Comparison of eating disorders and body image disturbances between Eastern and Western countries

Vol. 13: 73-80, June 2008

73

ORIGINALRESEARCH

PAPER

Key words: Tibet, Eating Attitudes Test(EAT), Figure Rating Scale(FRS), cross-cultural,ethnicity, socio-economicstatus (SES), body imagedistortion, media. Correspondence to:Marci E. Gluck, PhD,NIH/NIDDK/ODCRS, 4212North 16th Street, Room541, Phoenix, AZ 85016,USA E-mail: [email protected]: June 2, 2007Accepted: November 28,2007

Comparison of eating disorders andbody image disturbances betweenEastern and Western countries

INTRODUCTION

There are reports of higher rates of eat-ing disorders in industrialized Westernizednations (1-3), even among the ethnicminorities in those countries (4-8) than inless developed countries. Developing coun-tries, however, are experiencing a rise ineating disorders as they become moreindustrialized. Industrialization is associat-ed with a larger proportion of wealthier (9)and better educated individuals, who tendto have more exposure to media images ofthe idealized female body shape (2, 10, 11).Eating disorders have been reportedaround the world including France (12),India (13-15), Egypt (16), the Middle East(17, 18) and Korea (19). Tsai (20) reviewseating disorders in the Far East, but there

are no reports on Tibet, nor are there directcomparisons with the West.

The literature is inconsistent on whetheror not higher SES individuals in the US orother Westernized countries are more likelyto have eating disorders (21-23). In develop-ing cultures, the pattern is more evident. Ina Palestinian West Bank population, Steneet al. (24) found that higher SES individualswere more likely to suffer from eating dis-orders. Al-Subaie (25) found that youngSaudi girls who had high scores on theEDI- Drive for Thinness subscale were froma higher SES family than their classmates.Chadda et al. (26) reported a similar patternin girls living in India.

Images of idealized physical perfectionportrayed by the media in technologicallyadvanced countries may help induce feel-

B. Rubin*, M.E. Gluck**, C.M. Knoll*, M. Lorence*, and A. Geliebter*

*New York Obesity Research Center, Departments of Medicine and Psychiatry, St. Luke’s/Roosevelt HospitalCenter, Columbia University, College of Physicians and Surgeons, New York, and **Obesity and DiabetesClinical Research Section, Phoenix Epidemiology and Clinical Research Branch, National Institute ofDiabetes and Digestive and Kidney Diseases, National Institutes of Health, Phoenix, USA

ABSTRACT. Factors associated with the development of eating disorders in countries withnon-Western cultures have not been adequately investigated in relation to Westernizedcountries. We therefore studied 243 girls [age =16.5±1.2 (SD)], recruited from schools inIndia, Tibet, the US and France. They completed the Figure Rating Scale (FRS), the EatingAttitudes Test (EAT), and the Beck Depression Inventory (BDI). The Tibetan group had alower body mass index (BMI) than the other groups (p<0.0001), which did not differ fromeach other. All groups differed significantly on socio-economic status (SES), with those livingin India having the highest (p<0.0001). Prior to controlling for age, SES, and BMI, there wereno significant differences on any psychological measure between the individual countries, orwhen collapsed by East vs. West. However, after controlling for the same covariates, theTibetan group selected a significantly larger current (p<0.0001) and ideal body size (p=0.03),compared to all the other countries, and had more body image discrepancy than theAmerican group (p=0.04). After controlling only for BMI, the girls from the East had a largercurrent and ideal, but no difference on body image discrepancy. Body image discrepancyscores were best predicted by EAT scores and BMI, accounting for 35% of the variance(p<0.0001). EAT scores themselves were best predicted by mother’s education, BDI, bodyimage discrepancy, and drug and tobacco use, accounting for 33% of the variance(p<0.0001). Unlike some other studies, we did not observe greater body image discrepancyand eating pathology in Western cultures, whether or not controlling for age, SES, and BMI.There were no differences in eating and depression pathology between those in the US,France, or India. Indeed, the Tibetans, after controlling for their low BMI and SES, had thegreatest body image discrepancy. (Eating Weight Disord. 13: 73-80, 2008). ©2008, Editrice Kurtis

B. Rubin, M.E. Gluck, C.M. Knoll, et al.

ings of body dissatisfaction and increase therisk for developing eating disorders (27, 28). Inless developed nations, those women who aremore exposed to this ideal have begun to adoptharmful dieting behaviors (29). Additionally,recent immigrants in Westernized countriesoften adopt similar physical ideals (30) to thosealready living there, and some develop an eventhinner ideal body type (31, 32).

Within technologically advanced countries,there may still be cultural and racial differencesin body image and eating pathology. Gluck etal. (33) found that in the US, Caucasians hadgreater body image discrepancy than Asians,and more eating pathology than both Asiansand African-Americans. This became apparentonly after controlling for BMI, a statistical con-trol that has not been applied consistently inprevious studies (34). Although body image dis-tortion is widespread, women in “insulated”cultural religious groups, such as OrthodoxJews, are less likely to develop eating patholo-gy in part because less emphasis is placed onappearance (33). Additionally, eating disordersmay result from deep-seated individual psycho-logical issues, such as depression (35), andbeing religious may protect against depression(33).

Researchers have proposed that educationlevel and prevalence of eating disorders arepositively related (36). Silverstein et al. (35) sug-gest that women who surpass their mothersacademically and professionally are especiallylikely to develop eating pathology, includingbody image distortion. This may be associatedwith their greater drive for success, and theincreasing social demands experienced by suc-cessful women. Fisher et al. (37) observed thatindividuals scoring higher on the EatingAttitudes Test (EAT) were more likely to smokecigarettes and engage in alcohol and drug use.Although certain risk factors in Western cul-ture have been linked with the development ofeating pathology, it is not clear whether or notthese risk factors are related to body-imagedistortion or eating pathology in other coun-tries and cultures.

We therefore examined eating disorder psy-chopathology, body image discrepancy,depression, and potentially related variables,such as religiosity, empowerment, and per-ceived barriers to success. Middle and highschool female students were studied in twoWestern countries, United States and France,and in two Eastern countries, India, and Tibet.We expected that: 1) US students would havethe highest scores on the Eating Attitudes Test(EAT), endorse a thinner ideal body shape, andhave greater body image discrepancy followed

in order from most to least Westernized:France, India and Tibet. In all countries, wehypothesized that: 2) EAT scores and bodyimage discrepancy would be related to levels ofreligiosity, media exposure, empowerment,drug and tobacco use, and perceived barriersto success, and 3) parents’ level of educationwould be positively correlated with EAT anddepression scores.

MATERIALS AND METHODS

ParticipantsParticipants were 268 female students

recruited from middle and high schools in sev-eral countries. Ninety were from the USA andattended one public (n=62) and one privatehigh school (n=28) in New York City, 38 wereFrench high school girls in Paris, 40 wereIndian girls attending school in New Delhi and60 were Tibetan girls attending the central andmiddle schools in Lhasa.

Several groups (total=40) were excludedfrom analysis: students from a Tibetanorphanage (n=10), Indian students living inTibet (n=15), and Tibetan refugees living in theUnited States (n=15) because a majority inthese groups had missing data. The remaining228 females were 16.5 y.o. ±1.2 SD. Althoughthe differences in age were narrow, they weresignificant by country (F [3,227] =4.3, p=0.006),with those in India being younger than Tibetand France but not than those in the US.Given that the mean age was 16.5, when mostgrowth in females has ended, we used BMIrather than BMI percentile for age. Based ontheir reported current height and weight, BMIranged from 13.9-31.7 (19.8±2.8) and also dif-fered significantly across countries (F [3,211]=9.2, p<0.0001). Tibetans had the lowest BMIand were more likely than the other groups tobe underweight (χ2=31.1, p<0.0001), defined asBMI <18.5. Of the total sample, 72 (31%) wereunderweight (BMI <18.5), 129 (57%) were nor-mal weight (BMI >18.5 <25), 9 (4%) were over-weight (BMI >25), and 1 (0.4%) was obese(BMI >30). Eighteen subjects (7%) did notreport their height and weight. Participantsindicated their mother and father’s level ofeducation [which represents one factor of theHollingshead two-factor socioeconomic sta-tus (SES) index], which was averaged to esti-mate SES. All the groups differed from eachother on SES [F (3,221)=48.9, p<0.0001].Surprisingly, the girls in India had the highestSES (and may not be representative of mostgirls living there), and girls in Tibet had thelowest SES (Table 1).

74 Eating Weight Disord., Vol. 13: N. 2 - 2008

Eating disorders and body image in East and West

ProceduresCandidates were invited to participate during

one of their classroom sessions. All question-naires were translated into French and Tibetanby native speakers and were approved by theappropriate Institutional Review Boards. Thedistribution of the surveys was facilitated bynative residents in countries outside the UnitedStates. The same researcher (BR) traveled toand was present in all the countries exceptIndia, where the surveys were distributed by anative assistant. Students were invited by theresearcher or assistant to participate in a studyabout body image and eating habits. Studentswere assured that participation was voluntaryand that the survey was completely anony-mous. Questionnaires were distributed and col-lected during the same classroom session whilethe investigator or assistant was present. Allwere distributed and collected during a 2-month period.

Instruments Disordered eating attitudes and behaviors

were evaluated with the Eating Attitudes Test(EAT), a widely used and validated self-reportmeasure assessing the presence of eating dis-orders. A score >19 indicates a potential riskfor the development of an eating disorder(38).

The Figure Rating Scale [FRS (39)] was usedto assess perceptions of body image and bodydissatisfaction. The FRS contains nine femaleand nine male silhouettes. The participantswere asked to pick the female silhouette thatbest represented her current appearance(Current) and the silhouette she would mostlike to look like (Ideal). Ideal body image wassubtracted from the current appearance todetermine body image discrepancy.

The Beck Depression Scale [BDI (40, 41)] is awidely used and validated self-report measure,which assesses specific symptoms of depres-sion. A score <9 indicates no presence of

depression, 9-18 mild depression, 19-29 moder-ate depression, and >29 severe depression.

Additional questions were composed by theauthors to assess religiosity (33), empower-ment, drug and alcohol use, media exposure,and perceived barriers to success. Each ques-tion was scored on a 5-point Likert scale (either“strongly disagree,” “moderately disagree,”“neither agree nor disagree,” “moderatelyagree,” “strongly agree,”; or “never,” “occa-sionally,” “sometimes,” “often,” or “all thetime”). The responses to these questions weregrouped together and summed to provide sub-scale scores: “religiosity” comprised question10 (I practice my religion regularly), question11 (People in my community belong to thesame religion as I do), question 12 (If someonesaid something bad about my religion, I wouldalmost feel as if they had said something badabout me), question 13 (My religious group isimportant to me), question 14 (I feel proudabout my religious group), and question 15 (Iagree in principle with the lifestyle my religiondictates regarding modes of dress and/or socialinteractions with members of the opposite sex)(33). The “empowerment” score was based onQuestion 19 (men and women ought to betreated equally), question 20 (It is appropriatefor a woman to work outside the home), andquestion 23 (It is important that I have profes-sional success). The “perceived barriers to suc-cess” scale was composed of question 21(Women worry more about their appearancesthan men do), question 22 (Being a woman canhinder professional success), question 24(Attractiveness is very important if you want toget ahead in our culture), and question 25 (Inmy culture, it is easier for a man to have pro-fessional success than a woman). “Drug andtobacco use” scale comprised questions 4 (Howoften do you use tobacco?) and 5 (How often doyou use alcohol, marijuana or other leisuredrugs?). A “media” subscale was composed ofquestion 6 (How frequently do you use theInternet?), question 7 (How frequently do youwatch television?), and question 8 (Do you readfashion magazines?).

Additional questions concerned the highestlevel of education achieved by parents, used toestimate SES, and the level of education theparticipant planned to achieve.

Data analysisData were analyzed using ANOVA with coun-

try as the independent variable to test for effectson the psychological scales. Because BMI, SESand age differed significantly between the coun-tries (Table 1), they were entered as covariates insecondary analyses (ANCOVA). Additionally, we

75Eating Weight Disord., Vol. 13: N. 2 - 2008

US France Tibet India p-value*

n 90 38 60 40

Age 16.4±1.7a 17.0±0.7a,b 16.6±1.4c 16.1±1.0b,c 0.006

BMI 20.5±2.6a 20.3±2.3b 18.2±2.1a,b,c 19.8±3.7c <0.0001

SES 4.3±1.5a,b,c,d 2.8±2.1a,b,c,d 1.9±2.0a,b,c,d 5.7±.62a,b,c,d <0.0001

*Overall comparison among the four countries a,b,c,dAcross rows, groups with similar postscripts indicate significant differences bypost-hoc tests.

TABLE 1Participant characteristics.


collapsed countries into Western (US andFrance) and Eastern (India and Tibet). We con-trolled for BMI only in this analysis, because itwas higher in the West than East (F=19.7,p<0.000). LSD post-hoc tests were performedwhen the overall F was significant. Partial corre-lation coefficients were used to describe relation-ships between continuous variables when con-trolling for BMI, SES, and age. Spearman’s rhowas used for correlations between education andcontinuous variables, and chi-square was usedfor categorical data, i.e., level of parent educationby country. Results are presented as M±SD.Two-tailed p<0.05 was considered significant.Data were analyzed with the Statistical Packagefor the Social Sciences (SPSS version 14.0.1,2005, Chicago, IL).

RESULTSPerception of body size

There were no significant differencesbetween the four countries on the FRS for cur-rent appearance (F [3,205] =0.04, p=0.99, n.s.),ideal appearance (F [3,205] =1.6, p=0.19, n.s.),or body image discrepancy (F [3,205] =0.65, p=0.58, n.s.). However, after controlling forage, BMI, and SES, the Tibetans had the largestcurrent appearance, which differed from all theother 3 countries (F [6,193] =7.8, p<0.0001). TheTibetans now also endorsed a significantlylarger ideal appearance than the US andFrance, but not India (F [1,203] =3.1, p=0.03).Additionally, they now had significantly morebody image discrepancy compared to the US (F [6,201] =2.8, p=0.04). (Fig. 1). After control-ling for BMI, the girls from the East had a larg-er current (F [2,198] =11.0, p=0.001) and idealappearance (F [2,198] =6.3, p=0.01) than thosefrom the West, but no difference in discrepan-cy (F [2,198] =1.6, p=0.21).

Eating disorders and depression Neither the Tibetan Middle School nor

Tibetan Central School filled out the EAT or theBDI and were thus excluded from these analy-ses. The mean EAT score was 10.5 (±10.7), andthere were no differences between countrieseither before (F [2,167] =0.78, p=0.46, n.s.) orafter controlling for age, BMI and SES (F[5,154] =0.52, p=0.59, n.s.). The risk for develop-ing an eating disorder (score >19) also did notdiffer between countries (χ2 =2.1, p<0.37, n.s.).The mean BDI score was 21.2 (±7.2) and alsodid not differ between countries either before(F [2,163] =0.43, p=0.65, n.s.), or after enteringthe covariates (F [5,151] =1.1, p=0.35, n.s.) (Fig.2). There were no differences between theWestern and Eastern countries on EAT (F

[2,156] =0.001, p=0.98, n.s.) or depressionscores (F [2,153] =1.8, p=0.18, n.s.).

Risk factor subscales We conducted the analyses for the eating dis-

order risk factor subscales for religion, mediaexposure, drug and tobacco, barriers to suc-cess, and empowerment, both with and withoutcontrolling for age, BMI or SES, and the resultswere similar. Thus, only results from thecovariate analyses are presented. Girls fromTibet and India reported more religiosity thantheir US counterparts (F [6,169] =4.0, p=0.009),and girls from Tibet reported significantly lessmedia exposure than all three other groups (F[8,215] =38.2, p<0.0001). Drug and tobacco usein the American and French girls was signifi-cantly higher than in both India and Tibet, and


40

35

30

25

20

15

10

5

0US France Tibet India

BodyDiscrepancy

AGE

SES

BMI

Current

Current adjustedIdeal

Ideal adjusted

EAT BDI

Sco

re

25

20

15

10

5

0

USAFranceIndia

FIGURE 2There were no significant differences between countries on EAT

(F [5,154] =0.52, p=0.59, ns) or BDI (F [5,151] =1.1, p=0.35, ns)scores.

FIGURE 1For the actual scores, there were no significant differences between thecountries on the FRS for current, ideal or body image discrepancy. Afteradjusting for age, BMI, and SES, the Tibetans had the largest currentand ideal appearances and had more body discrepancy than the US.


significantly higher in the French compared toAmerican girls (F [8,216] =11.8, p<0.0001).Perceived barriers to success differed signifi-cantly between the countries (F [6,203] =5.8,p=0.001) with American girls scoring higherthan Tibetan and Indian girls, and French girlshigher than Tibetan girls. Levels of empower-ment did not differ by country (F [3,222] =1.8, p=0.16). Girls in the East had significantly higherreligiosity scores than those in the West (F [1, 164] =11.6, p=0.001). Girls in the West hadsignificantly more media exposure (F [2,211]=35.9, p<0.0001), drugs and tobacco use (F [2, 211] =28.4, p<0.0001) and perceived barriersto success (F [2,210] =19.2, p<0.0001) than thosein the East, but there were no differences onempowerment (F [2,206] =2.4, p<0.12, n.s.).

Education There were small but significant differences

in the level of education that the participantsplanned to obtain (F [6,234] =4.7, p<0.0001).Girls from India planned to obtain the mosteducation, followed by US, France and thenTibet, each significantly different from eachother. However, these differences were not sig-nificant after controlling for SES. Without con-trolling for any variables, all groups differedsignificantly in the amount of education theirmothers (F [4,205] =41.8, p<0.0001) and fathers(F [4,207] =34.0, p<0.0001) had achieved.Parents from India had the highest, followed bythe Americans, French and Tibetans, respec-tively, indicative of the differences in SES

between the groups. When comparing West vs.East, there were no differences in mother’s orfather’s level of education.

CorrelationsPartial correlations controlling for age, SES

and BMI are presented in Table 2. As hypothe-sized, the EAT, BDI and body image discrepan-cy scores were significantly correlated witheach other. EAT, but not body image discrep-ancy, was also correlated with drug and tobac-co use, but contrary to our hypothesis, neitherEAT nor body image discrepancy scores werecorrelated with religiosity, media exposure,empowerment or perceived barriers to success.Mother’s but not father’s level of education (r=0.21, p=0.004) was positively correlated withEAT scores, but neither was correlated withBDI scores or body discrepancy (Table 2).

To examine predictors of both body dissatis-faction and EAT scores, we used simultaneousmultiple linear regression (MLR) and enteredthe variables that were significantly correlatedwith each, as well as country and BMI. ForEAT scores, mother’s education (p=0.005), BDI(p=0.001), body image discrepancy (p<0.0001)and drugs and tobacco (p=0.04) accounted for33% of the variance (p<0.0001), while BMI,country, and ideal body shape were not signifi-cant predictors. EAT scores (p<0.0001) andBMI (p<0.0001) significantly predicted bodyimage discrepancy, accounting for 35% of thevariance (p<0.0001). BDI and country were notsignificant predictors.


TABLE 2Partial correlations controlling for age, SES, and BMI.

EAT Current Ideal Discrepancy BDI Drug Use Media Religiosity Perc. Empower- Father Mother Barriers ment Ed Ed

EAT 0.21* -0.33*** 0.47*** 0.31** 0.20* 0.14 0.05 -0.01 -0.07 0.13 0.21**

Current 0.45*** 0.63*** 0.15 -0.09 -0.19* 0.13 -0.002 -0.10 -0.06 0.04

Ideal -0.41*** -0.05 -0.12 -0.21** -0.01 -0.05 -0.08 -0.09 -0.06

Discrepancy 0.18* 0.01 -0.02 0.14 0.04 -0.04 -0.03 0.04

BDI 0.14 0.13 0.06 0.30** -0.06 -0.08 -0.05

Drug Use 0.25** -0.30*** 0.14 -0.18* 0.10 0.14*

Media 0.14 0.47*** -0.10 0.31*** 0.40***

Religiosity 0.18* -0.07 0.14 0.07

Perc. Barriers 0.02 0.02 0.10

Empowerment 0.04 0.08

Father Ed 0.72***

*p≤0.05, **p≤0.01, ***p≤0.0001: EAT (Eating Attitudes Test); BDI (Beck Depression Inventory); Current (FRS: current appearance); Ideal (FRS: ideal appearance);Discrepancy (FRS: current – ideal); Drug Use (drug and tobacco use); Perc. Barriers (perceived barriers to success); Father Ed (Father’s highest levels of educa-tion); Mother Ed (Mother’s highest level of education).


DISCUSSION

In contrast to our main hypothesis, after con-trolling for age, SES, and BMI, the girls fromthe US had less body image discrepancy thangirls living in Tibet. The Tibetans, who had thelowest actual BMI and SES, selected the largestcurrent appearance and considered the largestideal to be the most attractive. In addition, theyhad the greatest body image discrepancy; theywanted to be smaller than their actual currentappearance. Although a distorted body imageis commonly viewed as being pathological, it ispossible that these girls selected a larger idealbody size because of the positive relationshipbetween high social class and larger bodyobserved in many non-Western societies (42).They also had the least exposure to media, andmedia was negatively correlated with the idealbody size. Thus, their endorsement of a largerideal could be a reflection of their decreasedexposure to the Western “thin ideal (10).”

Also contrary to our hypothesis, girls in theUnited States did not differ in eating disorderpathology and depression from girls in Franceand India, both before and after controlling forage, SES, and BMI. It appears then, that girlsfrom the high SES area in India are equally aslikely to exhibit eating psychopathologyregardless of their non-Western location. Thisis in line with previous studies that observedhigher rates of eating disorders in individualswith higher SES living in developing countries(43, 44). It also supports a study which foundthat girls from Canada and India scored simi-larly on core features of eating disorders aftercontrolling for BMI, although SES was notassessed in that study (45). It is somewhat sur-prising though, that the girls from India did notreport more psychopathology than the USgroup, given their higher SES. Perhaps there isa ceiling effect for the highest SES groups.

Unlike another recent study (33), religiositywas not significantly correlated with eatingpathology, depression, or body image.Religiosity was higher in India than in the USor France, yet they still had similar eating dis-order and depression scores. However, com-pared to Gluck et al.’s study (33), which focusedon an extremely religious insulated group, it islikely that most girls living in India share simi-lar levels of religiosity. Thus, although they dif-fered from their US counterparts, they mightnot differ from others in their own country.Moreover, they had rates of media exposurethat were similar to the other Western groups,rather different than other insulated religiousgroups.

Drug and tobacco use, which was lower in

India, was significantly correlated with eatingpathology. Girls from the US perceived feelingthe most barriers to success, yet this was onlyrelated to depression scores, and not to EDI orbody image scores. There were no differencesin empowerment after controlling for BMI andSES, and it was not correlated with any vari-able. Thus, SES appears to play a large role ineating disordered pathology, given that thosein India, with the highest SES, still had equallevels of eating pathology and depression, inspite of having lower rates of risk factors suchas drug and tobacco use and higher levels ofreligiosity. Nevertheless, risk factors such asdrug and tobacco use and media exposure didnot differ whether or not we controlled for BMIand SES, and appear to be more related to thecountry. Conversely, observed differences oneating pathology and body image appear to bemore influenced by BMI and SES.

In our study sample, high educational statusin the mothers, although not fathers, predictedeating pathology, and neither was correlatedwith depression. We observed a positive corre-lation between mother’s educational successesand EAT scores, consistent with the literaturethat highlights family values of achievementand competitiveness as being important factorsin the development of eating disorders (46).Surprisingly, however, media exposure, drugand alcohol use, and country did not predicteating disordered or body image pathology.

This study has several limitations. Althoughthere was complete data in all countries forbody image, there was no available data on eat-ing pathology and depression for the Tibetangirls who had the lowest SES and the lowestBMI. As we were interested primarily in exam-ining cultural differences between countries,we did not ascertain race. In the US, studieshave demonstrated that Asians have less eatingpathology and lower BMI (47), but we wereunable to assess this in the current study.Additionally, with limited sample sizes, not nec-essarily representative of the four differentcountries, the data should be considered withcaution. Lastly, our study did not assess bingeeating disorder and/or loss of control over eat-ing, topics which would be worthy to examinein future studies.

This study demonstrated rates of eatingpathology that were similar in the US, Franceand India. These results are somewhat incon-sistent with other findings that eating disordersare more prevalent among higher SES groupsand more educated people. Instead, our find-ings revealed that the higher levels of religiosi-ty, as observed in India, might be a protectivefactor that overrides the influence of SES.




REFERENCES

1. McCarthy M.: The thin ideal, depression and eatingdisorders in women. Behav. Res. Ther., 28, 205-215,1990.

2. Bardwell Mumford D., Choudry I.Y.: Body dissatisfac-tion and eating attitudes in slimming and fitness gymsin London and Lahore: A cross-cultural study. Eur.Eat. Disord. Rev., 8, 224, 2000.

3. Pritts S.D., Susman J.: Diagnosis of eating disorders inprimary care. Am. Fam. Physician, 67, 297-304, 2003.

4. Mumford D.B.: Eating disorders in different cultures.Int. Rev. Psychiatry, 5, 109-113, 1993.

5. Pritts S.D., Susman J.: Diagnosis of eating disorders inprimary care. Am. Fam. Physician, 67, 297-304, 2003.

6. French S.A., Story M., Neumark-Sztainer D., DownesB., Resnick M., Blum R.: Ethnic differences in psy-chosocial and health behavior correlates of dieting,purging, and binge eating in a population-based sam-ple of adolescent females. Int. J. Eat. Disord., 22, 315-322, 1997.

7. Miller M.N., Pumariega A.J.: Culture and eating disor-ders: a historical and cross-cultural review. Psychiatry,64, 93-110, 2001.

8. le Grange D., Stone A.A., Brownell K.D.: Eating distur-bances in White and minority female dieters. Int. J.Eat. Disord., 24,-395-403, 1998.

9. Striegel-Moore R.H., Silberstein L.R., Rodin J.: Towardan understanding of risk factors for bulimia. Am.Psychol., 41, 246-263, 1986.

10. Hargreaves D., Tiggemann M.: The effect of televisioncommercials on mood and body dissatisfaction: therole of appearance-schema activation. J. Soc. Clin.Psychol., 21, 287-308, 2002.

11. Stice E., Schupak-Neuberg E., Shaw H.E., Stein R.I.:Relation of media exposure to eating disorder sympto-matology: an examination of mediating mechanisms. J.Abnorm. Psychol., 103, 836-840, 1994.

12. Tordjman S., Zittoun C., Anderson G.M., Flament M.,Jeammet P.: Preliminary study of eating disordersamong French female adolescents and young adults.Int. J. Eat. Disord., 16, 301-305, 1994.

13. Khandelwal S.K., Sharan P., Saxena S.: Eating disor-ders: an Indian perspective. Int. J. Soc. Psychiatry,41,132-146, 1995.

14. Srinivasan T.N., Suresh T.R., Jayaram V.: Emergenceof eating disorders in India: Study of eating distresssyndrome and development of a screening question-naire. Int. J. Soc. Psychiatry, 44, 198, 1998.

15. Sjostedt J.P., Schumaker J.F., Nathawat S.S.: Eatingdisorders among Indian and Australian university stu-dents. J. Soc. Psychol., 138, 351-357, 1998.

16. Jackson R.T., Rashed M., Saad-Eldin R.: Rural urbandifferences in weight, body image, and dieting behav-ior among adolescent Egyptian schoolgirls. Int. J. FoodSci. Nutr., 54, 1-11, 2003.

17. Stene L.C., Giacaman R., bdul-Rahim H., Husseini A.,Norum K.R., Holmboe-Ottesen G.: Food consumptionpatterns in a Palestinian West Bank population. Eur. J.Clin. Nutr., 53, 953-958, 1999.

18. Al-Subaie A.S.: Some correlates of dieting behavior inSaudi schoolgirls. Int. J. Eat. Disord., 28, 242-246, 2000.

19. Ko C., Cohen H.: Intraethnic comparison of eating atti-tudes in native Koreans and Korean Americans using aKorean translation of the eating attitudes test. J. Nerv.Ment. Dis., 186, 631-636, 1998.

20. Tsai G.: Eating disorders in the Far East. Eat. WeightDisord., 5, 183-197, 2000.


22. Snooks M.K., Hall S.K.: Relationship of body size, bodyimage, and self-esteem in African American, EuropeanAmerican, and Mexican American middle-classwomen. Health Care Women Int., 23, 460-466, 2002.

23. Story M., French S.A., Resnick M.D., Blum R.W.:Ethnic/racial and socioeconomic differences in dietingbehaviors and body image perceptions in adolescents.Int. J. Eat. Disord., 18, 173-179, 1995.

24. Stene L.C., Giacaman R., bdul-Rahim H., Husseini A.,Norum K.R., Holmboe-Ottesen G.: Food consumptionpatterns in a Palestinian West Bank population. Eur. J.Clin. Nutr., 53, 953-958, 1999.


26. Chadda R., Malhotra S., Asad A.G., Bambery P.: Socio-cultural factors in anorexia nervosa. Indian J.Psychiatry, 29, 111, 1987.

27. French S.A., Story M., Neumark-Sztainer D., DownesB., Resnick M., Blum R.: Ethnic differences in psy-chosocial and health behavior correlates of dieting,purging, and binge eating in a population-based sam-ple of adolescent females. Int. J. Eat. Disord., 22, 315-322, 1997.

28. Striegel-Moore R.H., Silberstein L.R., Rodin J.: Towardan understanding of risk factors for bulimia. Am.Psychol., 41, 246-263, 1986.

29. Furnham A., Alibhai N.: Cross-cultural differences inthe perception of female body shapes. Psychol. Med.,13, 829-837, 1983.


31. Ko C., Cohen H.: Intraethnic comparison of eating atti-tudes in native Koreans and Korean Americans using aKorean translation of the eating attitudes test. J. Nerv.Ment. Dis., 186, 631-636, 1998.


33. Gluck M.E., Geliebter A.: Body image and eatingbehaviors in Orthodox and Secular Jewish women. J.Gend. Specif. Med., 5, 19-24, 2002.

34. Akan G.E., Grilo C.M.: Sociocultural influences on eat-ing attitudes and behaviors, body image, and psycho-logical functioning: a comparison of African-American, Asian-American, and Caucasian collegewomen. Int. J. Eat. Disord., 18, 181-187, 1995.

35. Silverstein B., Perlick D.: The cost of competence: Whyinequality causes depression, eating disorders, and ill-ness in women. New York, Oxford University Press,1995.

36. Brumberg J.J.: Fasting girls: The emergence of anorex-ia nervosa as a modern disease. Cambridge, MA,Harvard University Press, 1988.

37. Fisher M., Schneider M., Pegler C., Napolitano B.:Eating attitudes, health-risk behaviors, self-esteem andanxiety among adolescent females in a suburban highschool. J. Adolesc. Health, 12, 377-384, 1991.

38. Garner D.M., Garfinkel P.E.: The Eating Attitudes Test:an index of the symptoms of anorexia nervosa.Psychol. Med., 9, 273-279, 1979.

39. Stunkard A.J., Sorensen T., Schulsinger F.: Use of theDanish adoption register for the study of obesity and



thinness. Res. Publ. Assoc. Res. Nerv. Ment. Dis. 60,115-120, 1983.

40. Beck A.T., Rush A., Shaw B., Emery G.: Cognitive ther-apy of depression. New York, Guilford Press, 1979.

41. Beck A.T., Ward C.H., Mendelson M., Mock J.,Erbaugh J.: An inventory for measuring depression.Arch. Gen. Psychiatry, 4, 561-571, 1961.


43. Al-Subaie A.S.: Some correlates of dieting behaviorin Saudi schoolgirls. Int. J. Eat. Disord., 28, 242-246,2000.

44. Snooks M.K., Hall S.K.: Relationship of body size, body

image, and self-esteem in African American, EuropeanAmerican, and Mexican American middle-classwomen. Health Care Women Int., 23, 460-466, 2002.

45. Gupta M.A., Chaturvedi S.K., Chandarana P.C.,Johnson A.M.: Weight-related body image concernsamong 18-24-year-old women in Canada and India: anempirical comparative study. J. Psychosom. Res., 50,193-198, 2001.

46. Ogden J., Chanana A.: Explaining the effect of ethnicgroup on weight concern: finding a role for family val-ues. Int. J. Obes. Relat. Metab. Disord., 22, 641-647,1998.

47. Gluck M.E., Geliebter A.: Racial/ethnic differences inbody image and eating behaviors. Eat. Behav., 3, 143-151, 2002.

Documents

Comparison of eating disorders and body image disturbances between Eastern and Western countries