European Eating Disorders ReviewEur. Eat. Disorders Rev. 14, 5465 (2006)
Eating and Body Image DisturbancesAcross Cultures: A Review
Nerissa L. Soh1*, Stephen W. Touyz1,2
and Lois J. Surgenor31Discipline of Psychological Medicine, University of Sydney, Australia2School of Psychology, University of Sydney, Australia3Department of Psychological Medicine, University of Otago, New Zealand
Objective: To critically review the literature examining theimpact of acculturation, socio-economic status, family functioningand psychological control in relation to eating and body imagedisturbances across cultures.Method: A review of the literature on eating disorders, eatingand body image disturbances, psychological control, body compo-sition, socio-economic status and family functioning on differentcultural groups.Results: Of the empirical studies undertaken, few investigated acultural groups eating pathology in both its country of origin anda Western country using the same methodology. To date, theresearch findings are mixed and it is still unclear if the presen-tation of an eating disorder differs across cultures. Acculturationhas not been consistently taken into consideration and psycholo-gical control has not been examined in relation to eating distur-bances in non-Western groups.Discussion: This review focuses on some of the methodologicallimitations of previous research and attempts to delineate thesalient issues which warrant further scientific enquiry. Copyright# 2006 John Wiley & Sons, Ltd and Eating Disorders Association.
Keywords: body image; cross-cultural; eating disorders; psychological control
Although eating disorders, and particularly anor-exia nervosa (AN), have long been identified in Wes-tern societies (Vandereycken & van Deth, 1994), inthe non-Western world such disorders have onlybeen described in the international literature sincethe late 1970s. In explanation, it is often argued thateating disorders are Western culture-bound syn-dromes and that individuals from non-Westernsocieties have some immunity to such disorders
(Gordon, Perez, & Joiner, 2002; Lai, 2000; Nasser,1997; Timimi & Adams, 1996; Tsai, 2000; Wildes,Emergy, & Simons, 2001).
The stereotypical AN patient is depicted as young,North-European Caucasian, female, well educatedand from the upper socio-economic class. Conven-tional wisdom dictates that it is the influence andadoption of Western values that has led to the rise ofeating disorders in the non-Western world (Rieger,Touyz, Swain, & Beumont, 2001; Weiss, 1995). Therehave been a number of reasons put forward as towhy eating disorders were thought to be rare in suchsocieties, including the following: non-Westernsocieties traditionally did not greatly value thin-ness and instead valued plumpness (Afifi-Soweid,Najem Kteily, & Shediac-Rizkallah, 2002; Buhrich,
Copyright# 2006 John Wiley & Sons, Ltd and Eating Disorders Association.
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/erv.678
* Correspondence to: Nerissa L. Soh, Discipline of Psycholo-gical Medicine, Blackburn Building DO6, The University ofSydney NSW 2006, Australia. Tel:61 2 9515 5844. Fax:61 29515 7778. E-mail: firstname.lastname@example.org
1981; Lee, Leung, Lee, Yu, & Leung, 1996; Nasser,1997; Tsai, 2000); the collectivistic structure of familyand society offered some degree of protection (Lee &Lee, 1996; Tsai, 2000); and that eating disorders sim-ply were not recognised within the society and inturn, were not brought to the attention of the appro-priate clinicians (Becker, Franko, Speck, & Herzog,2003; Buhrich, 1981; Gordon et al., 2002; Silber,1986; Striegel-Moore & Smolak, 2000; Tsai, 2000).
However, variants of eating disorders have beenknown in the non-Western world for some centuries.For example, fushokubyo, or non-eating illness,was described by Kagawa in 1718th century Japan(Nishizono-Maher, 1998; Nogami, 1997). Most of thepatients were women and the condition was thoughtto have a psychological origin. Nogami (1997) repor-ted that, before World War II, no articles on AN werefound in Psychiatria et Neurologia Japonica, the offi-cial journal of the Japan Association of Psychiatryand Neurology, and that only one case of AN inJapan was documented in 1941. The comparison ofprevalence figures for eating disorders across cul-tures is complicated by the variety of diagnosticcriteria used as well as changes in criteria over time.Nonetheless, the data suggest that while the preva-lence of eating disorders tends to be lower in non-Western countries, it has not been consistentlyreported as such (Figures 1 and 2). Furthermore,there are also differences in prevalence withincountries and communities (Fichter, Elton, Sourdi,Weyerer, & Koptagel-Ilal, 1988).
There is a strong debate regarding whethereating disorders in non-Western patients presents
differently when compared to eating disorderspatients of North-European backgrounds, not leastbecause of differing cultural values and family envir-onment. Much of the research into eating pathologyand body image disturbance in ethnic minoritygroups has been conducted on the African-Americangroup (Crago, Shisslak, & Estes, 1996; Wildes et al.,2001). Numerous Japanese studies were publishedafter 1970 (Nogami, 1997), albeit in the Japanese lan-guage, but otherwise relatively little work has beenundertaken in non-Western groups. Also, few studieshave investigated such groups in both their countryof origin as well as a Western country (Wildes et al.,2001) and to our knowledge, no study has yet beenundertaken on Western groups now living in anon-Western country.
This review is organised around key questionsand issues frequently addressed in the literature.It will first focus on comparing eating and bodyimage disturbance profiles between Western andnon-Western societies, then discuss the impact ofculture and socio-economic status on the apparentemergence of eating pathology and eating disordersin non-Western society. It will then summarise thespeculated differences in the presentation of eatingdisorders, particularly between Asian and Westernpatients, and then move to suggest that core con-structs such as psychological control and accultura-tion may provide a way forward in further teasingout the relationships between eating disorders andculture. Finally, it will discuss the potential impactof differences in body composition across ethnicgroups on the management of eating disorders
China Italy Japan UK SouthAustralia
(1) (2) (3)(4)
Figure 1. Prevalence of anorexia nervosa in women1Chun et al., 1992; 2Santonastaso et al., 1996; 3Nakamura et al., 2000; 4Pyle, 1983; 5Ben-Tovim & Morton, 1990; 6Fichteret al., 1988; 7Fichter et al., 1988; 8American Psychiatric Association, 1994.
Eating and Body Image Disturbances Across Cultures 55
Copyright # 2006 John Wiley & Sons, Ltd and Eating Disorders Association. Eur. Eat. Disorders Rev. 14, 5465 (2006)
and summarise the implications for research ineating and body image disturbances acrosscultures.
EATING AND BODY IMAGEDISTURBANCES: BY GEOGRAPHY
Studies investigating eating disorders and eatingpathology in different cultural groups have beenmostly conducted within Western nations, withmany of these studies suggesting that ethnic min-ority groups in such countries have the lower riskof eating pathology. This is possibly due to a lowerlevel of acculturation to the mainstream society,acculturation being the adoption of another cul-tures values, attitudes and society standards onexposure to them and embodies physical, biologi-cal, political, economic, cultural and psychologicalchanges in identity and attitude (Berry, 1989).Examples of such studies have found North-European Caucasian women to have greater levelsof body dissatisfaction compared to their Asianand African-American counterparts and to bemore likely to attempt to lose weight (Altabe,1998; Lowry, Galuska, Fulton, Wechsler, Kann, &Collins, 2000). Asian women who had moved tothe USA and were studying at a college were foundto have lower levels of restrained eating (16%)compared to their USA counterparts (33%) (Tsai,Hoer, & Song, 1998). Similarly, in a study of femaleuniversity students in Australia, Hong Kong-bornwomen who were more acculturated to Western
culture were found to have more positiveeating attitudes than Australian-born women(Lake, Staiger, & Glowinski, 2000), suggesting thatthe Asian women had the lower risk of eatingpathology.
A number of studies undertaken in Westerncountries found no difference across ethnic groups.A cross-sectional study in the USA conductedwith public school students from five differentethnic groups found that they had similar levels ofeating pathology (French et al., 1997). Similarly,Australian-born Greek- and Anglo-Australian yeareight schoolgirls showed no significant differencein their profiles for eating disorder risk factors,despite 33% of the GreekAustralian girls speakingpredominantly Greek at home (Mildred, Paxton, &Wertheim, 1995). However, other studies found thatethnic minority groups expressed greater frequen-cies of eating disturbances when compared toNorth-European Caucasians. Although Wildes(2001) meta-analysis found little difference acrossethnic groups for bulimia nervosa (BN), it found thatethnicity played a more significant role in subclinicaleating disturbances. Schoolgirls of South Asianbackground living in the UK have been foundto have significantly more unhealthy attitudes toeating than their North-European counterparts(McCourt & Waller, 1995; Mumford & Whitehouse,1988). Also in the UK, the level of acculturation inmale and female students was not associated withBulimia Investigation Test Edinburgh (BITE)scores within each sample of three different ethnicgroups: North-European, AfricanCaribbean and
Japan Italy UK(Caucasian)
China Egypt USA UK (SouthAsian)
Figure 2. Prevalence of bulimia nervosa in women1Nakamura et al., 2000; 2Santonastaso et al., 1996; 3Mumford & Whitehouse, 1988; 4Chun et al., 1992; 5Nasser, 1994;6American Psychiatric Association, 1994; 7Mumford & Whitehouse, 1988.
56 N. L. Soh et al.
Copyright# 2006 John Wiley & Sons, Ltd and Eating Disorders Association. Eur. Eat. Disorders Rev. 14, 5465 (2006)
South Asian. However, South Asian subjects yieldedthe highest raw BITE scores and were more likely tokeep fasts, feel that their lives were dominated byfood, to think about food and to be compulsiveeaters, although they did not have evident concernsabout their body weight (Bhugra & Bhui, 2003). Inthe USA, HispanicAmerican women were found tohave a 9.6% prevalence of binge-eating disorder(BED), compared to 3.9% for AfricaAmericanwomen and only 1.8% for North-European Americanwomen (Fitzgibbon et al., 1998); Hispanic womenwere also significantly more dissatisfied with theirbodies than North-European women (Robinsonet al., 1996).
Studies undertaken solely in an ethnic groupscountry of origin have also yielded mixed findings.For example the incidence of AN in Curacao wasfound to be similar to that of Western nations,despite Curacao having a culture in which it issocially acceptable to be overweight (Hoek, vanHarten, van Hoeken, & Susser, 1998). In a similarvein, there was no difference in drive for thinnessin Japanese schoolgirls before they left Japan for1 year abroad as exchange students and when theyreturned (Furukawa, 2002); however, 90% of thefemale subjects already wanted to be thinner beforethey left Japan. There was also no significant differ-ence in the drive for thinness between SingaporeanChinese schoolgirls and USA undergraduate women,although the former had a higher level of body dis-satisfaction than the female undergraduates in theUSA (Kok & Tian, 1994a). In contrast, the 16% of sec-ondary school girls in Saudi Arabia who scoredabove the screening threshold for the drive for thin-ness subscale of the Eating Disorders Inventorywere more likely to have lived in a Western countryfor at least six months and/or speak a Western lan-guage (Al-Subaie, 2000), both markers of exposure toWestern culture. Also, a preference for thinness wasassociated with speaking English at home in Singa-porean Chinese first-year undergraduate women(Wang, Ho, Anderson, & Sabry, 1999). Yet in HongKong undergraduate women, 27% wanted to weightmore than they already did, compared to just 3.1% ofUSA women (Lee et al., 1996).
It is worth noting that when such studies makecomparisons between countries, they often do soby comparing with the research data from other stu-dies, instead of comparing with data from the samestudy using the same methodologies. However, stu-dies which do utilize the same methodology in aparticular ethnic group in both a Western and non-Western country have again produced mixedresults. For example Chinese girls in Beijing dieted
more than Chinese girls in Sydney, although this dif-ference was not significant and their Anglo-SaxonSydney counterparts still dieted significantly morethan either group (Gunewardene, Huon, & Zheng,2001). In another study, Kenyan-South Asians stillliving in Kenya responded more positively to obesesilhouettes than Kenyan-South Asians living inBritain, the latter having similar responses to theirBritish Caucasian counterparts (Furnham & Alibhai,1983). In contrast, a study of Iranian women living inIran and in the USA revealed that, despite Westernmedia being banned in Iran since 1978 and that bylaw women must wear some form of dark, full-bodycovering that obscures body shape and size, therewere relatively few differences in eating pathologysymptoms between the two groups. Of the differ-ences present, it was the women living in Iran whowere more likely to exercise vigorously to controltheir weight or shape and more likely to desire anempty stomach (Abdollahi & Mann, 2001). Whilethe authors noted that their subjects mothers wouldhave been exposed to Western media and cultureprior to the revolution, they did not address the phe-nomenon that in Middle-Eastern societies wherewomen wear full-body coverings in public, thewomen have a great deal of interest in personalappearance and the fashions they wear beneath(Mahmoody & Hoffer, 1987; Sasson, 1992). Also sup-porting the theme of higher levels of eating and bodyimage disturbances in individuals with less accul-turation to Western society, a study of Taiwanesewomen in both Taiwan and the USA found thatthose in Taiwan not only identified significantlymore strongly with Taiwanese culture than thosein the USA, but they also exhibited significantlyhigher body dissatisfaction and eating disturbances(Tsai, Curbow, & Heinberg, 2003). In both samples,eating and body image disturbances were signifi-cantly and positively associated with greater identifi-cation withTaiwanese culture, contradicting the viewthat traditional cultural values offer protection...