personality characteristics of restrained/binge eaters versus unrestrained/nonbinge eaters

5
AddictiveBehaviors, Vol. 11, pp. 207-211, 1986 Printed in the USA. All rights reserved. 0306-4603/86 $3.00 + .OO Copyright e 1986 Pergamon Journals Ltd BRIEF REPORT PERSONALITY CHARACTERISTICS OF RESTRAINED/BINGE EATERS VERSUS UNRESTRAINED/NONBINGE EATERS FREDERICA E. EDWARDS and DANIEL B. NAGELBERG Georgia Southern College, Statesboro, GA 30460 Abstract-The present study used the Restraint Scale (Herman & Polivy, 1975) and the Binge Scale (Hawkins & Clement, 1980) to identify two distinct groups of college students: those who scored high on both scales (“restrained/binge eaters”) and those who scored low on both scales (“unrestrained/nonbinge eaters”). Following subject identification, 26 restrained/binge females, 25 unrestrained/nonbinge females, 14 restrained/binge males, and 13 unrestrained/ nonbinge males were administered the Beck Depression Inventory (BDI; Beck, Ward, Mendelson, Mock & Erbaugh, l%l), the California Psychological Inventory (CPI; Gough, 1975), and a questionnaire on bulimic symptoms (based on the Diagnosfic and Statistical Munual of Mental Disorders, American Psychiatric Association, 1980) that allowed each sub- ject to be classified as bulimic, binge eater, or normal eater. All of the students identified as bulimic were from the restrained/binge group and that group contained twice as many binge eaters as the unrestrained/nonbinge group. Students in the restrained/binge group were found to be more depressed than students in the unrestrained/nonbinge group and less well adjusted in areas of socialization, maturity, responsibility, and intrapersonal structuring of values. Re- sults may be suggestive of personality factors tht predispose one to develop eating disturbances and/or useful in directing therapeutic intervention. In recent years, reports of binge eating and bulimia have begun to permeate the psycho- logical literature. One research issue involves the relationship between bulimic behaviors and personality variables. The present study used the Restraint Scale (Her- man & Polivy, 1975), a measure of the tendency to restrict one’s food intake, and the Binge Scale (Hawkins & Clement, 1980), a measure of binge eating behaviors, to iden- tify two distinct groups of college students: those who scored high on both scales (“re- strained/binge (R/B) eaters”) and those who scored low on both scales (“unrestrained/ nonbinge (unR/nonB) eaters”). The primary goal of this study was to compare per- sonality characteristics of these two groups of students. METHOD Subjects To obtain samples of “restrained/binge eaters” and “unrestrained/nonbinge eaters”, a slightly modified version of the 11-time Restraint Scale (RS) and the 9-item Binge Scale (BS) were administered to 112 female and 92 male students taking introductory psychology courses at a college in south Georgia in the spring of 1983. Order of presen- tation of the RS and BS was counterbalanced and found not to be significantly related to scores on either scale. The correlation between RS and BS scores was .78 (p < .05) for females and .57 (p < .05) for males. Within sex, students who scored in the upper This article is based on Frederica E. Edwards master’s thesis, which was conducted under the supervision of Daniel B. Nagelberg while at Georgia Southern College, Statesboro, Georgia. Inquiries or requests for reprints should be sent to Dr. Nagelberg at Savannah Psychological Consultants, P.C., One St. Joseph’s Professional Plaza, Savannah, Georgia 31419. 207

Upload: daniel-b

Post on 01-Jan-2017

226 views

Category:

Documents


6 download

TRANSCRIPT

AddictiveBehaviors, Vol. 11, pp. 207-211, 1986 Printed in the USA. All rights reserved.

0306-4603/86 $3.00 + .OO Copyright e 1986 Pergamon Journals Ltd

BRIEF REPORT

PERSONALITY CHARACTERISTICS OF RESTRAINED/BINGE EATERS VERSUS UNRESTRAINED/NONBINGE EATERS

FREDERICA E. EDWARDS and DANIEL B. NAGELBERG Georgia Southern College, Statesboro, GA 30460

Abstract-The present study used the Restraint Scale (Herman & Polivy, 1975) and the Binge Scale (Hawkins & Clement, 1980) to identify two distinct groups of college students: those who scored high on both scales (“restrained/binge eaters”) and those who scored low on both scales (“unrestrained/nonbinge eaters”). Following subject identification, 26 restrained/binge females, 25 unrestrained/nonbinge females, 14 restrained/binge males, and 13 unrestrained/ nonbinge males were administered the Beck Depression Inventory (BDI; Beck, Ward, Mendelson, Mock & Erbaugh, l%l), the California Psychological Inventory (CPI; Gough, 1975), and a questionnaire on bulimic symptoms (based on the Diagnosfic and Statistical Munual of Mental Disorders, American Psychiatric Association, 1980) that allowed each sub- ject to be classified as bulimic, binge eater, or normal eater. All of the students identified as bulimic were from the restrained/binge group and that group contained twice as many binge eaters as the unrestrained/nonbinge group. Students in the restrained/binge group were found to be more depressed than students in the unrestrained/nonbinge group and less well adjusted in areas of socialization, maturity, responsibility, and intrapersonal structuring of values. Re- sults may be suggestive of personality factors tht predispose one to develop eating disturbances and/or useful in directing therapeutic intervention.

In recent years, reports of binge eating and bulimia have begun to permeate the psycho- logical literature. One research issue involves the relationship between bulimic behaviors and personality variables. The present study used the Restraint Scale (Her- man & Polivy, 1975), a measure of the tendency to restrict one’s food intake, and the Binge Scale (Hawkins & Clement, 1980), a measure of binge eating behaviors, to iden- tify two distinct groups of college students: those who scored high on both scales (“re- strained/binge (R/B) eaters”) and those who scored low on both scales (“unrestrained/ nonbinge (unR/nonB) eaters”). The primary goal of this study was to compare per- sonality characteristics of these two groups of students.

METHOD

Subjects To obtain samples of “restrained/binge eaters” and “unrestrained/nonbinge eaters”,

a slightly modified version of the 11-time Restraint Scale (RS) and the 9-item Binge Scale (BS) were administered to 112 female and 92 male students taking introductory psychology courses at a college in south Georgia in the spring of 1983. Order of presen- tation of the RS and BS was counterbalanced and found not to be significantly related to scores on either scale. The correlation between RS and BS scores was .78 (p < .05) for females and .57 (p < .05) for males. Within sex, students who scored in the upper

This article is based on Frederica E. Edwards master’s thesis, which was conducted under the supervision of Daniel B. Nagelberg while at Georgia Southern College, Statesboro, Georgia.

Inquiries or requests for reprints should be sent to Dr. Nagelberg at Savannah Psychological Consultants, P.C., One St. Joseph’s Professional Plaza, Savannah, Georgia 31419.

207

208 F.E. EDWARDS and D.B. NAGELBERG

third of the distribution on both scales (n = 50) were identified as restrained/binge (R/B) eaters whereas students who scored in the lower third of the distribution on both scales (n = 43) were identified as unrestrained/nonbinge (unR/nonB) eaters. (An original attempt to form restrainedinonbinge and unrestrained/binge groups proved futile as so few students could be classified into either group). A total of 78 subjects agreed to participate further in this study and earned extra-credit points toward their psychology course grade. Of these subjects, 26 were R/B females (M RS = 20.15, SD = 3.57; MBS = 10.58, SD = 2.32), 25 were unR/nonB females (MRS = 4.92, SD = 2.24; MBS = .60, SD = .87), 14 were R/B males (MRS = 17.29, SD = 4.83; M BS = 8.50, SD = 2.68), and 13 were unR/nonB males (M RS = 4.00, SD = 2.04; M BS = 1.23, SD = 1 .Ol). The mean age of this sample was 19.4 years.

Procedure Approximately two weeks after group selection, subjects were contacted and ap-

pointments were scheduled. The California Psychological Inventory (CPI) (Gough, 1975), the Beck Depression Inventory (BDI) (Beck, Ward, Mendelson, Mock, & Er- baugh, 1961), and a questionnaire based on the DSM-III criteria of bulimia which allowed each subject to be classified as bulimic, binge eater, or normal eater were all administered in one session to small groups of students over a three-week period of time.

RESULTS

DSM-III Groups Classification of subjects into DSM-III groups (bulimic, binge eater, normal eater)

on the basis of their responses to the questionnaire on bulimic symptoms resulted in the identification of 12 bulimics (7 females, 5 males), 15 binge eaters (6 females, 9 males), and 51 normal eaters. A chi-square analysis indicated that the relationship between group (R/B, unR/nonB) and DSM-III classification was significant for both males, x2 (2, N = 27) = 9.75, p < .Ol, and females, x2(2, N = 51) = 9.33, p < .Ol. Almost 80% of the unR/nonB males and 92% of the unR/nonB females were identified as nor- mal eaters. All of the bulimics were from the R/B groups; 35.7% of R/B males and 26.9% of R/B females were identified as bulimic. For both males and females, twice as many individuals in the R/B groups compared with the unR/nonB groups were clas- sified as binge eaters.

Personality measures The means and standard deviations for the BDI and each of the CPI scales are

presented in Table 1 as a function of group (R/B eaters and unR/nonB eaters) and sex. As high intercorrelations among the CPI scales contraindicated the use of multivariate analyses, these data were analyzed in a series of analyses of covariance with two main factors (group and sex), one covariate (weight deviation) and one of 19 dependent vari- ables (BDI and CPI scores).

R/B subjects scored higher on the BDI than unR/nonB subjects, F (1, 73) = 6.9, p < .Ol , indicating that concern with weight and binge eating was related to depression scores. A chi-square analysis indicated that the relationship between group (R/B, unR/nonB) and category of depression (normal, mild, moderate, severe) was signifi- cant for females, x2 (2, N = 51) = 10.51, p < .Ol, but not for males, x2 (2, N = 27) = .96, p > .05. Whereas all 25 of the unR/nonB females scored in the normal range on the BDI, 9 of 26 R/B females (34.6%) scored in the mild or moderate range of depres- sion. The proportions of R/B and unR/nonB males who scored in the depressed range

Binge/restraint 209

Table 1. Means and Standard Deviations of BDI and CPI Scores, by Group and Sex

Males Females unR/nonB R/B unR/nonB R/B (?I = 13) (n = 14) (n = 25) (n = 26)

M SD M SD M SD M SD

BDI** 5.15 3.63 8.36 4.53 2.56 2.84 7.73 6.91

CPI Do CS

SY SP Sa** Wb** Re*** so** sc** To* Gi* Cm AC’ A i Ie

Py* Fx Fe

47.78 7.75 51.21 5.79 49.96 12.31 52.04 11.35 40.00 9.82 40.86 9.08 41.96 10.04 36.31 11.10 45.77 8.43 45.57 6.00 50.52 10.09 50.65 11.33 48.31 6.45 50.29 11.57 50.36 9.06 49.65 11.09 52.62 8.86 59.00 8.04 53.36 8.75 58.42 11.54 33.08 17.38 27.00 13.09 43.76 9.89 32.08 10.79 41.08 7.51 33.00 12.96 42.92 7.74 34.15 10.15 41.54 8.62 42.21 10.39 49.52 11.48 40.54 10.36 40.23 12.24 31.57 9.45 45.84 11.47 35.65 10.73 34.31 10.40 32.71 13.39 41.48 11.43 0.71 13.17 37.85 12.31 34.79 5.41 42.86 10.19 36.54 10.08 50.46 10.97 46.29 13.04 49.68 13.66 51.65 5.86 42.46 10.57 35.36 12.55 47.04 10.01 38.65 10.68 43.62 9.79 40.50 12.31 49.28 10.40 39.39 9.32 35.15 9.45 33.64 10.35 43.00 15.45 35.15 14.07 44.23 9.31 42.57 8.93 50.40 7.44 41.54 9.95 47.54 10.37 41.86 9.16 47.84 12.53 45.19 9.55 48.92 8.29 48.93 9.09 50.76 10.66 48.04 10.55

Note. CPI scale names are dominance (Do), capacity for status (Cs), sociability (Sy), social presence (Sp), self-acceptance @a), sense of well-being (Wb), responsibility (Re), socialization (So), self-control (SC), tolerance (To), good impression (Gi), communality (Cm), achievement via conformance (AC), achievement via independence (Ai), intellectual efficiency (le), psychological-mindedness (Py), flexibility (Fx), and femininity (Fe). Data were analyzed in a series of analyses of covariance with two factors (group and sex), one covariate (weight deviation), and one of 19 dependent variables (BDI and CPI scores). Asterisks indicate significant differences between the R/B and unR/nonB groups; *p < .05, **p < .Ol, ***, < .OOl.

were 28.6% and 23.1%, respectively. Group differences were also found on nine of the 18 CPI scales. R/B subjects scored lower (indicating poorer adjustment) than unR/nonB subjects on the following eight scales: sense of well-being (Wb), F( 1, 73) = 7.4, p < .Ol; responsibility (Re), F (1, 73) = 11.9, p < .OOl; socialization (So), F (1, 73) = 6.8, p < .Ol; self-control (SC), F(l, 73) = 10.3, p < .Ol; tolerance (To), F(l, 73) = 4.8, p < .05; good impression (Gi), F(l, 73) = 5.3, p < .05; achievement via conformance (AC), F (1, 73) = 5.5, p < .05, and psychological-mindedness (Py), F (1, 73) = 5.8, p < .05.

R/B subjects scored higher than unR/nonB subjects on one scale, self-acceptance (Sa), F (1, 73) = 6.2, p < .Ol. Sex was a significant main effect for only one scale, sense of well-being (Wb), with males scoring lower than females, F (1, 73) = 7.0, p < .Ol. None of the group by sex interactions was significant.

DISCUSSION

Present results indicate that, on the basis of standardized psychological tests, students who were identified as restrained/binge eaters displayed personality deficits relative to peers ‘Identified as unrestrainedjnonbinge eaters. Although mean depression (BDI) scores for both groups were within the normal range, 0.0% of the unR/nonB females, 34.6% of the R/B females, 23.1% of unR/nonB males, and 28.6% of the R/B

210 F.E. EDWARDS and D.B. NAGELBERG

males scored in the depressed range of the BDI. In interpreting these findings, it is im- portant to realize that approximately 25% of college students are suffering from depression at any one time (Beck & Young, 1978); Nagelberg, Pillsbury, & Balzer (1983) found that 28.9% of female and 16.5% of male “class attenders” scored in one of the depressed categories of the BDI. What is most striking about the present data is that none of the unR/nonB females were classified as depressed. This raises the in- teresting possibility that the absence of eating problems may be associated with a lowered risk of depression, at least in the case of college women.

On eight of the CPI scales, students in the R/B group scored lower, on the average, than students in the unR/nonB group. In examining Table 1, it can be seen that mean scores generally fall at or below the average score (T5J derived from the standardization sample. The fact that mean scores of unR/nonB eaters were generally lower than those obtained by college students in the standardization sample (Cough, 1975) suggests that group differences obtained in the present study were not due to unusually high control (unR/nonB) group scores.

Compared with students in the unR/nonB group, R/B students appeared less able to minimize their worries and complaints (sense of well-being), less conscientious, respon- sible, and dependable (responsibility), less mature socially (socialization), less able to self-regulate impulsive and self-centered behavior (self-control), less permissive, ac- cepting, and nonjudgmental (tolerance), less able to create a favorable impression and less concerned about how others react to them (good impression), less likely to achieve in situations where conformance is necessary (achievement via conformance), and less interested in and responsive to the inner needs, motives, and experiences of others (psy- chological-mindedness). In particular, R/B eaters scored lower than unR/nonB eaters on five of the six “class II” CPI scales, which measure socialization, maturity, respon- sibility, and intrapersonal structuring of values.

A somewhat inconsistent finding in the present study was that R/B students scored higher, on the average, than unR/nonB students on the self-acceptance scale. How- ever, high scorers on this scale are not always described in desirable terms; they tend to be intelligent, verbally fluent, and self-confident but outspoken, demanding, ag- gressive, and self-centered. Overall, results are consistent with previous studies that found personality deficits associated with eating difficulties. However, in the absence of premorbid personality data, it is not known whether such personality factors predis- pose one to develop dysfunctional eating patterns or if the personality deficits are secondary to the eating disorder. In particular, researchers have debated whether there is a causal link between depression and bulimia (see Gandour, 1984). Independent of cause and effect relationships, treatment may be directed toward clearly defined per- sonality factors that are associated with eating problems.

Because of the lower prevalence of eating disorders among men, they have often been excluded from studies involving binge eating and bulimia. In the present study, a sizeable number of male college students reported restrained and binge eating behaviors and of these, 35.7% fulfilled the DSM-III criteria for bulimia, Furthermore, the same pattern of personality deficits were found in this group as in the restrained/binge female students. These data at least suggest that it may be premature to not examine carefully eating related problems in male populations. It is possible, for example, that excessive exercise for weight control purposes, which is not a DSM-III “symptom” of bulimia, is more prevalent and acceptable in men than in women; a relationship be- tween “obligatory” running and eating disorders has already been suggested (Yates, Leehey, & Shisslak, 1983).

Binge/restraint 211

In summary, both male and female college students identified as binge/restrained eaters were found to have personality deficits relative to students who were classified as unrestrained/nonbinge eaters. Consistent personality differences between these two groups of students may be suggestive of predisposing personality factors involved in eating disorders and/or directions for therapeutic intervention.

REFERENCES

American Psychiatric Association (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, D.C.: Author.

Beck, A.T.,& Young, J.E. (1978, September). College blues. Psychology Today, pp. 85-86; 89-92. Beck. A.T.. Ward. C.H.. Mendelson, M., Mock, J., & Erbaugh. J. (1961). An inventory for measuring de-

pression. Archives of General Psychiatry, 4, 561-571. - Gandour, M.J. (1984). Bulimia: Clinical description, assessment, etiology and treatment. The International

Journal of Eating Disorders, 3, 3-38. Cough, H.G. (1975). Manual for the California Psychological Inventory (Rev. ed.). Palo Alto, California:

Consulting Psychologists Press. Hawks, II, R.C., & Clement, P. (1980). Development and construct validation of a self-report measure of

binge eating tendencies. Addictive Behaviors, 5, 219-226. Herman, C.P., & Polivy, J. (1975). Anxiety, restraint, and eating behavior. Journal ofAbnormal Psychol-

ogy, 84, 666-672. Nagelberg, D.B., Pillsbury, E.C., & Balzer, D.M. (1983). The prevalence of depression as a function of

gender and facility usage in college students. Journal of College Student Personnel, 24, 525-529. Yates, A., Leehey, K., & Shisslak, C.M. (1983). Running-an analogue of anorexia? New England Journal

of Medicine, 308, 25 l-255.