. Marcus et al. (1988) created clinical cutoff scores for the BES representing none-to-minimal (<17 total score), moderate (18–26), and severe (>27) binge eating problems. Importantly, the BES was created before binge eating disorder (BED) was officially recognized as a psychiatric diagnosis (American Psychiatric Association 2013) and thus is not intended to detect the presence of this disorder. Rather, it has been suggested that this measure be used as a brief screening tool to identify the severity of binge eating behavior in overweight and obese adults, to tailor obesity interventions, and to track treatment outcomes (Gormally et al. 1982; Marcus et al. 1988). Although the BES has been used with children and adolescents sparingly, we only present data from adults in this chapter, as this measure was created specifically for this age group.
Estimated internal consistency of the measure is generally acceptable (Cronbach’s alphas were reported above .8) across samples, including men and women from the community, college students, treatment-seeking adults, racially/ethnically diverse groups, and adults from the United States and abroad (e.g., Celio et al. 2004; Freitas et al. 2006; Kelly et al. 2012; Minnich et al. 2014; Ricca et al. 2000; Timmerman 1999). Adequate test-retest reliability has also been reported among adult females’ samples over a ~2-week interval (e.g., Timmerman 1999).
The initial development of the scale (Gormally et al. 1982) yielded two eight-item factors representing the behavioral and cognitive/emotional aspects of binge eating. More recent examinations of the scale (e.g., Kelly et al. 2012) further support this two-factor solution, suggesting that the existing subscales remain appropriate across a range of samples. Of interest, Kelly et al. reported measurement invariance between White/Caucasian and Black/African American female college samples, suggesting that, while the same two-factor solution may be adequate for both racial/ethnic groups, the items might be assessing different constructs (see Race/Ethnicity subsection for further discussion).
In support of the validity of the BES, numerous studies have identified significant associations with other questionnaires assessing related attitudes, behavior, and mood symptoms. For instance, Mitchell and Mazzeo (2004) examined the BES in a sample of 259 undergraduates, including 73 Black/African American women and 131 White/Caucasian women, and found that BES scores in both groups of women were moderately to strongly correlated with a measure of general eating disorder symptomatology (r = 0.80 and 0.52, respectively). In Ricca et al.’s (2000) research involving 344 outpatient men and women with obesity, BES scores were significantly associated with state (r = 0.25) and trait anxiety (r = 0.32), depressive symptoms (r = 0.30), and BMI (r = 0.13) providing further evidence of concurrent validity.
Although the BES has primarily been studied in women, researchers have specifically examined the BES in men. For example, Minnich et al. (2014) examined the BES in a sample of 302 primarily White/Caucasian (88.8 %) undergraduate men. Concurrent validity was established with significant correlations in the expected directions at two separate time points with measures of body dissatisfaction (r = 0.52, 0.31), drive for muscularity (r = 0.19, 0.15), self-esteem (r = 0.39, 0.25), depressive symptoms (r = 0.54, 0.50), anxiety symptomatology (r = 0.33, 0.27), and BMI (r = 0.27, 0.32). BES scores also appear consistent with established gender differences in BED prevalence, such that women endorse significantly higher average BES scores than men (e.g., Ricca et al. 2000).
The BES has been used among diverse adult samples in both the United States and abroad and has been translated into several different languages (e.g., Freitas et al. 2006; Ricca et al. 2000). Because the BES was originally developed and normed in a primarily White/Caucasian sample, research has since investigated whether racial/ethnic variations in this measure exist. Data from Kelly et al. (2012) indicate that White/Caucasian college women (n = 1467) reported a significantly stronger association between body dissatisfaction and BES scores than Black/African American college women (n = 741). White/Caucasian women also endorsed higher scores on the BES than their Black/African American peers (M = 11.42, SD = 8.16 and M = 8.69, SD = 6.80, respectively). Taken together, data indicate that, on average, binge eating behavior among Black/African American men and women may occur less frequently, may be less emotionally distressing, and may be less associated with body image concerns compared to their White/Caucasian peers. Qualitative data are needed to further clarify whether binge eating behavior manifests differently among various racial/ethnic groups.
Clinical Usefulness to Assess Binge Eating Episode Size
A core feature of BED according to DSM-5 criteria (American Psychiatric Association 2013) is the presence of recurrent objective bulimic episodes (OBEs), which involves the consumption of unambiguously large amounts of food, in addition to a sense of loss of control (LOC) while eating. However, subjective bulimic episodes (SBEs), the consumption of smaller amounts of food with LOC, are also associated with significant eating disorder symptomatology and general psychopathology. While the BES was only designed to capture OBEs, research has examined whether this measure adequately captures both types of binge eating episodes.
Timmerman (1999) compared BES scores to data from 28-day food records in a sample of 56 healthy, primarily White/Caucasian (91 %) women who engaged in OBEs at least twice per month in the absence of regular compensatory behaviors. Participants received training in measuring and recording food intake accurately prior to beginning the study. The BES demonstrated significant, moderate associations (r = 0.39–0.40) with several indicators of SBEs, including calories consumed during SBEs, number of SBEs, and number of SBE days. Small-to-moderate, significant correlations (r = 0.29–0.32) were also noted between the BES and similar indicators of OBEs. Importantly, BES scores were not correlated with overall caloric intake. Timmerman (1999) also divided participants into groups based on BES clinical cutoffs. Significant differences emerged between the mild (≤17) and severe (≥27) groups, such that the severe group endorsed higher levels of the following: number of calories consumed during SBEs, number of calories consumed during SBEs and OBEs combined, number of total SBEs, number of OBE days, number of SBE days, and number of combined OBE and SBE days. In Celio et al.’s (2004) research, the BES was also correlated with frequency of OBEs and number of days, but no other form of overeating, including SBEs or objective overeating without LOC. Taken together, existing research indicates that the BES does not successfully discriminate between SBEs and OBEs, but appears to be a good indicator of severity of LOC eating.
Comparisons to Diagnostic Interviews
Although the BES was not developed with the intention of diagnosing BED, some researchers have examined the potential utility of this measure as a replacement for more costly, time-consuming diagnostic clinical interviews that require trained mental health staff. Therefore, it is important to consider the concordance between the BES and standard diagnostic clinical interviews.
Eating Disorder Examination (EDE, Fairburn and Cooper 1993)
Celio et al. (2004) compared the BES with the gold standard interview for eating disorder symptom assessment, the EDE, in a sample of 157 primarily White/Caucasian (70.3 %) women seeking treatment for BED. Results suggested that the BES (using a severe cutoff score of 27) was reasonably sensitive in detecting individuals with BED (85.1 %), but demonstrated low specificity (20 %), such that many women without BED obtained BES scores above the severe cutoff.
Structured Clinical Interview for DSM Disorders (SCID; First et al. 2007)
Freitas and colleagues (2006) examined the clinical utility of the Portuguese version of the BES as a measure of BED in a sample of 178 Brazilian women with obesity. Comparisons were made between the SCID, a semi-structured psychiatric interview, and a BES cutoff score of 17. Again, the BES demonstrated higher sensitivity (97.8 %) than specificity (47.7 %); over half of the participants who were not diagnosed with BED according to the SCID scored above the clinical cutoff on the BES, while only ~2 % of those diagnosed with BED scored below the cutoff. A large sample of outpatient men and women with obesity (N = 344; Ricca et al. 2000) completed the SCID and the BES; their data yielded a sensitivity of 84.8 % and a specificity of 74.6 %.
Based on the existing research, it appears that the BES may demonstrate slightly stronger concordance with clinical interviews in community samples compared to clinical samples, although the potential for false positives remains a concern in the majority of studies. As such, it is recommended that the BES not be used independently to diagnose BED, but may be a useful initial screening tool (as originally suggested by Gormally et al. (1982)).
Summary and Future Directions
Overall, scores on the BES (Gormally et al. 1982) have demonstrated good reliability, and the measure appears to be valid for the assessment of binge eating severity for clinical, community, and college samples in the United States and abroad. Given discordance between the BES and clinical interviews, particularly the tendency to overdiagnose BED when using the BES clinical cutoffs, it is suggested that the BES be used as a brief screening device rather than a diagnostic indicator. The BES may also be more useful in terms of capturing general LOC eating patterns rather than identifying specific binge episodes. Although the BES may be used with diverse patient and community samples, there remains a need to further examine the validity of this measure, particularly among Asian and Hispanic/Latino groups. Additional research is also needed with men outside of college samples, including community and clinical settings. Longitudinal studies are also needed to evaluate whether the BES is a prospective indicator of disordered eating pathology as the majority of research cited herein reported cross-sectional data.
References and Further Reading
- Celio, A. A., Wilfley, D. E., Crow, S. J., Mitchell, J., & Walsh, B. T. (2004). A comparison of the binge eating scale, questionnaire for eating and weight patterns-revised, and eating disorder examination questionnaire with instructions with the eating disorder examination in the assessment of binge eating disorder and its symptoms. International Journal of Eating Disorders, 36(4), 434–444.CrossRefPubMedGoogle Scholar
- Fairburn, C., & Cooper, Z. (1993). The eating disorder examination. In C. Fairburn & G. Wilson (Eds.), Binge eating: Nature, assessment and treatment (12th ed., New York, NY : Guilford, pp. 317–360).Google Scholar
- First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. (2007). SCID-I/P.Google Scholar