Eating Problems, Body Image Disturbances, andAcademic Achievement: Preliminary Evaluation of the

Eating and Body Image Disturbances AcademicInterference Scale

Tovah Yanover, MAJ. Kevin Thompson, PhD*

ABSTRACT

Objective: To examine the relation-

ships between a new scale, the Eating

and Body Image Disturbances Academic

Interference Scale (EBIDAIS), and meas-

ures of eating disturbance, body image,

and academic achievement.

Method: One thousand five hundred

eighty-four college undergraduates com-

pleted the measures in an online survey

and were awarded class credit for their

participation. Measures included the Eat-

ing Disorder Inventory Bulimia, Drive for

Thinness, Body Dissatisfaction, and Per-

fectionism subscales. Grade point aver-

age (GPA) was also reported.

Results: Academic interference and

GPA were significantly correlated, indicat-

ing that higher interference scores were

related to lower GPA. EBIDAIS was also

significantly correlated with drive for

thinness, bulimia, and body dissatisfac-

tion, but was not significantly associated

with perfectionism. The correlation be-

tween interference and GPA was substan-

tially higher for a subsample of individu-

als who scored in the elevated range on

eating and body dissatisfaction.

Conclusion: Academic interference may

be a relatively unexamined, but poten-

tially important, outcome for individuals

who experience eating problems and

body image disturbance. VVC 2007 by

Wiley Periodicals, Inc.

Keywords: academic performance; eat-

ing disturbances; grade point average;

body image

(Int J Eat Disord 2008; 41:184–187)

Introduction

The eating disorders of anorexia nervosa and buli-mia nervosa fall at the most severe end of the con-tinuum of eating disturbance. Anorexia nervosa ischaracterized by intense fear of fatness, refusal tomaintain a healthy weight, and body image distur-bances.1 Disturbed body image is also a core fea-ture of bulimia nervosa along with recurrentepisodes of binging and purging or other compen-satory behaviors such as excessive exercise, laxativeand diuretic use, or fasting.1 Eating disorders repre-sent a serious problem with sequelae includingpsychiatric comorbidity,2,3 and high rates of mor-bidity and mortality.4 However, eating and bodyimage disturbances may also be present, to a less

severe extent, in nonclinical samples.5 Whether

clinical or subclinical symptoms exist, it is impor-

tant to determine the specific role that eating and

body image problems play in altering one’s ability

to engage in effective psychosocial and occupa-

tional functioning.

One area that has received a limited degree of

attention is the role that eating problems and body

image disturbances might play in interfering with

optimal cognitive performance. Research has

shown that eating disturbance has acute, detrimen-

tal effects on performance on a range of cognitive

tasks6–8 including problem solving ability.7 Could it

also be that academic performance and achieve-

ment might suffer due to the inhibiting effects of a

focus on the issues associated with eating and body

image problems? For instance, binge eating and

purging, and the aftereffects of purging, such as fa-

tigue, might conceivably interfere with one’s ability

to study or complete classroom assignments. Addi-

tionally, body image dissatisfaction might lead to

a high level of class absenteeism due to social anxi-

ety regarding appearance. Importantly, one moder-

ating factor that could affect the relationship

Accepted 8 August 2007

Department of Psychology, University of South Florida, Tampa,

Florida

*Correspondence to: J. Kevin Thompson, PhD, Department of

Psychology, University of South Florida, Tampa, FL 33620.

E-mail: thompsonjk2003@yahoo.com

Published online 5 October 2007 in Wiley InterScience

(www.interscience.wiley.com). DOI: 10.1002/eat.20483

VVC 2007 Wiley Periodicals, Inc.

184 International Journal of Eating Disorders 41:2 184–187 2008—DOI 10.1002/eat

BRIEF REPORT

between academic achievement and eating dis-turbance is the presence of perfectionism. Severalstudies have found a link between perfectionismand eating problems (see Refs. 9 and 10 forreviews). Dura and Bornstein11 also found thatschool achievement was higher than would beexpected given IQ for school-age females with an-orexia nervosa. Therefore, some individuals withstrong perfectionistic tendencies may not allowtheir eating problems to interfere with theirachievement.

For the present study, a new scale, the Eating andBody Image Disturbances Academic InterferenceScale (EBIDAIS), was developed. This new scale isdesigned to assess the degree to which some ofthe core features and symptoms of eating disturb-ance interfere with one’s academic achievement.The purpose of the present study was to examinethe relationships between this new scale and mea-sures of eating and body image disturbance, gradepoint average (GPA), and perfectionism.

Method

Participants

Initially, 1,632 individuals participated in the study. Of

these, 31 were removed because they had completed less

than 75% of the measures or their data contained incon-

sistencies leading to questionable validity of the data.

Sixteen individuals were removed because their data had

been recorded incorrectly due to computer error. One

individual was removed because he had participated

twice. The final sample consisted of 1,584 undergradu-

ates (1,181 women and 397 men) from a large southeast-

ern state university, who obtained extra course credit for

participation. The mean age of the sample was 20.43

years (SD 5 3.714). The sample was predominantly Cau-

casian Non-Hispanic (63.4%) with 12.8% of the partici-

pants self-identifying as Hispanic, 11.4% self-identifying

as African-American Non-Hispanic, 5.2% as Asian, and

6.6% as other.

Measures

Drive for Thinness. The Eating Disorder Inventory 2-

Drive for Thinness (EDI-DT; Ref. 12) subscale was used to

measure drive for thinness. This scale measures restrict-

ing tendencies, desire to lose weight, and fear of weight

gain. The EDI-DT shows good internal consistency in

clinical samples of adults and adolescents in both the

United States and internationally (all a [ 0.9). The test–

retest reliability is also very good (r 5 .95). Internal con-

sistency in the current sample was excellent (a 5 0.91).

Bulimic Symptoms. The Eating Disorder Inventory 2-

Bulimia (EDI-B; Ref. 12) subscale was used to assess

bulimic symptoms. The EDI-B assesses tendencies to

think about and engage in bouts of uncontrollable over-

eating (bingeing). The EDI-B shows good the test–retest

reliability (r 5 .94). Internal consistency is highest in

international samples (a\ 0.75). Internal consistency in

the present sample was very good (a 5 0.85).

Perfectionism. The Eating Disorder Inventory 2-Perfec-

tionism (EDI-P; Ref. 12) subscale was used to assess per-

fectionism. The EDI-P assesses the extent to which one

values the achievement of high goals and puts a pre-

mium on the highest possible standards of personal

achievement. The EDI-P shows good test–retest reliability

(r 5 .93). Internal consistency is also acceptable (a \0.74). Internal consistency in the present sample was

very good (a 5 0.83).

Body Dissatisfaction. The Eating Disorder Inventory 2-

Body Dissatisfaction subscale (EDI-BD; Ref. 12) is a 9-

item scale that assesses overall satisfaction with weight-

related body sites. The EDI-BD shows good internal con-

sistency in clinical samples of adults and adolescents in

both the United States and internationally (all a [ 0.9).

The test–retest reliability is also very good (r 5 .95). Inter-

nal consistency in the present sample was excellent (a 5

0.91).

Eating and Body Image Disturbances Academic Interfer-ence Scale. This 12-item scale assesses the degree to

which symptoms of eating problems and body dissatis-

faction interfere with class attendance, attention paid in

class, and homework completion. Participants respond

on a 5-point Likert-type scale from ‘‘never’’ to ‘‘always.’’

Internal consistency in the current sample was excellent

(a 5 0.92). A few sample items of this measure are: ‘‘How

often do you have difficulty concentrating in class

because you are worrying about food or eating?’’ and

‘‘How often do you miss class because you feel ill from a

binge or a purge?’’ An exploratory factor analysis indi-

cated the existence of one main factor which explained

55% of the variance. No factor loading for any of the 12

items was below .55.

Academic Achievement. Participants self-reported their

most recent GPA on a 4.0 point scale. Participants also

reported the number of hours per week of class in which

they were enrolled and the number of hours per week

they attended. They also reported the number of weekly

hours of homework received and number of hours of

weekly homework completed.

Procedure

Participants completed all of the measures online. De-

mographic information, including GPA, weight, and

PRELIMINARY EVALUATION OF EBIDAIS

International Journal of Eating Disorders 41:2 184–187 2008—DOI 10.1002/eat 185

height information, was obtained after completion of the

main study measures. Because of computer error during

data collection, height information was obtained only

from 1,305 individuals.

Results

The mean BMI of the sample was 24.03 (SD 55.09), which falls at the high end of the normalweight range. There were 81 underweight individu-als in the sample, 786 normal weight individuals,258 overweight individuals, and 157 obese individ-uals. The mean GPA of the sample was 3.17 (SD 50.50).

EBIDAIS and GPA were significantly correlated(r 5 2.07, p\ .01), indicating that higher interfer-ence scores were associated with lower GPA. EBI-DAIS was also significantly correlated with EDI-B(r 5 .44, p \ .01), EDI-DT (r 5 .30, p \ .01), andEDI-BD (r 5 .26, p\ .01), indicating that higher in-terference scores were associated with higher levelsof disturbance. Perfectionism was significantly cor-related with GPA (r5 .09, p\ .01), body dissatisfac-tion (r 5 .09, p \ .01), drive for thinness (r 5 .19,p\ .01), and bulimia (r 5 .13, p\ .01), but not withacademic interference (r 5 2.01, ns). None of theacademic variables were significantly correlatedwith EBIDAIS score: hours of weekly class (r 5 .01,ns), hours of class attended weekly (r 5 .00, ns),hours of weekly assigned homework (r 5 .00, ns),and hours of homework completed weekly (r 5 .11,ns; see Table 1).

The pattern of correlations when the sample wasdivided into males and females was similar to thatobtained for the complete sample, with a fewexceptions. The correlations between perfection-ism and the other variables were significant forwomen, but not for men. Interestingly, the associa-tion between perfectionism and interference was

nonsignificant for women (r 5 .04, ns), but signifi-cantly negative for men (r 52.14, p\ .01).

Although there was a significant correlationbetween EBIDAIS and GPA, the relationship is quiteweak at r 5 2.07, indicating a small effect size.Therefore, to further examine the issue, we createda set of extreme groups based on the eating disor-der inventory (EDI) subscales, by comparing thosewho scored 1 standard deviation above the meanscore (disturbed) to those who scored 1 standarddeviation below the mean (controls). (Therefore,we were comparing those who scored in the top16% on the measure with the bottom 16%.) We lim-ited these analyses to individuals whose BMI wasbelow 24.9 to produce a more typical eatingdisturbed sample as opposed to an overweightsample. We also wanted to partial out the effects ofperfectionism as a potential moderator, especiallybecause perfectionism was significantly associatedwith each of the EDI subscales.

For the EDI-DT, the disturbed group consisted of136 individuals and the control group consisted of199 participants. For the disturbed group, therewas a significant correlation between GPA and EBI-DAIS scores (r 5 2.23, p\ .01); however, the corre-lation in the control group was not significant (r 52.06, ns). Partialing out perfectionism from the sig-nificant correlation in the disturbed group did notchange the relationship, which remained at r 52.23. For the EDI-BD, the disturbed group con-sisted of 84 individual and the control group con-sisted of 219 individuals. Again, for the disturbedgroup, there was a significant correlation betweenGPA and EBIDAIS scores (r 5 2.30, p\ .01); how-ever, the correlation in the control group was notsignificant (r 5 2.08, ns). Partialing out perfection-ism score did not change the r for the disturbedgroup, which remained at r 5 2.30. For the EDI-Bmeasure, the disturbed group consisted of 93 indi-viduals and the control group consisted of 90 par-ticipants. Again, for the disturbed group, there wasa significant correlation between GPA and EBIDAISscores (r 5 2.33, p\ .01); however, the correlationin the control group was not significant (r 5 2.15,ns). Partialing out perfectionism from the r for thedisturbed group slightly attenuated the correlation(r 52.30), but it remained significant.

Conclusion

Overall, the present study indicates that higher lev-els of eating disturbance and body dissatisfaction

TABLE 1. Correlations among main study measures forfull sample and by sex

Variable EBIDAIS EDI-DT EDI-B EDI-P EDI-BD GPA

EBIDAIS — .34*/.28* .47*/.41* .04/2.14* .29*/.26* 2.05/2.07EDI-DT .30* — .61*/.49* .22*/.10* .68*/.59* .004/2.06EDI-B .44* .60* — .14*/.09 .48*/.46* 2.03/2.09EDI-P 2.01 .19* .13* — .12*/2.02 .11*/.06EDI-BD .26* .69* .49* .09* — 2.04/2.09GPA 2.07** .02 2.03 .09* 2.02 —

Notes: Values below the diagonal are for the full sample. Values abovethe diagonal are separated into female/male groups.

* p\ .01** p\ .05.

YANOVER AND THOMPSON

186 International Journal of Eating Disorders 41:2 184–187 2008—DOI 10.1002/eat

are associated with higher levels of interference inacademic achievement. The relationship was sig-nificant for the entire sample, but indicated arather small effect size. When the relationshipbetween EBIDAIS scores and GPA was examined ingroups of individuals whose scores for eating dis-turbed behaviors fell in the extreme range, signifi-cant and substantially higher correlations wereobserved, indicating a moderate effect size (in the.30 range). Significant correlations were notobserved in the control group.

One of the major strengths of this study was itssample size. The study is limited by its cross-sectional nature, which precludes causal interpre-tations of the findings. Furthermore, all of thescales were self-report. In addition, although wecreated extreme groups, we did not diagnose indi-viduals and assign them to DSM categories. It willbe important in future research to include individ-uals who are diagnosed with an eating disorder.

Despite the limitations of this study, the prelimi-nary, but novel finding that eating and body imagedisturbances are associated with interference inacademic functioning and achievement is impor-tant. The psychological and physical symptoms ofeating disorders that interfere with daily socialand occupational functioning have been welldelineated,1 however, the role of these symptomsas a possible deterrent to optimal academic per-formance has not been examined. Future researchis needed to expand upon these preliminary find-ings. In particular, it will be important to assess theinterference construct in childhood and adoles-

cence, where eating and body image related prob-lems may also have an impact on children’s schoolperformance.

References

1. Thompson JK, Roehrig M, Kinder B. Eating disorders. In: Hersen

M, Turner S, Beidel D, editors. Handbook of Psychopathology.

New York: Wiley & Sons. 2007;571–600.

2. Fichter MM, Quadflieg N. Six-year course and outcome of ano-

rexia nervosa. Int J Eat Disorder 1999;26:359–385.

3. Sullivan PF, Bulik CM, Carter FA, Joyce PR. Correlates of severity

in bulimia nervosa. Int J Eat Disorder 1996;20:239–251.

4. Reijonen JH, Pratt HD, Patel DR, Greydanus DE. Eating disorders

in the adolescent population: An overview. J Adolescent Res

2003;18:209–222.

5. Rodin J, Silberstein L, Striegel-Moore R. Women and weight: A

normative discontent. Nebr Symp Motiv 1984;32:267–307.

6. Green MW, Rogers PJ. Impaired cognitive-functioning during

spontaneous dieting. Psychol Med 1995;25:1003–1010.

7. Green MW, Rogers PJ. Impairments in working memory associ-

ated with spontaneous dieting behavior. Psychol Med 1995;28:

1063–1070.

8. Wing RR, Vazquez JA, Ryan CM. Cognitive effects of ketogenic

weight-reducing diets. Int J Obesity 1995;19:811–816.

9. Cassin SE, von Ranson KM. Personality and eating disorders: A

decade in review. Clin Psychol Rev 2005;25:895–916.

10. Franco-Paredes K, Mancilla-Dıaz JM, Vazquez-Arevalo R, Lopez-

Aguilar X, Alvarez-Rayon G. Perfectionism and eating disorders:

A review of the literature. Eur Eat Disord Rev 2005;13:61–70.

11. Dura JR, Bornstein RA. Differences between school achieve-

ment and IQ in anorexia nervosa. J Clin Psychol 1989;45:433–

435.

12. Garner DM. Manual for the Eating Disorder Inventory-2.

Odessa, FL: Psychological Assessment Resources, 1991.

PRELIMINARY EVALUATION OF EBIDAIS

International Journal of Eating Disorders 41:2 184–187 2008—DOI 10.1002/eat 187