CHAPTER TWO 
OVERVIEW OF LITERATURE 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 9 
2.1             Definitions of Eating Disorders 
     
Eating disorders are a group of disorders, characterized by dysfunctional eating patterns that 
conform to the American Psychiatric Association (Diagnostic and Statistical Manual of Mental 
Disorders: DSM-IV-R: APA, 1994) criteria for Anorexia Nervosa, Bulimia Nervosa and Eating 
Disorders Not Otherwise Specified (EDNOS). It has been estimated that 10% of females in 
western countries will suffer from a diagnosable eating disorder (APA, 1994) making it one of the 
more prevalent psychiatric disorders faced by women (Stice, Telch & Risvi, 2000).  
 
Anorexia Nervosa is characterized by extreme emaciation (body weight less than 85% of that 
expected for age and height), intense fear of gaining weight or becoming fat, despite a low body 
weight; disturbed perception of weight and shape, undue influence of weight or shape on self-
 evaluation or denial of the seriousness of low body weight; and amenorrhea. Anorexia Nervosa 
has a life-time prevalence of almost 1% amongst females, shows a chronic course, results in 
serious medical complications and is associated with psychiatric co-morbidity (Stice, Telch & 
Rizvy, 2000). Onset frequently occurs in adolescence with the highest risk period occurring 
between 15-19 years, where the average point prevalence across studies is estimated at 0.3% (van 
Hoeken, Seidell & Hoek, 2003).  
 
Bulimia Nervosa involves recurrent episodes of uncontrollable consumption of large amounts of 
food; compensatory behavior to prevent weight gain such as self-induced vomiting, laxative 
abuse, diuretic abuse, or excessive exercise; and undue influence of weight or shape on self-
 evaluation (APA, 1994).  Bulimia nervosa has a life-time prevalence of approximately 2% for 
females (Newman, Moffit & Caspi et al., 1996); is marked by a persistent course, and associated 
with high levels of co-morbid psychopathology (Garfinkel, Lin & Goering et al., 1995). Point-
 prevalence rates for bulimia nervosa have been indicated at 8% (Pyle, Halvorson & Neuman et 
al., 1986) and 19% (Halmi, Falk & Schwartz, 1981) of female college students, whereas later 
studies suggested a decline in point prevalence, with estimates ranging from 1.3% (Schotte & 
 10 
Stunkard, 1987) to 2.9%-3.3% (Drewnowski, Yee & Krahn, 1988).  The onset of bulimia nervosa 
frequently occurs in adolescence and the highest-risk group occurs in young females 20-24 years 
old, where the average point prevalence across studies has been estimated at 1% (van Hoeken et 
al., 2003).  
 
Subjects who display all of the symptoms of a full eating disorder syndrome, but who do not 
reach the required thresholds of symptom for disorder, such as weight criteria, are defined as 
having sub-threshold disorder. Those who do not fulfill one of the diagnostic features, such as 
amenorrhea, are considered to have partial-syndrome disorder (Santonastaso, 2005). 
Collectively these are grouped under the category Eating Disorders Not Otherwise Specified 
(EDNOS) which includes a variety of patients who do not qualify for a clinically significant eating 
disorder. This heterogeneity makes it a difficult category for epidemiological or etiological study. 
Community surveys among normal student populations suggest that sub-clinical eating disorders 
are five times more common than full-blown syndromes (Dancyger & Garfinkel, 1995), with up to 
20% of adolescent girls experiencing various forms of sub-clinical binge-eating, purging and 
unhealthy dieting practices (Schwartz, Thompson & Johnson, 1985; cited from Graber, Tyrka & 
Brooks-Gunn, 2003). Partial syndromes are usually brief and self-limiting (Patton, Coffey & 
Sawyer, 2003). 
 
Criteria for Binge-eating Disorder are also outlined, but to date, this disorder is categorized as a 
particular case of EDNOS. Binge-eating Disorder involves repeated episodes of uncontrollable 
binge-eating characterized by rapid eating or eating alone because of embarrassment; marked 
distress regarding binge eating; and the absence of compensatory behaviors. Binge-eating 
Disorder has a life-time prevalence of approximately 4% in the community and is present in 
about 30% of individuals attending weight control treatment (Brody, Walsh & Devlin, 1994; cited 
from Stice, Telch & Rizvy, 2000). This disorder appears to follow a persistent course and is 
associated with obesity, weight cycling, health complications and psychiatry co-morbidity (Stice, 
Telch & Risvi, 2000).  
 11 
2.2             Eating Disorders as Culture-bound Syndromes 
 
Until recently, anorexia was considered the ?quintessential western culture-bound syndrome? 
(Prince, 1985) and eating disorders were considered unique among psychiatric disorders in the 
degree to which socio-cultural factors may influence their epidemiology, development and 
etiology (Gordon, 2001). The term ?culture-bound? refers to a syndrome of signs and symptoms 
which is restricted to a particular culture or group of cultures (Pow Meng Yap, 1951; cited from 
Nasser, 1997). The Culture-specific approach (Kleinman, 1977; cited from Nasser, 1997) argues 
that a culture-bound disorder symbolizes core meanings and behaviors that are considered 
norms of that specific culture. The Generalist approach (Kraepelin, 1904, cited from Nasser, 
1997) disputes that there may be factors in any culture that are necessary and sufficient to 
account for causation and considers culture-bound syndromes as atypical, culturally local 
variants of universally recognized pathology. Anorexia is thus seen as an expression of neurotic 
distress that is shaped by western cultural values, such that western ideals of thinness become 
the ?socially sanctioned coloring of distress and not the cause? (Russell & Treasure, 1989).  
 
2.2.1           The Acculturation Hypothesis 
 
The culture-specific model attributes the development of eating disorders to specifically western 
cultural values where appearance is seen as a central evaluative dimension and the thin body 
type is idealized as a metaphor of beauty and attractiveness (Stice, Schupak-Neuberg & Shaw et 
al., 1994). Social pressures to conform to these ideals are believed to render women vulnerable to 
dissatisfaction with their bodies (Stice, Maxfield & Wells, 2003) and a preoccupation with weight 
and dieting behavior (Powell & Kahn, 1995), which has been seen as a powerful antecedent of 
eating disorders (Hsu, 1997). Dieting is seen to fall on a continuum of severity, with body image 
disturbances and dieting behavior at one end and anorexia and bulimia nervosa at the 
pathological end of this ?normative? discontent and corrective behavior (Nylander, 1971).  
 12 
Evidence for this model has largely been drawn from data which suggests that, the increasing 
prevalence of eating disorders in the west during the 20th century has been accompanied by an 
increasingly ?thin? body ideal as portrayed by popular role models and media images (Garner & 
Garfinkel, 1980) which has placed increasingly stringent expectations for thinness on women.  
 
2.2.2           Eating Disorders in the West 
 
Anorexia nervosa was first introduced in the medical literature (Gull, 1874; Laseque, 1873; cited 
from Di Nicola, (1990 a) as a somatic expression of hysteria (apepsia hysterica) to describe 
deliberate weight loss in adolescent girls. At this time, Picasso?s painting ?The Girl in Chemise? 
endorsed thinness as an expression of fashionability (Nasser, 1997). Cases of anorexia nervosa 
were documented among the western bourgeoisie and for some time, remained a condition that 
predominantly afflicted female teenagers in affluent families of Western Europe and North 
America (Jilek, 2001). By the end of the 19th century, physicians began to differentiate anorexia 
as a syndrome that is distinct from the more general categories of hysteria (Keel & Klump, 2003).  
 
During the 1920?s, fashion demanded a tubular style of clothing which deemphasized breasts, 
waist and hips and beauty, cosmetics and fashion became institutionalized in consumerist society, 
which began promoting the idea that thinness was a crucial dimension of female beauty 
(Brumberg, 1988); as recorded by Helena Rubenstein (1930; cited from Nasser, 1997) who noted, 
?an abundance of fat is not in accord with the principles that rule our conception of the beautiful.?   
 
Incidence rates of anorexia nervosa in Western Europe increased gradually from the mid 1930?s 
until the mid 1960?s and then accelerated into the 1970?s  and 1980?s so sharply that anorexia 
nervosa began to be referred to as a ?social epidemic? (Bruch, 1978). This increase appeared most 
substantial in females (15-42 years) who demonstrated a significant linear, increasing trend 
between 1935 and 1984 (Lucas, Beard, O?Fallon et al., 1991).   
 
 13 
This ?social epidemic? spread over Western and Southern Europe, finally involving young women 
of all socioeconomic classes (Jilek, 2001). This apparent increase in anorexia nervosa appears to 
coincide with drastic changes in the western feminine body-ideal, as propagated by Miss America 
and the centerfold of Playboy magazine which became increasingly thin and less curvaceous 
between 1959 and 1988 (Garner & Garfinkel, 1980; Morris, Cooper & Cooper, 1989). During the 
period 1979?1988, the majority of Miss America contestants (Wiseman, Gray & Mosimann et al., 
1992) and Playboy centerfold models (Szabo, 1996) were 15% or more below expected body 
weight; suggesting that western women were pursuing an  ?anorexic? body-ideal.  The popular 
?Twiggy? in the late 1960?s epitomizes the ?waif-like?, thin, non-curvaceous body ideal prevalent at 
this time.  
 
The number of magazine articles on dieting and exercise also increased significantly between 1959 
and 1988 (Garner et al., 1980; Wiseman et al., 1992) which also parallels the period of increasing 
incidence of eating disorders in the west. These trends appeared to continue through the 1990?s 
although the incidence of anorexia nervosa in Europe appears to have remained relatively stable 
after the 1980?s (Hoek, van Hoeken & Katzman, 2005).  
 
The term ?bulimia nervosa? (BN) was first coined by Russell in 1979 (cited from Halmi, 2005) to 
describe a syndrome of fat-phobia, and episodes of overeating followed by compensatory 
purging.  Soundy, Lucas & Suman et al., (1995) documented a rapid increase in the incidence of 
bulimia nervosa in females from 1980 to 1983; which then remained relatively stable from 1984 
to 1990. By the mid 1980?s, bulimia nervosa was more common than anorexia nervosa with 
18.6% of college students reporting a history of bulimia nervosa (Pope, Hudson & Yurgelen-Todd 
et al., 1984); while Whitehouse, Cooper & Vize et al., (1992) reported a prevalence of 1.5% for 
full-syndrome bulimia nervosa in the early 1900?s. Keel, Todd and Heatherton et al., (2006) 
found a steady decline in the point-prevalence of bulimia nervosa from 1982 to 1992 to 2002.  
 
 
 14 
2.2.3           Eating Disorders in non-western populations  
 
Until recently there were few reports of eating disturbances among black and non-western 
women (Dolan, 1991) and it was argued that these women were less vulnerable to eating 
disorders as they frequently value plumpness as a metaphor of attractiveness, fertility and 
prosperity (Furnham & Alibhai, 1983), are more accepting of larger body types (Rucker & Cash, 
1992; Powell & Kahn, 1995) and experience less social pressures towards thinness (Greenberg & 
La Porte, 1996).  
 
Since the 1990?s, however, reports of disturbed eating attitudes and cases of eating disorder have 
been reported from many non-western countries (Gordon, 2001) and in black and native ?
 American women in America (Crago, Shisslak & Estes et al., 1996) and South Africa (Szabo, Berk 
& Tlou et al., 1995). The acculturation hypothesis attributes this apparent increase of eating 
disturbances in non-western women to their exposure and assimilation of specific western 
cultural values of thinness, or ?western contagion?. Evidence for this hypothesis is drawn from 
epidemiological, cross-cultural, immigrant and acculturation studies.  
 
With the exception of Japan and Chile, the first case reports of anorexia appeared in the 1980?s, 
from India (Chakraborty, 1985); Singapore (Ong, Tsoi & Cheah, 1982); Hong Kong (Lee, Chiu & 
Chen, 1989); Taiwan (Tseung, Lee & Lee, 1989), South Korea (Lee, Kwak, Rhee et al., 1987), 
Egypt (Okasha, Kamel, Sadek et al., 1977) and Eastern Europe (Faltus, 1986; cited from Rathner 
2001). Reports of bulimia also began to emerge in Hong Kong in the early 1980?s (Lam & Lee, 
2000; Tseng et al., 1989). Many of these early cases of eating disorder did not suffer from body 
image concerns and were from lower socio-economic levels (Lee, 2000). Since the 1990?s, reports 
of dysfunctional eating attitudes and behaviors and cases of eating disorder, have been increasing 
across the non-western world. 
  
 
 15 
2.2.3.1         Asia 
 
Increasing numbers of cases of anorexia were found in Hong Kong between 1984 and 1998 (Lee, 
2000) and the number of clinic referrals increased from 2 per year in 1990 to one per week in 
2000 (Lee, 2001). By the mid 1990?s, several cases of anorexia and binge-purging behavior were 
reported in Singapore (Ung, Lee & Kua, 1997) and prevalence rates for South Korea were 
reported at 0.7% for anorexia nervosa and 0.8% for bulimia nervosa (Efron, 1997). Several cases 
of anorexia were also reported in India (Chandra, Shah & Shenoy et al., 1995) and studies 
indicated that a pattern of weight consciousness was emerging (Srinivasen, Sresh & Vasantha, 
1998; cited from Gordon, 2001). By the late 1990?s, weight consciousness was common amongst 
female students in Hong Kong (Lee, 2001) and high schoolgirls in South Korea (Tsai, 2000) and 
Singapore (Kok & Tian, 1994). Despite this apparent increase in weight consciousness, the 
prevalence of anorexia nervosa in Singapore, remained low (Gordon, 2001) and 60% of these 
cases reported no conscious fears of ?fat?.  
 
The Eating Attitudes Test (EAT40; Garner & Garfinkel, 1979; EAT26: Garner, Olmsted & Bohr et 
al, 1982) is a screening test that identifies risk for developing an eating disorder (Lee & Lee, 
2000) where about 25% of those scoring positively (greater than or equal to 20) may develop an 
eating disorder (Patton, Johnson-Sabine & Wood et al., 1990). By the late 1990?s, positive EAT 
scores were recorded in 10.8% of young females in Hong Kong and 18% of South Korean female 
undergraduates (Lippincott & Hwang, 1999). High rates of positive EAT scores were also found 
in Indian females 14-23 years (29%; King & Bhugra, 1989) and female medical and nursing 
students in Pakistan (21.7%; Barbar, Alam & Ali et al., 2002). Choudry and Mumford (1992) 
found that 7% of Urdu-speaking schoolgirls in Pakistan scored positively on the EAT26 while 
10.3% of Pakistani schoolgirls in an English-medium school scored positively (Mumford, 
Whitehouse & Platts, 1992). The authors concluded that dysfunctional eating increased with 
increasing westernization and was mediated through greater dissatisfaction with body shape. 
 
 16 
2.2.3.2         The Middle-East 
 
Eating disorders were considered rare in the Israeli Kibbutz until 1965, when the annual 
incidence of anorexia rose 800% over the next 25 years (Kaffman & Sadeh, 1989). Nasser (1986) 
found that 12% of female Arab students in Cairo scored positively on the EAT-40 while none 
qualified for anorexia nervosa or exhibited bulimic symptoms. Nasser (1994 b) found that 11.4% 
of a sample of female adolescents scored positively on the EAT26 while 1.2% qualified for a 
diagnosis of bulimia nervosa and 3.4% for partial syndrome bulimia nervosa. Apter, Abu Shah & 
Iancu et al., (1994) found that 27.3% of Israeli schoolgirls in Kibbutzim scored positively on the 
EAT26 while 19.4% of Bedouin and 18.6% of Arab Muslim schoolgirls scored positively. Greater 
risk of eating disorder was understood in terms of changing female role expectations. 
 
Abou-Saleh, Younis & Karim, (1998) reported on five cases of anorexia in the United Arab 
Emirates, one of whom had no contact with western influences; while female Saudi-Arabian 
adolescents demonstrated widespread fear of fatness, dieting and drive for thinness that was 
similar to Canadian adolescent females at the time (Al-Subaie, 2000). By the mid 2000?s, studies 
across the Middle East indicated that 23.4% of a sample of high school girls in the United Arab 
Emirates scored positively on the EAT while 2% met the full criteria for anorexia nervosa (Eapen, 
Mabrouk & Bin-Othman, 2006).  Nobakht & Dezhkam (2000) found that 24.16% of a sample of 
female adolescents in Tehran scored positively on the EAT26, while 0.9% qualified for anorexia 
and 3.23% for bulimia nervosa. In all, 47.8% of the subjects reported using some form of 
compensatory behavior to control their weight. The authors speculated that this may be due to 
conflict between conformity to Islamic values and western cultural influences. In another sample 
of females in Teheran, 58.6% were found to restrict food to control weight and shape; 42.4% 
fasted for long periods; 58.6% felt fat; 5.2% vomited to control weight; 3.4% took laxatives; and 
46.4% exercised strenuously (Abdollahi & Mann, 2001).  
 
 17 
Al-Adawi, Dorvio and Burke et al., (2002 a) found that 29.4% of female Omani adolescents 
scored positively on the EAT26. Elal, Altug and Slade, (2000) compared the EAT scores of two 
groups of Turkish female college students over a ten year period. The ?bulimia? factor and 
?ambivalence about eating' only emerged in the later study. This was interpreted in terms of 
increasing exposure to modern values that clashed with traditional Turkish values.  
 
2.2.3.3         Eastern Europe 
 
Apart from a few case reports of anorexia, eating disorders were generally considered rare in 
Eastern Europe before the dissolution of the Iron Curtain in 1989, and it was claimed that 
bulimia was scarcely known in these countries (Rathner, 2001). In general, the few studies 
conducted prior to unification, indicated that, overall, Eastern and Western European students 
showed similar weight indices and prevalence of disorder (Rathner, Tury & Szabo et al., 1995).  
 
Studies conducted immediately prior to unification, indicated that East-Berlin anorexics scored 
significantly lower than West-Berlin anorexics on subscales measuring drive for thinness and 
body dissatisfaction on the Eating Disorders Inventory (EDI; Garner et al., 1983); which is 
another screening test for eating disorder, (Steinhausen, Neumarker & Vollrath, 1992). In 
contrast, non-clinical groups of East-Berlin school girls scored significantly higher on the EAT 
than their West-Berlin peers, and scores were comparable to scores obtained by school girls in 
Britain and Belgium (Neumarker, Dudeck, Vollrath et al., 1992).  
 
Szabo and Tury (1991) found that 6.2% of female medical students in Hungary scored positively 
on the EAT while 1.3% met criteria for bulimia nervosa. Hungary was seen as more liberal than 
the rest of communist Europe and these prevalence rates have been interpreted as evidence of 
western influence (Rather, 2001). 
 
 18 
The study by Neumarker, Dudeck, Vollrath, et al., (1992) was replicated in East-Berlin, following 
German unification (Hein, Neumarker & Neumarker, 1998) and found that while the EAT scores 
of school girls were similar to those of East-Berliners in the 1980?s, there was a notable increase 
in body dissatisfaction, dieting behavior and a quest for thinness. Another study that was 
conducted following unification (Rathner, Tury & Szabo, 2001) demonstrated no significant 
differences on the EAT and similar rates of bulimia nervosa and sub-clinical bulimia between 
large population samples from Austria, Hungary, the Czech Republic and Poland.  No cases of 
anorexia were detected. Rathner et al., (2001) concluded that the emergence of eating disorders 
in Eastern Europe may be underscored by socio-political transformation which brought a 
western free-market economy and an over-identification with western values. 
 
2.2.3.4         South America  
 
Chile was first to report cases of anorexia nervosa and observed a notable increase in incidence 
during the 1970?s (Pumarino & Vivanco, 1982; cited from Gordon, 2001). These patients 
presented with a ?relentless drive for thinness? that was indistinguishable from that described in 
the west. Chile was undergoing ?momentous socio-political and economic change during the 
1970?s (Gordon, 2001).  
 
More recent studies have indicated that 29% of a sample of female adolescents in Argentina 
scored positively on the EAT (Bello, 1995; cited from Hoek, van Hoeken & Katzman, 2005) and 
9% of these high scorers qualified for a diagnosis of eating disorder.  A study in Brazil found that 
16.5% of women (12 to 29 years) scored positively on the EAT26 (Nunes, Barros &  Olinto, 2003) 
of whom 0.6% qualified for anorexia nervosa and 5.6% for bulimia nervosa in a follow-up study 
(Nunes, Camey, & Olinto et al., 2003). Eating disorders in Brazil have recently made newspaper 
headlines following the death of four prominent women from anorexia nervosa (The New York 
Times, March 15, 2007) and claim that 8.0% of Brazilian women suffer from an eating disorder.    
 
 19 
2.2.3.5         Africa 
 
Prior to the 1970?s, eating disorders were considered rare on the African continent. In an early 
review of cross-cultural attitudes, Powdermaker (1960; cited from Gordon, 2001) found that the 
traditional body ideal among Africans tended towards a large, full form where fatness was closely 
tied with fertility and pubescent girls were ?fattened up? to make them more marriageable.   
 
During the 1980?s a few case reports of eating disorder in black Africans began to emerge from 
Nigeria (Nwaefuna 1981; Famuyiwa, 1988), Zimbabwe (Buchan & Gregory, 1984) and Ethiopia 
(Fahy, Robinson, Russell et al., 1988); most of whom had exposure to western cultural 
influences. In a sample of black Zimbabwean females, Hooper and Garner (1986) found that high 
scores on the Eating Disorders Inventory (EDI; Garner et al., 1983) were more evident on the 
general psychological subscales of this screening test than subscales that focus on body image. 
Eddy, Hennessey & Thompson-Brenner (2007) also found that the majority of a sample of young 
black females in Tanzania, did not achieve subscale scores on the EDI that were suggestive of 
eating disorder; although 39.8% of  this sample displayed cognitive symptoms of eating disorder. 
Moreover,  10% of this sample had a Body Mass Index of less than 17.5; 1.9% met modified 
criteria for anorexia nervosa, 0.46% qualified for bulimia nervosa; 4.7% indicated clinically 
significant symptoms of eating disorder not otherwise specified; and 10% engaged in binge-
 eating and 5% purged. Symptoms were most common in those with exposure to western media. 
The authors concluded that the intended factor structure of the EDI was not supported. Bennett, 
Sharpe and Freeman et al., (2004) also found that while 1.5% of a sample of secondary 
schoolgirls in Ghana, were pathologically underweight due to self-starvation, these girls did not 
score any differently to control groups on the EAT40 and attributed their food restriction to 
religious motivations rather than body image issues.  In contrast, Oyewumi and Kazarian (1992) 
found that 18.6% of a sample of black female high school students in Nigeria scored positively on 
the EAT.   
 
 20 
 
2.2.3.6         South Africa 
 
The first published reports of eating disorders in South Africa appeared in the 1970?s but were 
focused on the emergence of eating disorders in the white female population (Beumont, George 
& Smart, 1976; Norris, 1979; Ballot, Delaney & Erskine et al., 1981). A survey of hospital 
admissions in Cape Town, between 1979 and 1989 (Nash & Colborn, 1994) revealed that no black 
South Africans had been referred for hospital treatment for either anorexia or bulimia nervosa, 
during this time. Norris (1979) documented a 20-fold increase in admissions of white female 
patients during the 5 years prior to this study.  
 
Shefer (1987) found that 11.8% of white female undergraduate students scored positively on the 
EAT40; 21.9% engaged in binge-eating and 6.3% used self-induced vomiting as a means of 
weight control. Until recently, eating disorders have thus been seen as the ?exclusive domain? of 
white South Africans who have been closely aligned with western cultural ideals (Norris 1979; 
Shefer 1987; Nash & Colborn, 1994; Le Grange, Louw, Breen & Katzman, 2004). 
 
In January 1993, Szabo first diagnosed an eating disorder in a black South African female, which 
led to the publication of a series of cases describing anorexia (bulimic subtype) in black South 
African females (Szabo et al., 1995), all of whom were urbanized, receiving higher education and 
presented with fears of fatness. Szabo et al., 1995 concluded that cultural homogenization within 
this setting may be more relevant than race as an underlying factor in eating-related 
psychopathology. Since 1995, several surveys among racially diverse groups of adolescents and 
young adults also suggest the emergence of disturbed eating attitudes in young black women in 
South Africa. In an article entitled ?Africans look for beauty in the western mirror? (Christian 
Science Monitor, Dec 1999), the current views of black women in South Africa were reported as 
as: ?Its embarrassing to be a fat African mama now. We are more aware since we got democracy. 
We want to be healthy, independent women who look good? (cited from Gordon, 2001).  
 21 
 
In a study conducted in 1996, Szabo and his colleagues (Szabo & Allwood, 2004 a & b) found that 
18.7% of a sample of urban black high-school girls scored positively on the EAT26 while only 3% 
of a sample of rural black school girls scored positively; suggesting that an increased risk of 
eating disorder may have been associated with urbanization.  Szabo and Hollands (1997) found 
that 37.5% of the black girls in an exclusive urban high school scored positively on the EAT26. 
 
In a sample of male and female college students Le Grange, Telch and Tibbs (1998) found that 
13% scored positively on the EAT40 while 5% scored positively on the Bulimic Investigatory Test 
Edinburgh (BITE; Henderson and Freeman, 1987). Caradas, Lambert and Charlton (2001) found 
that 17.9% of a sample of black female adolescents aged 15 to 18 years had positive scores on the 
EAT26; while Senekal, Steyn and Mashengo et al., (2001) found that 14% of black female college 
students scored positively. Edwards and Moldan (2003) found that 10% of black female college 
students scored positively on the Bulimia Test (BULIT; Smith & Thelen, 1984).  
 
Le Grange, Louw and Russell et al., (2006) found that 11.9% of a sample of black, male and 
female adolescents and young adults scored positively on the EAT26 while 5.6% scored positively 
on the BITE. This study eliminated results obtained from all of the 128 black subjects from a 
black school in an informal settlement as these authors concluded that they ?yielded a much 
higher proportion of potentially eating disordered subjects than one would expect from any 
cohort? and self?starvation was found to have meanings that were more about inadequacy of food 
resources than cosmetic ideals of attractiveness. Le Grange et al., (2004) note that there does not 
appear to be any studies that suggest anorexia nervosa among black females in South Africa.  
Other research has suggested that there may be a tendency amongst black South Africans to 
present with bulimic symptoms (Szabo et al., 1995). 
 
 
 
 22 
2.2.3.7         Limitations and Potential of Prevalence Data 
 
The acculturation hypothesis rests on the assumption that the ?sine non qua? of eating disorders 
is a specifically western over-evaluation of body weight and shape which may be acquired by 
non-western individuals who are exposed to these cultural ideals. The rise of eating disorders in 
non-western communities since the 1990?s is seen as evidence for this hypothesis. Eating 
disorders are, however, probably the most difficult of all mental disorders to study 
epidemiologically (Lee, 2005) such that prevalence rates actually remain unclear and any 
increase in non-western countries remains conjecture (Palmer, 2001). 
 
Evaluating the impact of western ideals of thinness on the prevalence of eating disorders in non-
 western cultures requires reliable base-line rates of eating disorders before and after 
westernization (Bisaga, 2001). The low prevalence of eating disorders makes it necessary to study 
a large number of subjects (Hoek, van Hoeken & Katzman, 2005) which is time and cost 
intensive. Epidemiological data on eating disorders in non-western populations is, therefore, 
extremely limited, particularly prior to the 1990?s (Dolan, 1991) and most prevalence data has 
been gathered from case reports or records. Case reports and records notoriously underestimate 
the magnitude of any medical problem in the community (Dolan, 1991) as they only report on 
subjects who have presented for treatment, which is heavily influenced by the help-seeking 
patterns of subjects, the degree of public awareness of eating disorders, the availability of user-
 friendly services and Eurocentric assumptions of clinicians, which may influence the  diagnosis 
and referral of cases, with whites more likely to receive a diagnosis than blacks (Dolan, 1991). 
Many developing communities also do not have the infrastructure to gather salient data on eating 
disorders over time (Furnham, 2005). Existing epidemiological evidence does not, therefore, 
exclude the possibility that eating disorders existed in non-western communities in some form, 
before westernization. Conversely, the establishment of treatment facilities typically leads to 
sudden increases in the clinical prevalence of eating disorders (Lee, 2005), suggesting that 
apparent increases in non-western communities, after westernization, may be artifact. 
 23 
Secondly, the use of different methods of case finding, different population samples, different 
prevalence measures and the adoption of different definitions of illness have also had 
considerable impact on prevalence estimates for anorexia and bulimia nervosa and seem as likely 
to account for variation between studies as real population differences (Okasha, 2005). Most 
surveys have, for example, used convenience samples, which range from high risk adolescent 
female samples (Szabo & Allwood, 2004), to mixed gender and age-group samples (Le Grange et 
al., 1998). Keel and Klump (2003) note that earlier prevalence studies tended to produce higher 
estimates of symptoms than interview assessments, which were more often used by later studies.  
These authors also point out that early definitions of bulimia nervosa did not specify a minimum 
frequency of weekly binge-eating and that stipulating a frequency of binge-eating more than once 
a week, reduces these early point prevalence rates of bulimia nervosa from 8% in college females 
(Pyle et al, 1986) to 1% in 1980 and 3% in 1983. Studies relying solely on formal diagnoses may 
thus be tracking changing diagnostic criteria over time rather than a true increase in prevalence.  
 
The Diagnostic and Statistical Manual for Mental Disorders (APA, 1994) also stresses the 
importance of weight and shape concerns in the diagnosis of anorexia and bulimia nervosa 
(Gowers and Shore, 2001). An accumulating amount of evidence has, however, demonstrated 
that in non-western countries where thinness has not been traditionally valued as a symbol of 
attractiveness, anorexia nervosa has not been accompanied by weight concerns, but by other, 
more culturally relevant complaints (Lee, Lee, & Leung, 1998; Hsu & Lee, 1993), although non-
 western women who live in the West and develop eating disorders also demonstrate weight 
phobia (Woodside, 2005). Lee, Ho and Hsu, (1993) found that 58.6% of Chinese anorexic 
patients in Hong Kong displayed no conscious fear of fatness and attributed their food refusal to 
somatic complaints and other rationales although they resembled western anorexic patients. 
When placed in a western treatment program, these eastern anorexics frequently ?acculturate? 
towards a more western anorexia and body image concerns and fear of fatness replaces their 
earlier concerns (Woodside, 2005).  
 
 24 
Most community samples have also used western screening tests such as the EAT and the EDI to 
establish risk for eating disorder. Lee et al., (1998) showed that while the psychological profile of 
Chinese, ?fat-phobic? anorexics and bulimics, on the EDI, was similar to that of Canadian patients 
with eating disorder, non-fat-phobic Chinese anorexics scored significantly lower on cardinal 
features such as drive for thinness. Lee, Kwok, Liau and Leung (2002) demonstrated that 61.6% 
of Chinese anorexics scored atypically low on the EAT26, where only 11.4% of non-fat phobic 
anorexics scored positively, while 58.5% of fat-phobic anorexics and 76.1% of bulimics scored 
positively. Surveys in non-western countries have traditionally revealed a low prevalence of 
anorexia nervosa, suggesting that some potential cases of anorexia may have been missed by 
western screening instruments which endorse ?fat phobia? as a cardinal symptom of anorexia 
(Lee & Katzman, 2002). 
 
 Indeed, although early reports from Eastern Europe indicated that eating disorders were rare 
prior to unification in 1989, the few studies conducted shortly before unification, indicated that 
the prevalence of eating disorders was very similar across East and Western Europe (Rathner et 
al., 1995). Anorexics in East-Berlin also reported significantly less body dissatisfaction and drive 
for thinness than West-Berlin anorexics (Steinhausen et al., 1992), suggesting that eating 
disorders in East-Berlin may have lacked the presence of fat phobia prior to westernization and 
thus failed to be included in early reports from Eastern Europe.  
 
?The application of western diagnostic criteria and instruments in non-western communities may 
therefore, merely realize a self-fulfilling prophecy that the disorder is non-existent or extremely 
rare in these settings or, as in more recent years, be used to explain the emergence of familiar 
eating disorders due to the homogenizing toxin of westernization? (Lee, 2004). Following 
Woodside (2005), evidence for the increasing incidence of eating disorders with westernization 
may, therefore, simply reflect the westernization of existing pathology. 
 
 
 25 
Consideration of the historical context within which self-starvation and other symptoms of eating 
disorders have occurred, also suggests that eating disorders may indeed have existed since 
ancient times, within diverse cultures and communities (Halmi, 2005), long before the western 
cosmetic culture of thinness. Spiritually-motivated self-starvation was noted in early eastern and 
western religions as a means of penance, purification or ascetic control of the body. Many 
Christians who practiced piety and self-starvation were elevated to sainthood; which has been 
viewed as a variant expression of anorexia nervosa or ?holy anorexia? (Bemporad, 1996). Many 
Victorian girls explicitly followed the ascetic aspirations of St Catherine and were named the 
?miraculous maids? or ?fasting girls? (Keel & Klump, 2003). Chipley (1860; cited from Keel & 
Klump, 2003) described cases of emaciation among females who actively refused to eat as 
?sitomania?. A condition called ?chlorosis? was also used to describe a syndrome similar to 
anorexia nervosa. Di Nicola (1990 a) refers to these syndromes as ?clinical analogues? of anorexia. 
Parry-Jones (1985) studied inpatient admissions in England between 1812 and 1917, and found 
40 cases of anorexia as well as a range of other admissions that could be described as historical 
analogues of anorexia nervosa.  
 
The practices of binge-eating and vomiting have also been widely prevalent, since early history 
and across diverse cultures (Halmi, 2005). Galen (AD 130-200) described the episodic ingestion 
of considerable amounts of food as bulimos (James 1743, from Halmi, 2005) which could be 
treated by tickling the back of throat to induce emesis.  Ancient Egypt, Greece, Rome and Arabia 
also describe orgies of eating and Rome is notorious for the ?vomitorium? which allowed them to 
relieve the effects of excessive eating (Miller & Pumariega, 2001). Many of the fasting saints were 
also known to engage in binge-eating (St. Veronica) and self-induced vomiting (St. Catherine) 
(Keel & Klump, 2003). Medical literature of the 17th and 18th century defined periods of insatiable 
appetite as  bulimy or canine appetite (Parry-Jones and Parry-Jones, 1991) and ?cynorexia? which 
described eating that is out of control followed by compulsive vomiting (Forestus, 1602; cited 
from Keel & Klump, 2003).   
 
 26 
Interestingly, none of these early descriptions of self-starvation included any reference to weight 
and shape concerns or fears of fatness and even the first medical definitions focused on anorexia 
as a somatic expression of hysteria such as ?bradypepsia? and ?visceral neuroses? (Brumberg, 
1988). Typical ?fat phobia? was not described until the 1930?s (Brumberg, 1988) and it was not 
until the mid 20th century that weight concerns came to dominate the rationale for food refusal. 
Negative findings do not, therefore, mean that eating disorders did not exist in non-western 
countries before 1990 or before the culture of thinness in western countries.  Apparent increases 
in non-western communities could, therefore, be artifact or due to factors other than the 
assimilation of western ideals.  
 
A review of cross-cultural literature (Rieger, Touyz, Swain, et al., 2001) indicated that when 
weight loss is defined in terms of positive value instead of ?fear of fat?, anorexia is not limited to 
contemporary western cases of anorexia, suggesting that other cultures may also value thinness, 
that may be underscored by culturally different meanings and motivations.  These authors 
speculated that while assumptions focusing on western ideals of thinness remain ?remarkably 
resilient?, they may also have exaggerated the role of westernization in non-western cultures; and 
that the intrinsic ideologies, values and practices of non-western cultures also need to be 
considered in the etiology of eating disorders.  
 
Meta-analysis of a wide range of sources (Keel & Klump, 2003) found that anorexia nervosa has 
indeed been observed in every non-western region of the world and in the absence of western 
influence. Food refusal was reported for all cases, but the presence of weight concerns was not 
universal. These authors also found only a moderate effect size for an increase in the incidence of 
anorexia in western society over the 20th century which has been seen to coincide with increasing 
idealization of thinness. They concluded that, although cultural factors may influence rates of 
anorexia, cultural ideals of thinness may not be sufficient or necessary in the background of 
anorexia and that deliberate or self-motivated food refusal and starvation may have many 
different motivations, including weight concerns.  
 27 
Miller (2005) concluded that ?given all the methodological problems, it is doubtful whether an 
epidemiological approach to the question of cultural or temporal trends in the incidence of eating 
disorders, is of any use at this stage?. Despite these limitations, the increase of eating disorders in 
non-western cultures has been interpreted as ?evidence for the exportation of western culture and 
the culture-bound nature of eating disorders, which sets up eating disorders and western 
influence as both evidence and explanations for each other, thereby setting up a process of 
circular reasoning wherein the notion of culture is reified and is used to bolster existing 
epistemologies rather than challenge them? (Lester, 2004). Moreover, atypical cases which may 
shed light on the etiology of eating disorders are systematically excluded. ?This gives the illusion 
of cross-cultural substantiation for psychiatric diagnostic criteria while obscuring genuine 
cultural analysis and other factors which may have etiological salience? (Lester, 2004).  
 
Nasser (1997) noted that although epidemiological studies have typically demonstrated a higher 
prevalence of eating disorder in the United States of America than in Britain and Europe, these 
differences within western countries have, traditionally, been ignored; although exploration of 
differences within cultures as well as the similarities between cultures may provide significant 
insight into potential etiological factors. Surveys within non-western countries (see previous 
section) have indeed demonstrated highly variable scores on screening tests between different 
samples, with some recent studies (Nobakht & Dezhkam, 2000) recording higher percentages of 
positive EAT scores than Canadian samples.  High scores on the EAT in non-western populations 
are frequently attributed to a lack of validity in the trans-cultural application of these screening 
tests (King & Bhugra, 1989). Following Nasser and Katzman (2003) however, these variable 
findings could provide valuable etiological insights.  
 
Apter et al., (1994), for example, found that the EAT profiles of middle-eastern adolescents were 
most similar to the profile of diagnosed anorexics when they were exposed to western body ideals 
and the presence of conflict between traditional and modern expectations of the female role. The 
EAT scores of girls in an Israeli Kibbutzim were the highest (27.3%) and were most similar to the 
 28 
anorexic profile, followed by the Bedouin (19.4% EAT-positive) and Arab Muslim (18.4% EAT-
 positive) groups; while the Circassians scored lowest (8%). The authors concluded that changes 
in the female role within the Kibbutz had placed these adolescents under severe role conflict 
which may have been associated with elevated EAT scores; while the Circassians had maintained 
traditional female roles. Bedouin society was also undergoing rapid social change from a 
pastoral, nomadic lifestyle, to an urban, industrialized existence at the time; suggesting that 
variable EAT scores may have reflected varying degrees of social transition and role conflict. 
Kaffman and Sadeh (1989) also identified an association between the increased prevalence of 
eating disorders in the Israeli kibbutz and changes in the structure of family life and the role of 
women. Nasser (1997) noted that the Arab world has been experiencing a feminist movement 
during the last decades; suggesting that variable EAT scores within these non-western 
populations may reflect varying degrees of female emancipation and socio-cultural transition.  
 
Research also indicates that shortly before unification in Eastern Europe, East-Berlin high school 
girls scored significantly higher on the EAT than West-Berlin schoolgirls (Neumarker et al., 1992) 
suggesting an over-identification with western cultural norms within the rapidly changing socio-
 political context of East-Berlin at the time. Surveys conducted in East-Berlin shortly before 
unification, also demonstrated that clinical samples were only discriminated from non-clinical 
samples by significantly higher scores on general psychological subscales of the EDI; 
ineffectiveness, perfectionism and maturity fears (Steinhausen et al., 1992). Following Apter et 
al., (1994) this suggests that the onset of clinical disorder may have depended on the 
simultaneous presence of individual conflict within the rapidly changing socio-political context of 
Eastern Europe at the time. Hooper and Garner (1986) also found that high scores on the EDI 
were most evident on the general subscales in black Zimbabwean females. The authors 
speculated that subjects may have been experiencing acculturative stress encountered during a 
process of westernization where western ideals are juxtaposed with traditional values. In 1980, 
Zimbabwe achieved a black-majority government after years of white-dominated rule, suggesting 
that Zimbabwe may also have been experiencing a period of sociopolitical transition at the time. 
 29 
Lee et al (1998) found that fat-phobic Chinese anorexics also scored higher than non-clinical 
samples in ineffectiveness and maturity fears and suggested that drive for thinness needs to be 
fueled by more general pathology to produce disorder. Hungarian female students also 
demonstrated higher EDI scores and a higher rate of eating disorder than Austrian female 
students (Rathner et al., 1995) and scored highest on the General Health Questionnaire-28 
(Goldberg & Hillier, 1979) which measures general psychological distress.  The authors 
speculated that exposure to cultural and social change may lead to significant stress and changes 
in health, which may be seen as an ?epidemiological risk? factor for the development of eating 
disorders. Chile also observed an increase in the incidence of eating disorders during the socio-
 political changes that occurred during the 1970?s (Gordon, 2001).  
 
These considerations suggest that the mechanisms that link culture and eating disorders may be 
more complex than once believed and that the acculturation hypothesis may not be sufficient or 
even perhaps necessary as an etiological explanation of eating disorders. These considerations 
also suggest that conflicts and stress associated with socio-cultural transition may provide an 
important risk factor in the development of eating disorder, particularly in societies undergoing 
rapid socio-cultural transition.   
 
South Africa achieved black-majority rule in 1994, after years of white-minority political 
oppression and discrimination that were institutionalized in the racist policies of apartheid. 
Since the dissolution of apartheid in 1994, South Africa has also experienced a period of intense 
socio-political transition. Rapid urbanization has occurred and economic, employment and 
educational opportunities have escalated for the previously disadvantaged black community 
(Szabo & Le Grange, 2001). Previously disadvantaged black South Africans have also experienced 
upward social mobility and changes in the structure and functioning of families and the roles and 
expectations of women (Szabo & Le Grange, 2001); suggesting that apparent increases in 
dysfunctional eating and cases of eating disorder in South Africa may also reflect conflicts and 
stress that have been associated with rapid socio-cultural and political transition. 
 30 
2.2.4           Westernization or process of socio-cultural transition? 
 
The apparent rise of eating disorders in the west has coincided with a period of rapid social 
modernization that has occurred in the second half of the 20th century such as industrialization 
and urbanization which have been paralleled by technological advances, the emergence of a 
consumer economy, media saturation, individualism and achievement orientation, upward 
mobility, changes in the structure and functioning of families and the roles and expectations of 
women, and increasing intergenerational conflict (Gordon, 2001).  
 
Rathner (2001) noted that the 20th century has witnessed transformations ?from full-blown 
industrial society to globalization, from colonialism to post-colonialism, and from industrial 
production to the post-modern information age?. Westernization has also been ?dismantled into 
the changes pursuant to industrialization? including urbanization, female emancipation, 
individualism, economic transformations, the translation of social conflicts into individual 
medical morbidity, and the disappearance of traditional idioms of distress? (Katzman & Lee, 
1997); suggesting that socio-cultural transition may be the common denominator associated with 
the development of eating disorders in westernizing societies and that eating disorders may 
represent an idiom of distress or protest within this changing socio-cultural context. Important 
questions arise from this such as ?what do we mean by westernization of a society? Does this refer 
to the acculturation of particular ?western? ideas and behaviors or does it refer to a process of 
socio-cultural transition to a more modernized society?  
 
As data emerges from various cultures and contexts, these social and psychological tensions 
associated with modernization have also emerged as potential risk factors that are shared across 
cultures, suggesting that an increased risk of eating disorders may be associated with the process 
of socio-cultural change or modernization, as opposed to a simple compliance to a western 
cosmetic ideal; particularly in communities undergoing rapid social transition. 
 
 31 
2.2.4.1         Industrialization and Urbanization 
 
Japan has demonstrated a temporal increase in eating disorders that parallels the apparent 
increase in western countries, and similar patterns of weight preoccupation and dieting as 
western countries (Nakamura, Hoshino & Watanabe et al., 1999; Nakai, 2005). Japan was also 
the first non-western nation in contemporary history to become a major industrialized economic 
power and the rise of eating disorders in Japan has correlated with increasing industrialization, 
urbanization and consumerism and the fraying of traditional family forms (Gordon, 2001).  
 
The incidence of anorexia nervosa was also found to be four times higher in Japanese cities than 
in rural Japan (Azuma & Henmi, 1982; cited from Nasser, 1997) and three times more prevalent 
in densely populated areas of Japan (Ohezeki, Hanaki & Motozumi et al., 1990; cited from 
Nasser, 1997). Weight concerns were also found to be greatest in Hong Kong and least prominent 
in rural Hunan, suggesting that ?fat phobia? may be grounded in the global culture of modernity 
which is characterized in many rapidly urbanizing parts of the world (Lee & Lee, 2000).  
 
Ruggiero (2001) noted that the process of modernization in Italy has created two distinct 
territories, the industrialized north, and the rural South. The prevalence of eating disorders in 
the south was comparable to other western European countries while rates in the industrialized 
north were lower than the South. During that time, life in the south was in the process of social 
change in areas such as economy and female gender roles. Ruggiero et al., (2000; cited from 
Ruggerio, 2001) found that while southern girls were less sensitive to western influences on body 
ideals than their northern counterparts, they had lower self-esteem, greater emotional confusion 
and maturity fears. Ruggiero (2001) concluded that modernization places the individual in a 
position of conflict between traditional and modern expectations and may be accompanied by 
feelings of disconnection, social insecurity and lower self-esteem; which may be just as powerful 
a psychological determinant for eating disorder as body dissatisfaction and media influences.  
 
 32 
Becker, Burwell, Gilman et al., (2002) also demonstrated a dramatic increase in disturbed eating 
in adolescent girls during the years (1995-1998) following the introduction of western television 
to a predominantly agricultural, Fijian community undergoing rapid social and economic 
transition. Positive EAT scores increased from 12.7% (1995) to 29.2% (1998) and self-induced 
vomiting increased from 0% (1995) to 11.3% (1998). Becker (2004), however, found that the shift 
towards thinness in Fiji was more about economic and social competition than ideals of 
attractiveness as these young women saw thinness as a route to ?getting a job?. Becker concluded 
that Fijian women were attempting to reshape their identities by reshaping their bodies in 
accordance with television characters that were seen as role models for success in modern 
consumerist lifestyles.  
 
Fairburn (1997) noted that ?eating disorders are known to flourish at times when there is a need 
for self-change, particularly when a sense of competence and efficacy is lacking. At these times, 
individuals may attach themselves to role models that exemplify the ?new? social order in an 
attempt to navigate the changing social environment (Becker, 2004). Eating disorders have thus 
been seen as one of the most sensitive barometers of culture change and its impact on human 
beings (Di Nicola, 1990; Nasser, 1997; Nasser and Di Nicola, 2001). 
 
2.2.4.2         Socio-economic Status  
 
Modernization may also be associated with participation in a global market economy (global 
market capitalism) and upward economic and social mobility which frequently conflicts with 
traditional resources, which are ill-equipped to assist the individual through the changing social 
landscape. Socio-economic changes that take place in non-western and developing societies also 
promote an emphasis on education and achievement orientation; which has long been linked to 
the development of eating disorders.  
 
 33 
Towards the end of the 19th century in the west, anorexia was overrepresented amongst upper 
socio-economic groups which endorsed the idea that anorexia was a disorder of upper classes 
(Hoek, van Hoeken & Katzman, 2005) and perpetuated the notion of thinness as a symbol of 
competitiveness, achievement and success. The picture that emerges for bulimia nervosa is 
entirely different. Keel and Klump (2003) found that while anorexia nervosa (using a wider 
definition of anorexia) appeared more stable across different times, cultures and socio-economic 
groups, bulimia nervosa appeared to have increased against the background of westernization, 
industrialization, urbanization, upward social mobility, predisposition to overweight and 
increased accessibility to large quantities of food; suggesting that processes and stresses 
associated with social transition may be more important than mere exposure to western ideals of 
thinness in the development of bulimia nervosa.  
 
Contrary to former beliefs, the risk of eating disorders may thus actually be as high or even 
higher among ethnic minority and socially disadvantaged young women (Crago et al., 1996; 
Pumariega, Gustavson & Motes et al., 1994; Smith & Krejci, 1991). Ethnic minorities with eating 
disorders have also been found to have a slightly higher socioeconomic level than their non-
 eating disordered controls (Andersen & Hay, 1985), suggesting that upward social mobility may 
be more important in the development of eating disorders than social class per se (Anderson-
 Fye & Becker, 2003). Katzman, Hermans & van Hoeken et al., (2004) found that women who 
presented with anorexia on the island of Curacoa saw thinness as a sign of financial 
accomplishment and those who were more financially accomplished reported a greater desire to 
lose weight.  These women were also those who were ?straddling? several worlds simultaneously. 
They aspired to fit into the mobile elite subgroup yet were unable to feel acceptance within that 
group and they encountered frustrations and tensions when re-entering Curacoa after a time 
overseas. Katzman et al., (2004) concluded that eating disordered women were those caught 
between worlds of educational, social and economic aspiration, and the ?typical? reality of an 
island woman; supporting the idea that upward social mobility and transition may be more 
important in the development of eating disorders than social class per se. 
 34 
2.2.4.3         Obesity  
 
Early theorists accounted for the western cultural specificity of eating disorders in terms of the 
affluence and abundance of food in western and industrialized nations which were termed 
?societies of abundance? (Prince, 1983). As Selvini-Pallazoli (1985) argued ?people do not play 
games around food when availability is limited and the specter of hunger is pervasive?. A major 
post-industrial shift was, however, made from production towards consumption which brought 
changes in diet (fast foods) and lifestyle (reduced activity). This shift heralded increased obesity, 
particularly in American populations (Nasser, 1997); which heralded a concurrent rise in ?fat 
phobia? as a defining feature of eating disorders in the west (Hsu & Lee, 1993). 
  
Recognized risk factors for obesity in western societies include lower socio-economic status and 
minority group membership; suggesting that thinness may set upper classes apart from lower 
classes and be an outer sign of achievement, success and upward mobility (Nasser, 1997) in a 
developing economy. Where plumpness may have traditionally been valued as a sign of 
prosperity in a subsistence economy, thinness may therefore, herald success in a consumer 
economy. Conversely, obesity is stigmatized, resulting in acute tension between the factors which 
lead to weight gain in the general population and the forces that drive a need for thinness (Hebl 
& Heatherton, 1998); suggesting that the starving and binge-eating of bulimia may form a 
parallel to the production-consumption dialectic and represent the ?unity of opposites in 
contemporary society? (Fine, 1998; cited from Neumarker & Hein, 2001). 
 
A meta-analysis of epidemiological studies (Keel & Klump, 2003) found a highly significant 
increase in the incidence of bulimia nervosa in the west during the period 1970-1993 with a 
secular trend of increasing incidence in the non-west. These authors concluded that bulimia may 
well be a western ?culture-bound syndrome? which occurs as a result of western weight concerns 
against the background of industrialization and increased accessibility to large quantities of food.  
 35 
Gowers and Shore (2001) also noted that bulimia nervosa may be predated by binge-eating 
rather than dieting; with weight concern and then dieting ensuing. Purging behavior has also 
been found to be a common form of weight control in 21% of females at a low-income population 
family practice clinic (Martin & Wollitzer, 1988). Early onset obesity has also been associated 
with distorted body image and the development of eating disorders (Stunkard, 1959). The higher 
the weight among native -American women, the higher was the desire to lose weight and engage 
in weight-reducing measures (Rosen, Hafer, Dummer, et al., 1988).  
 
Negative attitudes towards obesity have also developed from recognition of associated health 
risks and the promotion of fitness. The ?thin ideal? has thus become synonymous with a trend of 
health, youthfulness and athleticism suggesting that the recent rise in weight consciousness in 
the west may be a reaction against obesity rather than mere compliance to a cosmetic ideal. 
Interestingly, a condition known as ?orthorexia nervosa? has recently been popularized to account 
for extreme obsessiveness around maintaining healthy or ?pure? eating patterns (Bratman, 2003), 
and a condition known as ?activity anorexia? has been proposed to characterize excessive 
exercising to achieve or maintain perfect body weight (Epling & Pierce, 1996); suggesting that 
new variants of anorexia may be emerging within a socio-cultural context of healthfulness rather 
than a simply cosmetic ideal of attractiveness.  
 
2.2.4.4         Gender Role Confusion    
 
During the latter part of the 20th century, increasing industrialization and urbanization has also 
heralded an increasing fragmentation of the family with an emphasis on the nuclear family that 
has demanded changes in the roles, expectations and identity of women and children; and has 
resulted in conflicts regarding gender roles and intergenerational conflict that have introduced 
great strain and confusion, particularly into the experiences of adolescents (Gordon, 2000). 
Changes in the social position of women has led to an increase in the number of educated and 
working women which requires transitions in female role and identity. Expectations of women in 
 36 
a male-dominated workplace, conflict sharply with demands for traditional feminine postures; 
promoting increased uncertainty, self-doubt, powerlessness and identity confusion (Gordon, 
2000, cited from Gordon, 2001). Feminist writers argue that contemporary ideals of ?thinness? 
may be a way of resolving conflict between two sets of contradictory social expectations for 
women, namely, traditional female stereotypes of dependency, passivity, frailty and helplessness 
along with the qualities required of the modern women such as autonomy, power, self- control, 
achievement and success (Nasser, 1997). ?Thinness promises that women can look like a woman 
and succeed like a man: the female superwoman that has it all (Gutwill, 1994)?. Silverstein and 
Perlick (1995; cited from Catina & Joga, 2001) coined the term ?gender ambivalence? to describe 
the ambiguities in the female role during periods of social transition which, they believe, leave 
women vulnerable to develop eating disorders. The pursuit of thinness is seen as metaphorical to 
a woman?s struggle to formulate a new identity in the face of changing social roles (Orbach, 
1986); and the degree of gender ambivalence is related to the degree of general social 
ambivalence, at  macro-economic and socio-political levels (Catina & Joga, 2001). Interestingly, 
when anorexia was first identified as a medical syndrome, both London and Paris were 
witnessing their first feminist movement (Nasser, 1997).  
 
The 1920?s was also a period of female emancipation in the west; when women were finally given 
the vote and many women started working outside of the home. ?A woman with a slender body 
distinguished herself from the plump Victorian matron and her old-fashioned ideas of 
nurturance, service and self-sacrifice? (Brumberg, 1988). The rail thin, androgynous style of the 
1920?s ?flapper? may, therefore, have reflected contradictions during the emancipation of women 
at the time and have been an expression of competition and success in a male-dominated work-
 place. Similarly, the introduction of oral contraception (1965) marked an era of sexual-liberation 
for women who were faced with contradictory expectations to be sexually liberated and 
aggressive yet also traditionally demure and naive.  Could it be co-incidence that Twiggy was 
launched as a super-model in 1966 to convey an image of ?a skinny kid with a face like an angel?? 
(www.twiggylawson.co.uk/ fashion.html). This official Twiggy website also notes that ?with her 
 37 
waif-like, slim figure and short boyish hair? Twiggy represented ?youthfulness and the new-found 
Sixties freedom?. Twiggy is also described as a ?reflection of an era? rather than the cause and may 
have been socially created as a pure, child-like, boyish, non-sexual alternative to the sexual 
(Hippie) revolution occurring at the time. These considerations suggest that a ?culture of 
thinness? may have emerged in western society as a solution to the psycho-social conflicts of 
women, rather than simply a cosmetic ideal. Interestingly Miss America (1967) was a similar 
weight (86.5% of expected body weight) to Miss America of 1959 (87% of expected body weight) 
and Miss America 1971, (88.5% of expected body weight), suggesting that although the weights of 
these popular icons decreased in the few years after Twiggy?s launch, Twiggy may not have made 
as significant an impact on ideals of thinness as popular literature would suggest. 
 
Pike and Borovoy, (2004) also found that shifting gender role expectations in contemporary 
Japan may contribute towards an increased risk of eating disorders. These authors found that 
division of labor in Japan has become increasingly gender?polarized such that marriage and 
motherhood are seen as the only possible route to female maturity yet also as a mechanism of 
social entrapment. Japanese women thus appear to have created some developmental space to 
delay maturity by the ?culture of cute?, portrayed in icons such as ?Hello Kitty? and the teenage 
schoolgirl figure (shojo) which is neither girl nor mother, neither asexual nor a fully realized 
woman (mother).  Conforming to this culture may thus represent a form of escape or protest 
against these gender-role expectations, reminiscent of Crisp?s account (1980) of eating disorders 
as a strategy for delaying maturation. These authors thus question the importance of western 
ideals of thinness as a reason for self-starvation, and point instead to local Japanese meanings of 
starvation (Lester, 2004); suggesting that western ideals of thinness may not be necessary as an 
etiological explanation for eating disorders, and that a more culturally sensitive search for the 
meaning of eating disorders is required. 
 
 
 
 38 
2.2.4.5         Socio-cultural Transition and Eating Disorders 
 
Periods of rapid social transition have been characterized as a ?social predicament? which is seen 
as a ?complex, unstable, morally charged and painful social situation? that requires new and 
adaptive socio-cultural mechanisms (Taylor, 1985; cited from Nasser & Katzman, 2003). Such 
changes may result in loss of certainty, disruption of familiar social networks and patterns of 
every day life, which may be a time of psychological vulnerability in countries experiencing rapid 
social transition. Confrontation with modernization has been seen as a? social predicament? 
(Nasser & Katzman, 2003) and especially disturbing for agrarian populations where exposure to 
new cultural expectations destabilizes the society?s traditional norms and requires new and 
adaptive socio-cultural mechanisms (Rathner, 2001).  
 
Jilek (2001) noted that imposed modernization may transform small tradition-directed 
communalist societies consolidated over many centuries, into modern mass societies and create 
an anomic situation for vulnerable groups. This may also lead to conflict between modern and 
traditional values and create cultural identity crisis and confusion (Jilek, 2001). Transitional 
processes have thus been seen to produce a number of ill effects on mental health (Rathner, 
2001) which may, in a hitherto tradition-directed population, lead to the emergence of types of 
psychosocial pathology previously unknown in that population? (Jilek, 2001). 
 
Rathner (2001) argued that, as hysteria was the most prevalent psychiatric disorder among 
women at the end of the 19th century, eating disorders may now be seen to reflect social pressures 
associated with westernization. Rathner (2001) continued to note that one of the main features of 
a western free-market economy is individualism, which ?tends to translate social conflicts into 
socially acceptable, individualized morbidities? such that social distress becomes articulated 
through medical and psychiatric terminology. Skarderud (2001) also argued that a major 
characteristic of eating disorder is the degree of concrete communication, whereby symbolic 
conflicts and boundaries are communicated and expressed through the explicit and concrete 
 39 
medium of the body (Skarderud and Nasser, 2007). The body may thus also become ?the 
dominant means by which the tensions and crises of society are thematized? (Turner, 1992; cited 
from Nasser and Di Nicola, 2001).  
 
Nasser & Katzman (2003) argued that, historically, during periods of cultural transition, the 
locus of power has been displaced onto the body; suggesting that eating disorders may be 
powerful markers of social change (Nasser & Di Nicola, 2001). Di Nicola (1990 b) described 
anorexia as a ?culture change syndrome? where women undergoing rapid socio-economic or 
cultural change may experience ?enormous acculturation stresses? which may place them at risk 
for anorexia nervosa.  According to this hypothesis, migrants and individuals in cultures that are 
in rapid economic and socio-cultural transition are especially vulnerable to eating disorders. 
Jilek (2001) proposed that culture-bound syndromes are not exclusively linked to a particular 
culture but rather related to a particular cultural emphasis which can become important in 
diverse societies at different historical periods. Nasser and Di Nicola (2001) concluded that the 
body ?has gone through phases and manifestations reflecting significant milestones in human 
history ? from the fasting of medieval saints to the secularization of the body into a commodity 
in our modern era, suggesting that body history may have mirrored human history?. Di Nicola 
(1990 a) noted the changing face of anorexia which he described as ?anorexia multiforme; a 
medical chameleon that changes with the times?.  
 
Indeed, even within the west, the ?thin? ideal appears to have become a symbolic expression of 
attributes such as power, and self-efficacy; self-control, assertiveness, autonomy and self-
 determination, competitiveness, achievement and sexual liberation, as well as a measure of 
beauty and attractiveness. Positive stereotypes such as beauty, health, class and success thus 
appear to have merged with ?normal? western dieting behavior (Nasser, 1997) and the body is 
seen as a commodity that is created by the individual such that the attainment of bodily 
perfection exemplifies individual self-regulation control and success which is equated with self-
 value and self-worth. Deliberate self-starvation has also been used as a demonstration of 
 40 
personal distress (Katzman & Lee, 1997) and protest (Brumberg, 1988). Given the plurality of 
meanings and motivations behind self-motivated starvation, Rieger, Touyz, Swain et al., (2001) 
recommend replacing weight phobia with ego-syntonic emaciation as a diagnostic criterion for 
anorexia nervosa. These considerations suggest that popular trends of body modification may 
reflect society?s conflicts during periods of social transition rather than simply a western cosmetic 
ideal and that eating disorder symptoms may be an expression of individual conflict and distress 
within a changing socio-cultural context.  
 
Nasser (1997) argued that the concept of Ethnic Disorder (Devereux, 1955; cited from Nasser, 
1997) may provide a more plausible alternative to the ?culture-bound? approaches to eating 
disorder as it is identified  as  a condition that occurs frequently within a particular culture and 
expresses core conflicts and psychological tensions that are pervasive in that society. Symptoms 
are seen as the final common pathway for the expression of a wide variety of personal problems 
and represent exaggerations of normal attitudes and behaviors that are prevalent in that culture 
that have become highly patterned and imitated ?idioms of distress.? This approach provides a 
theoretical framework for the individual-social dialectic such that ?culture? reflects core, conflicts, 
contradictions and tensions that are pervasive within in a particular society, while also providing 
an ?idiom of distress? for the expression of  individual problems.  
 
Ethnic Disorder may also embrace the broad array of cultural forces that are shared by a large 
number of societies rather than a particular geographic locale (Gordon, 2004). Following this 
approach, it may thus be speculated that the increased prevalence of eating disorders in western 
and non-western women during the 20th century may well have been underscored by stress 
encountered within a period of rapid modernization and socio-cultural transition that has taken 
place during this time. Stress encountered during the process of change or socio-cultural 
transition may thus be considered as an important risk factor for the development of eating 
disorders in modernizing or westernizing populations such as black women in post-apartheid 
South Africa. 
 41 
2.2.5           Cross-cultural Studies 
 
It was previously believed that eating disorders occurred almost exclusively in Caucasian women 
(Bruch, 1973). Reports of eating disorders appear, however, to be increasing in a wide range of 
ethnic groups (Miller, 2005). This has led to a plethora of cross-cultural studies, which the 
acculturation hypothesis has used as evidence that, with continued exposure to western culture, 
black women may be adopting western ideals of thinness; and that traditionally protective 
cultural factors may be rapidly eroding; leaving black women as vulnerable to eating disorders as 
white women (Perez & Joiner, 2003).  
 
Many studies have suggested that anorexia nervosa is quite limited amongst African-Americans 
(Striegel-Moore, Dohm, Kraemer et al., 2003; Taylor, Caldwell, Baser et al., 2007). Katzman et 
al., (2004) also found few, if any, cases of anorexia nervosa in the black population on the island 
of Curacoa. Hooper and Garner (1986) also found drive for thinness, body dissatisfaction and 
anorexia were more common in white than black Zimbabwean schoolgirls. Several studies have 
however, shown higher levels of disordered eating among African-American women than in their 
white counterparts. Black-American females have been found to demonstrate greater drive for 
thinness (Striegel-Moore, Schrieber, Pike et al., 1995) and are more likely to report fasting, 
recurrent binge-eating, and abuse of laxatives or diuretics than white-American women (Striegel-
 Moore, Wilfey, Pike et al., 2000). Normal-weight female black-American dieters have also been 
also been found to be more likely to use risky weight-loss strategies than their white-American 
counterparts (Beiner & Heaton, 1995). Pumariega et al., (1994) found that 53.6% of African-
 American women scored positively on the EAT; 38% engaged in binge eating, 3.5% purged by 
vomiting; and 16% abused laxatives. Dysfunctional eating patterns, with a tendency for bulimia 
was also found to be more prevalent in the African-Caribbean population in the United Kingdom 
than their white counterparts (Reiss, 1996) and more commonly reported in black females than 
anorexia nervosa (Nasser, 1997).  
 
 42 
Miller, Verhegge, Miller et al., (1999) found that 24.8% of a sample of rural native?American  
adolescents, scored positively on the EAT26 and a sample of rural native-Americans and 
Hispanics demonstrated that 87% were worried about being too fat; 11% met criteria for bulimia 
and 53% fasted for prolonged periods (Snow and Harris, 1989; cited from Miller & Pumariega, 
2001).  Rosen, Shafer, Dummer et al., demonstrated that 75% of native-American female dieters, 
were using risky weight-loss techniques. Smith & Krejci (1991) found that native-American 
adolescents demonstrated more symptoms of disturbed eating attitudes and behaviors than their 
white counterparts. Binge-eating was also found to be higher in native-Americans (14.2%) than 
white-Americans (10.1%), (Smith & Krejci, 1991).  Rosen et al., (1988) found that 74% of native 
Americans were trying to lose weight and concluded that native-Americans should be considered 
at risk for eating disorder.  
 
Meta-analysis of cross-cultural studies (Wildes, Emery & Simons, 2001) revealed that heavier-
 weight, black, high school populations may be particularly at risk for eating-related disturbances; 
suggesting that self-evaluation that is based upon thinness may thus be particularly detrimental 
to overweight, minority women (Abrams et al., 1993; Crago et al., 1996; Striegel-Moore et al., 
1995). Meta analysis also revealed that although white women tended to report greater levels of 
body dissatisfaction and eating disturbances than black women; whites did not score significantly 
higher than non-whites above the 90th percentile on measures of clinical pathology; and that the 
differences between groups, at the 99th percentile, was less than 1% in measures of restrictive 
eating. These findings suggest that while western cosmetic ideals may play a role in the 
development of sub-clinical eating disturbances, other factors may be needed to account for the 
onset of clinical eating disorder.  
 
Silber (1986) studied seven African-American and Hispanic adolescent anorexics in private, 
predominantly white schools, one of whom was the only black pupil in the school.  These girls 
were dissatisfied with their ?bigger bodies?; felt different, suffered from low self-esteem, and had 
a powerful need to be accepted and to ?fit in?;  which they sought through dieting and adopting an 
 43 
extreme social standard of slimness. Altering body weight may thus be a powerful way of 
accelerating acculturation and minority adolescents may ?overshoot? these western ideals in order 
to be accepted. Stressors associated with inter-ethnic experiences may, therefore, exaggerate 
prevailing acculturative pressures towards thinness and provide an important risk factor in the 
development of eating disorder in ethnic minority groups. Several other case reports of anorexia 
in black communities of the United States of America and the United Kingdom, reflect the 
psychological problems that girls struggle with as they try to achieve a racial-identity while fitting 
into a new society (Lacey & Dolan, 1988; Thomas & Szmukler, 1985) and it has been suggested 
that thinness has become a parameter of black achievement and social mobility within the wider 
(but thinner) American middle-class society (Schwartz, 1985). These considerations suggest that 
western cultural influences may not be sufficient to explain the development of clinical disorder 
in African-American women, and that individual (within-group) ?reactive? responses to inter-
 ethnic differences may be necessary to explain any differences in the prevalence of eating 
disorder in ethnic minority groups.  
 
2.2.5.1         Cross-cultural Studies in South Africa 
 
During the years of apartheid, black and white South Africans, were kept apart by 
institutionalized racist policies in almost every aspect of daily life. Since the dissolution of 
apartheid in 1994, South Africans of all ethnic groups are engaging in every aspect of human 
activity and sharing their diverse cultural backgrounds in neighborhoods, schools, work, leisure 
and sporting activities.  
 
In 1996, Szabo and his colleagues (Szabo & Allwood, 2004 a & b) conducted a survey in a sample 
girls in recently integrated urban high schools, and found equal numbers of white (18.6%) and 
black (18.7%) students scored positively on the EAT26. Earlier, Szabo and Hollands (1997) found 
that black female adolescents in an exclusive private school, scored a markedly higher percentage 
of positive EAT scores (37.5%) than their white peers (20.67 %). These scores were also higher 
 44 
than those of schoolgirls in Canada (22%; Leichner, Arnett & Rallo, 1986); North America 
(17.5%; Fischer, Pastore & Schneider, 1994) Britain (18.6%; Cooper & Goodyer, 1997) and Nigeria 
(18.6%; Oyewumi et al., 1992); and closer to prevalence rates recorded for high risk groups such 
as dancers (38%) and models (34%), (Garner & Garfinkel, 1980). During apartheid, this school 
had been an elite ?all-white? school which, at the time of the study, had become racially 
integrated. Szabo and Hollands (1997) speculated that the ?impact of acculturation on a young 
black girl entering a formerly whites-only school may be akin to that experienced by other 
minority groups? in other formerly white, western-dominated contexts who have also been found 
to demonstrate a greater risk of eating disorder than their white counterparts (Pumariega, 1986; 
Mumford, Whitehouse & Platts, 1991). Szabo and Hollands (1997) concluded that the increased 
risk of eating disturbances found among the black students in their study may be due to 
acculturative stress ?particularly in view of the major sociopolitical changes that South Africans 
are experiencing.? The authors commented on the need for further study on acculturative stress 
as a risk factor for eating disorders in South Africa.  Interestingly, although desire to be thinner 
and attempts to lose weight predicted higher mean EAT scores in this study, these attitudes and 
behaviors were found to be almost universal within the sample. Following Patton, Johnson-
 Sabine & Wood (1990) these findings suggest that dieting may be a relatively benign practice in a 
large percentage of school girls, and that the progression from dieting to disorder may require 
other vulnerability factors. Following Silber (1986), acculturative stressors associated with 
interethnic contact may promote an over-identification with western cultural norms and conflict 
necessary to promote the progression from vulnerability to disorder.  
 
Le Grange et al., (1998) also found that a sample of black college students scored significantly 
more positive scores on the EAT40 than their white peers. In an item which asked whether 
respondents had ever had a serious eating difficulty in the past, 14.3% of whites, 15% of blacks 
and 13% of mixed race and 3.5% of Asians responded affirmatively. The authors concluded that 
higher levels of disturbance in black subjects might be due to exposure to extreme pressures 
associated with westernization. Wassenaar, Le Grange and Winship et al., (2000) found that 
 45 
while white women showed the highest body dissatisfaction on the Eating Disorders Inventory 
(EDI: Garner et al., 1983) black women showed the highest drive for thinness, perfectionism and 
maturity fears. White subjects demonstrated significantly lower body mass index (BMI) than 
black subjects and higher BMI correlated with body dissatisfaction and drive for thinness. 
Following Davis and Katzman (1997), these authors speculated that high BMI, along with high 
scores on perfectionism and drive for thinness among black subjects, suggest that these subjects 
may be attempting to alter their body weight (high BMI) in order to adapt to western cultural 
expectations. They continued to speculate that high scores on these indices, amongst black 
subjects, may reflect the psychology of cultural transition and that acculturative stress may 
explain their results, particularly in view of dramatic socio-political changes that have taken 
place in South Africa since the lifting of apartheid in 1994. 
 
Silber (1986) also notes that altering their body weight may be a powerful way for black females 
to reduce inter-ethnic differences and be accepted; suggesting that  higher-weight black South 
African girls may be at particular risk for eating disorder in the multiethnic context of the new 
South Africa. Higher BMI in young black females has been associated with serious dieting and 
vulnerability for dysfunctional eating attitudes and behavior. Senekal et al., (2001) found that 
while black females were more satisfied with their body shape than white subjects in other South 
African studies, higher BMI was associated with poor self-concept, body dissatisfaction, dietary 
restraint and higher mean EAT scores. Caradas et al., (2001) also found that increasing BMI was 
associated with increasing body-shape concerns; although  a comparable percentage of black and 
white adolescents had positive scores on the EAT. These authors concluded that their findings 
?reinforce the notion that eating disorders are culture-reactive rather than culture-bound? and 
argued that ?considering the near epidemic proportions of obesity in black women?, obesity may 
indeed be a risk factor for eating disorder in young black westernizing South African women?.  
Reactive responses to inter-ethnic and inter-cultural differences may thus be considered as a risk 
factor for the development of eating disorder in black South African women, particularly those 
who are obese and in a context of social comparison with white western peers. 
 46 
2.2.6           Immigration Studies 
 
Most initial case reports of eating disorders were of individuals from different ethnic and cultural 
backgrounds who had emigrated or escaped to western countries (Lacy & Dolan, 1988) and who 
developed a disorder within a short time of their arrival (Kope & Sack, 1987), fuelling speculation 
that their eating disorders were due to exposure to white western culture (Nasser, 1997). Some 
studies have also demonstrated that non-western women who have emigrated to a western 
culture may present a higher risk of eating disorder than their peers in their homeland (Nasser, 
1986; Fichter et al, 1983) and have been used as evidence for the acculturation hypothesis. 
 
Further study has revealed however, that ?western? concern with body weight and symptoms of 
eating disorder may have been existent in non-western societies presumed to have different 
values (Abdollahi & Mann 2001; Shuriquie, 1999) and have highlighted the effects of 
encountering a ?different? social milieu by demonstrating an increased risk of eating disorders in 
trans-location between different ?western? countries (Van Den Broucke & Vandereycken, 1986). 
Furukawa, (1994) found that Japanese exchange students reported weight gain and disturbed 
eating attitudes during their stay in a new country, regardless of whether their host country was 
weight conscious or not. After six months in the host country, symptoms of bulimia were 
significantly correlated with interpersonal distrust, ineffectiveness and with scores on the 
General Health Questionnaire-30 (Goldberg, 1972; cited from Furukawa, 1994); suggesting that 
stresses encountered in a ?different? social milieu may be associated with an increased risk of 
eating disorder; whether this milieu is western or non-western. Furukawa concluded that non-
 specific acculturative stresses may mediate disturbed eating in vulnerable individuals.  
 
Other studies have demonstrated that immigrants may be even more at risk for eating disorder 
than their western counterparts in the host culture (Fichter et al., 1983; Mumford et al, 1991; 
Waller & Matoba, 1999; Furnham & Adam-Saib, 2001). Furnham and Alibhai (1983) found that 
Kenyan women in Britain favored thinner body shapes than Kenyan women in Kenya and 
 47 
reacted more positively to thinner shapes than British women. The researchers argued that 
exposure to different, western ideals may result in an over-identification with western values of 
thinness and the adoption of more extreme ideals of thinness. Nasser (1986) found that 22% of 
Arab students in London scored positively on the EAT-40 which approached rates reported in 
high risk groups such as dance students (38%) and modeling students (34%), (Garner % 
Garfinkel, 1980).  Moreover, 12% of the London group qualified for a diagnosis of bulimia 
nervosa, which was the highest rate of bulimia reported in Britain at the time. Nasser concluded 
that the London students may have become more competitive and achievement-orientated in 
reaction to their exposure to western culture and over-identified with western ideals of thinness; 
suggesting that factors encountered during the process of intercultural transition may be 
associated with an increased risk of eating disorder. Mumford and Whitehouse (1988) also found 
that Asian schoolgirls in Britain scored a higher percentage of positive EAT scores (15%) and a 
higher rate of bulimia (3.4%) than their British peers (12% positive EAT scores and 0.6% 
bulimia). Jackson et al (2006) found that female Korean immigrants to America scored 
significantly higher EAT26 scores than American-born, Korean-Americans but similar scores to 
native Koreans in Korea. The authors note that Korea has been undergoing rapid socio-economic 
changes since World War II, suggesting that high scores among immigrant and native Korean 
women may reflect common stressors encountered during rapid social transition.  
 
 These considerations suggest that exposure to western cultural ideals of thinness may not be 
sufficient to explain the development of eating disorders in migrant women, and that factors 
associated with inter-cultural contact and transition may be necessary to explain the risk of 
eating disorder in this group. Di Nicola (1990 b) introduced the concept of anorexia as ?culture-
 reactive? to explain the incidence of eating disorder in non-western women exposed to western 
influences. Under these circumstances, non-western immigrants may demonstrate an extreme 
reaction against their own culture, such that acculturation stress may be a key factor in triggering 
eating disorder in immigrants and other communities undergoing rapid socio-cultural transition 
(Di Nicola, 1990 b). 
 48 
2.2.7           Acculturation Studies 
 
Traditionally, research has attempted to demonstrate a relationship between individual levels of 
acculturation to western culture and the prevalence of eating disorder, using criteria and 
instruments designed to measure level of acculturation to western culture. Studies that 
demonstrate an association between assimilation of western culture and an increased incidence 
of eating disorders have also been used as evidence for the acculturation hypothesis.  
 
A number of studies have found a correlation between level of acculturation to western culture 
and morbid concerns over weight and dysfunctional eating amongst immigrant and black 
populations in the United States of America (Abrams, Allen & Gray, 1993; Davis & Katzman, 
1997; Chamorro & Florez-Oritz, 2000; Ball & Kenardy, 2002). Other studies have found no 
association between level of acculturation to western culture and dysfunctional eating attitudes 
or behaviors (Abrams, Allen & Gray, 1993; Furnham & Patel, 1994; Akan & Grilo, 1995; Joiner & 
Kashubeck, 1996; Abdollahi & Mann, 2001). 
 
Synonomous with the belief that eating disorders are related to the acculturation of western 
norms is the belief that strong identification with culture of origin provides some protection from 
the effects of westernization. Some studies have supported this belief (Pumariega, 1986; Jane, 
Hunter & Lozzi, 1999; Humphrey and Ricciardelli, 2003). Other researchers have found that a 
strong identification with culture of origin may even be associated with an increase in eating 
disturbances, in particular settings (Lake et al., 2000; Tsai, Curbow & Heinberg, 2003). Meta-
 analysis of acculturation studies (Wildes, et al., 2001) demonstrated inconclusive findings with 
some studies indicating that acculturating individuals who identify with western culture 
demonstrate a higher prevalence of eating disorders; while other studies indicated that those 
who remain strongly identified with their culture of origin demonstrate a higher prevalence of 
eating disorders than those who identify with western culture.  
 49 
Wildes et al., (2001) concluded that limited and discrepant research findings provide no support 
for a positive relationship between acculturation and eating pathology. Jackson, Keel & Lee 
(2006) also found that female Korean immigrants to America scored significantly higher on the 
EAT26 than second-generation Korean-American women and that EAT scores were unrelated to 
measures of western acculturation. These authors concluded that their results contradict the 
acculturation hypothesis and that any understanding of eating disorders needs to look beyond 
the influence of western acculturation to the stresses encountered during rapid socio-cultural 
transition. These authors also speculated that the increased levels of disordered eating found in 
Korean immigrants may reflect conflict between traditional Korean and Western values, thereby 
supporting the acculturative stress hypothesis. Lake et al (2000) found that Chinese-Australian 
immigrants who were highly identified with traditional Chinese culture were significantly more 
dissatisfied with their bodies than those who identified with western culture. The authors 
speculated that elevated body dissatisfaction found in traditional Chinese-born subjects may be 
due to a clash between western cultural expectations and a strong traditional orientation (culture 
clash).  Tsai et al., (2003) speculated that traditional women may compare their physical 
appearance to western girls around them (the ?girl-next-door hypothesis?) and feel invalidated 
and dissatisfied with their appearance if they are not similar.  
 
These findings suggest that the acculturation of western ideals may be not be sufficient to 
account for the development of eating disorders in subjects exposed to western culture and that 
the stressors involved in the process of acculturation, particularly factors involved in the 
differences or the ?clash? between cultures in inter-culture contact, may be an important risk 
factor and may account for the apparent increased prevalence of eating disorders in westernizing 
populations. Di Nicola (1990 b) emphasized the importance of the process of acculturating to a 
new environment where different and conflicting socio-cultural influences may be a trigger for 
acculturative stress.  
 
 50 
Traditionally, research on acculturation has, however, attempted to define and measure 
acculturation as a uni-dimensional continuum (Suarez-Orozco, 2001) which assumes that change 
is a unidirectional progression towards complete assimilation by the dominant culture and that 
the adoption of values and practices of the dominant culture will be accompanied by a 
corresponding loss of their original culture (Trimble 2003). In recent years, this model has been 
called ?uni-linear? because it describes only one type of acculturation, namely assimilation, 
(Flannery, Reise & Yu, 2001) and ?an individual is placed somewhere between a traditionalist 
pole to a fully acculturated position? (Trimble, 1989). Following this tradition, many studies 
assume that low ethnic identity is synonomous with high acculturation and visa versa; and no 
allowance is made for a bicultural option or that the process of acculturation may interact with 
other socio-cultural or psychological variables. These models also imply that what is occurring is 
an unquestioning ?absorption? of western values and images; although literature suggests that 
women are more complex and strategic, and may select portions of the dominant culture that fit 
their original worldview, whilst at the same time striving to maintain parts of their traditional 
culture (Trimble, 2003).  
 
More contemporary views also argue that it cannot be assumed that complete assimilation will be 
the outcome, and challenge the unidirectional view of acculturation for forcing and an inverse 
relationship between two dimensions that they see as orthogonal (Flannery, Reise & Yu, 2001). 
Measures used to determine acculturation on a single, linear dimension are thus seen as 
inappropriate (Sue, 2003) and orthogonal models have eclipsed more linear understandings of 
the process of acculturation.  
 
 
 
 
 
 
 51 
2.3             Dieting To Disorder 
 
In order to account for the pathogenesis of eating disorders, the culture-specific approach needs 
to demonstrate that exposure to western culture is necessary and sufficient as a causal 
explanation for eating disorders. To do this, this approach needs to provide significant causative 
associations between specific western ideals of thinness; dieting behaviors as normative for 
western culture; and eating disorders as the extreme form of this culturally normative behavior.   
 
Evidence for this approach has been drawn from sub-populations and occupations who place a 
greater focus on their body shape and weight such as jockeys, dancers and models and athletes 
and who have been found to demonstrate higher levels of body image disturbances, dieting, over-
 exercise, vomiting to control their weight, positive EAT scores (N=30% in models) and anorexia 
nervosa (N=7%), (Garner & Garfinkel, 1980). Imm and Pruitt (1991) found that high-frequency 
exercisers had greater body dissatisfaction than moderate exercisers and tended to pursue 
exercise even when feeling ill.  Hsu (2005) questions whether the increased incidence amongst 
these subgroups is a result of characteristics of the profession or of the people attracted to these 
professions and argues that ?research needs to understand the genetic or epigenetic mechanisms 
that drive the individual to find fulfillment in starvation within that particular environment or 
subculture.?  
 
A meta-analytic review (Groesz et al., 2002) found main effects that body satisfaction is 
significantly lower after viewing media images depicting ?thin images? and that the thinner the 
model, the greater the degree of body dissatisfaction, particularly in females under the age of 19 
years. This may be mediated by stimulating social comparisons and using the body as a ?quick 
and concrete barometer by which to measure oneself and one?s worth? (Rhodin et al., 1985). 
These findings are used to support the view that women use media images as a reference for 
evaluating their own body image and that these images are internalized and contribute towards 
body dissatisfaction, dieting and ultimately eating disorders (Stice et al., 1994). 
 52 
Meta-analysis (Groesz et al., 2002) of research on the effects of media have, however, 
demonstrated little to no immediate effect of thin media images on body satisfaction and other 
studies have produced inconsistencies within a single study This meta-analysis suggested that 
the experimental effect may be seen most clearly in the activation rather than cultivation of a 
thinness schema.  Hamilton & Waller, (1993) found that eating disordered women were affected 
by idealized body images but control women were not affected by the nature of the pictures; 
suggesting that while there may be an association between media images of thinness and eating 
disorders, it may be that women who already have disordered eating, are more vulnerable to 
these images of thinness than non-eating disordered women. Studies that have attempted to 
identify the media in creating the ?prevailing pressure to be thin? in western communities, have 
thus been inconclusive (Okasha, 2005) and the specific mechanisms that mediate between media 
images, dieting and eating disorder remain poorly understood. Becker?s (2004) study also 
suggested the role of the media in eating disorders must be studied in tandem with the 
socioeconomic changes that configure the everyday lives and concerns of young people (Lee, 
2004). 
 
The association between prevalent dieting and eating pathology is also far from clear and is the 
subject of continuing debate (Nasser, 1997). The culture-specific approach argues that eating 
pathology lies on a continuum of severity with dieting at one end of the spectrum and the 
extreme forms of disorder at the other end (Nylander, 1971; Drewnowski, 2005). At the dieting 
end of the continuum, girls may simply score high on a screening test, while at the other they 
may qualify for a full-blown, life-threatening clinical anorexia or bulimia nervosa. In this 
spectrum, EDNOS would be the first stage of disease; suggesting an area of great importance 
from a public health and preventive perspective. Longitudinal studies (Drewnowski, 1988) 
suggest that incident cases of bulimia nervosa do arise largely from the EDNOS population. 
Conversely, that cases which no longer meet diagnostic criteria for full-blown disorder, may 
return to a diagnosis of EDNOS for a prolonged period. Dieting and disorder are thus seen to 
only differ in degree rather than kind; and disorder is seen as an extreme form of this normative 
 53 
behavior (Nylander, 1971). Research has indeed found that dieting is prevalent in 50-80% of 
young women who see themselves as overweight and report exaggerated concern with their 
weight (Nylander, 1971) and that, within the course of one year, dieters among adolescent 
schoolgirls, have an eight-fold increased risk of developing an eating disorder than non-dieters 
(Patton 1990). Others have argued that clinical eating disorders are categorically distinct from 
sub-clinical syndromes (Wonderlich, 2005) such that anorexia and bulimia represent 
qualitatively distinct entities or taxons rather than the opposite poles of a single continuum 
(Wildes et al., 2001); which contrasts with the dimensional approach and requires an 
explanation for the appearance of pathological eating disorders in a population of apparently 
?normal? dieters.  
 
 Schleimer (1983) found no direct causal link between dieting and anorexia which questions the 
basic premise of the continuum hypothesis. Schleimer conducted a 10 year prospective study and 
found that those girls who dieted and went on to develop anorexic behavior, had shown such 
behavior at an earlier age or signs of mental insufficiency. Patton et al., (1990) also questioned 
whether there is a cause and effect sequence between dieting and the development of eating 
disorders. These authors concluded that in the great majority of girls, dieting is essentially a 
benign practice without natural progression to eating disorder and yet the influences on dieting 
behavior (including weight and shape concerns) are the same in those with and without eating 
disorders; suggesting that factors other than media influence and dieting may be required to 
explain the progression from dieting to disorder.  
 
Patton, Coffey and Sawyer (2003) argue that partial syndromes are also more common in 
adolescent females, but are usually brief and self-limiting. Subjects who progressed to eating 
disorder tended to report anxiety and depression. Graber, Tyrka and Brooks-Gunn, (2003) 
concluded that co-existing depression may link sub-clinical forms of eating disturbance along a 
continuum to disorder; suggesting that factors other than dieting may be required to account for 
the progression from dieting to disorder.  
 54 
Botta (2000) demonstrated that black and white American girls watch similar amounts of 
television and similarly idealized female characters, compared their own bodies with these ideals 
and desired to be thin. Carney and Louw (2006) found that, while media exposure was associated 
with dysfunctional eating in South African college students, media exposure was only one of 
numerous factors that may predispose young women to pathological eating. They concluded that 
the interaction of many factors may be necessary to cause disorder. Cooper and Goodyear (1997) 
also found that while significant weight and shape concerns were present in 11-14 year-old, these 
concerns were only associated with significant, concurrent behavioral and ideational disturbance, 
similar to older bulimics,  in 15-16 year olds. The authors suggest that longitudinal studies are 
needed to identify the etiology of these disturbances in this age group. 
 
Halmi  (2005) concluded that dieting seems to be a common risk factor across the centuries and 
that while dieting behavior may play a role in the pathogenesis of clinical disorder, only a small 
proportion of all dieters proceed to develop a disorder (Hsu, 2005). Dieting, may, therefore be 
seen as a mediating factor in the development of the full-syndrome eating disorders; but the 
question remains unresolved as to what determines the progression from dieting to extreme 
dieting or partial disorder, to a definite clinical case (Nasser, 1997). The culture-specific 
hypothesis may thus be insufficient to account for extreme individual variations in ?normative? 
dieting, and transition from dieting to disorder may require a multi-factorial explanation, as it is 
likely that only those who are vulnerable to these disorders will respond to normative demands 
with symptoms of clinical eating disorder (Halmi, 2005). Conversely, while the Generalist 
approach may account for eating disorder symptoms as the ?social coloring? of underlying, 
universal pathology, it does not provide sufficient account for the underlying causes of eating 
disorder in particular, or the occurrence of drive for thinness and eating disorder, outside of a 
westernized context. Eating disorders are now thus generally considered to have a multi-factorial 
origin, with genetic, biological, developmental, personality, family, environmental and socio-
 cultural factors predominating as risk factors and the interaction of several risk factors is 
believed necessary for the development of an eating disorder (Gowers & Shore, 2001) 
 55 
2.3.1           Predisposing factors 
 
Genetic heritability has been estimated as approximately 50% for both anorexia nervosa (Kaye, 
1999) and bulimia nervosa (Kendlar, MacLean, Neale, et al., 1991). Other eating disorders, 
depression, substance abuse, anxiety disorders, particularly obsessive-compulsive disorder 
(Lilenfeld, Kaye, Greeno et al., 1998), personality disorders, particularly obsessive-compulsive 
personality traits such as perfectionism (Strober, 1995) are also frequently found in the family 
history of eating disordered individuals.  
 
Pre-and-co-morbid presentation of these factors is also frequently associated with the diagnosis 
of eating disorder (Fichter & Pirke, 1995; Treasure, 2007).  Suggested pre-morbid biological 
characteristics have included obesity (Striegel-Moore et al, 1995) disturbances of the leptin-
 melanocortinergic system (Hinney, Remschmidt & Hebebrand, 2000) and dysregulation of the 
serotonergic system, which is involved in the regulation of mood and appetite (Kaye, 1999). 
Personality traits such as perfectionism and obsessive-compulsiveness have also been suggested 
(Anderluh, Tchanturia, Rabe-Hesketh et al., 2003) where perfectionism may be associated with 
chronic dissatisfaction with body weight and shape (Slade, 1982), negative self-evaluation and 
low self-esteem; which have also been implicated in the development of eating disorders 
(Fairburn, Cooper, Doll, et al., 1999). Obsessive-compulsive traits may also be associated with 
rigidity, and inflexible thinking that is characteristic of anorexia nervosa (Goldner, 
Srkameswaran, Schroeder et al., 1999). Competitiveness and achievement-orientation have also 
been reported as pre-morbid risk factors (Garner & Garfinkel, 1980). Neuro-cognitive deficits 
such as cognitive inflexibility and over-attention to detail (Tchanturia, Anderluh, Morris, et al., 
2004) have also been suggested as possible pre-morbid dysfunctions, and season of birth (Eagles 
et al., 2001), prenatal and obstetric complications have been implicated (Favero, Tenconi & 
Santonastaso, 2006). Alexithymia has also been suggested as a stable pre-morbid characteristic, 
and may be associated with difficulties differentiating the somatic and emotional aspects of 
emotion, concrete thinking, and a paucity of fantasy (Schmidt, Jiwany &Treasure, 1993b). 
 56 
Earlier theories of etiology focused on the family as the microcosmic environment that may 
contribute towards the development and maintenance of an eating disorder. Family attitudes and 
beliefs around weight and dieting have been found to contribute towards risk of eating disorder, 
where over-concern with weight was more evident in the mothers of eating-disordered girls (Hill 
& Franklin, 1998). Vandereycken (1995) sees the characteristic family as having difficulty gaining 
the right balance between parental control and age-appropriate autonomy?. Over-protectiveness, 
rigidity, enmeshment, lack of conflict resolution (Minuchin, Baker & Rosman et al., 1978) and 
high-concern parenting (Shoebridge & Gowers, 2000) have also been associated with the 
development of eating disorders. Selvini-Pallazoli (1974) focused on marital disharmony and 
trans-generational coalitions and Szmukler, Eisler, Russell, et al., (1985) noted the lack of 
expressed emotion in families of anorexics. Other theorists have noted the association between 
eating disorders and insecure parental attachment, separation or loss of a parent through illness, 
death or divorce and sexual abuse (Eisler, 1995).  Bruch (1973) argues that in many cultures food 
occupies a very special place in family and social interactions and has become associated with 
degree of parental nurturance, thereby promoting universal conflicts around food and food-
 refusal as a symbol of distress or medium of control within family relationships. 
 
Adverse life events may also contribute towards the development of eating disorders. These 
events may contribute towards vulnerability for weight and shape consciousness or, less 
commonly, act as a trigger for the onset of disorder in vulnerable individuals (Gowers & Shore, 
2001). Some studies have found higher than expected rates of past physical and sexual abuse in 
eating disordered individuals (Welch & Fairburn, 1996), while others have found that these rates 
are probably no higher than those in other psychiatric disorders (Palmer & Oppenheimer, 1992). 
Freud saw difficulties with sexuality as a core conflict in the development of anorexia, where 
dietary restriction and emaciation was seen as a symbolic attempt to rid the individual of sexual 
drives, while others have focused on the negative effects of early sexual abuse on the 
development of self-esteem (Tice, Hall, Beresford et al., 1989). 
        
 57 
2.3.2           Precipitating Factors 
 
Adolescence has been seen as a particularly vulnerable stage for the development of eating 
disorders (Gowers & Shore, 2001) as it is a time when a number of physical and psychosocial 
challenges converge. Hormonal changes herald a sudden pre-pubertal increase in orexins 
(Treasure & Holland, 1995) with consequent deposition of body and an average increase from 8% 
body-fat in childhood to 22% after puberty (Tanner, 1989); the onset of puberty and escalating 
sexuality. Adolescence is also associated with the development of independence, autonomy and 
self-identity (Erik Erikson, 1959); as well as increasing social demands for mature 
responsibilities.  At the same time, the development of formal operations and abstract thought, 
allows the adolescent to reflect consciously on himself, often leading to morbid introspection, 
self-consciousness and self-criticism. Gowers and Shore (2001) suggest that these challenges 
may have a significant impact on confidence, self-esteem and sense of personal control and may 
lead to difficulties with individuation and the development of an independent ego-identity.  
Precocious puberty has been associated with elevated prevalence of eating disorders, where 
pubertal increases in body weight was associated with negative body image, body dissatisfaction 
and increased weight and eating-related concerns (Koff & Rierdan, 1993). 
 
Negative life-events and trauma may also trigger a disorder in vulnerable individuals (Gowesr 
and Shore, 2001). Negative comments about weight (Hill & Pallin, 1998), racial teasing (Iyer & 
Haslam, 2003) and social comparison (Silber, 1986) have also been cited as precipitants for 
dieting.  Some researchers argue that anorexics pursue dieting as an indirect way of coping with 
or expressing a wide range of problems or conflicts that they are unable to resolve directly 
(Katzman & Lee, 1997). Following Di Nicola (1990b), these findings suggest that conflicts and 
stresses involved in the process of intercultural contact and social transition may provide a 
relevant risk factor in the development of eating disorders in the rapidly transforming landscape 
of  black South African women and that adolescents may be particularly vulnerable. 
 
 58 
2.4             Contemporary Models of Acculturation  
 
Contemporary constructs of acculturation have moved beyond a focus on the cultural domination 
of one nation by another and inter-cultural contact is seen to potentially generate change in 
either or both groups or may be reactive such that cultural change is rejected (Berry, 2003). 
Contemporary models have thus defined and measured acculturation as a multidimensional and 
bi-directional process that needs to be described in terms of two cultural orientations; namely an 
individual?s relationship towards a new culture, whilst also allowing for a relationship towards 
their traditional or original culture (Berry, 2003; Ward, Bochner & Furnham, 2003)  
 
2.4.1           Dimensionality and Directionality  
 
Berry (1997, 2003) defined acculturation as a complex, multidimensional process that includes 
many levels of acculturative influence and potential change, including ecological, group and 
individual dimensions. This model also identifies acculturation as a continuous, interactive and 
bi-directional process between an established or traditional society and a ?new? socio-cultural 
environment or group, whereby aspects of both may be differentially selected and integrated by 
either group or the individual, at different stages of inter-group experience or phases of life. 
Acculturation refers to the process of adjusting to these changes, which is seen to lead to 
increased stress, known as acculturative stress and the greater the disparity between the ?new? 
and original society, the greater the potential for acculturative stress (Berry and Annis, 1974). 
                          
2.4.1.1         Ecological Dimension 
 
Berry (1976, 2003) described the ecological dimension as those aspects of the physical and social 
environment that provide the setting for socio-cultural change and are seen to influence change 
at the level of the group, such as weather patterns, demographic distribution of the population 
and national political, social and economic institutions and policies. Ecological change such as  
 59 
urbanization may, for example, be accompanied by the organization and stratification of society 
on the basis of role specialization and authority systems which are defined by a set of social 
expectations and govern individual interaction. Rural, extended families may thus be 
differentiated into nuclear family structures where the ?head? of the family is expected to be 
economically and socially independent of his relatives; and women may be required to play a 
more active role within the economic and authority systems of the family.  This dimension may 
thus be especially important when considering the sweeping ecological changes that have 
occurred with modernization during the 20th century.  
 
2.4.1.2         Group Dimension 
 
The group dimension is described as those socio-cultural components which may be influenced 
by ecological components or contact with a different cultural group and which include different 
cultural, social, linguistic, technical and religious systems (Berry, 2003).  Ecological changes such 
as industrialization and urbanization may, for example, be accompanied by different ?cultural 
syndromes? such as collectivism versus individualism, which may reflect differences in values 
such as loyalty towards the community or family versus individual goals; cooperation versus 
competition; authority and obedience versus autonomy and self-reliance; compliance versus 
assertion, responsibility versus achievement orientation; and nurturance and interdependency 
versus independence (Ward et al., 2003). These dimensions are also reflected in the norms of 
social interaction and the socialization of children.  
 
Bochner (1994) equated the socio-cultural dimensions of collectivism and individualism with 
non-western versus western cultural patterns, suggesting that the group dimension may be 
important when considering intercultural contact between western and non-western cultural 
groups and the globalization of western values, ideals and practices; particularly the exposure by 
non-western cultural groups to western ideals of beauty and attractiveness, and its possible 
associations with eating disorder. 
 60 
2.4.1.3         Individual Dimension 
 
Berry (1997, 2003) distinguished between change that may occur at the level of the group and the 
level of the individual, who needs to cope with new sets of social expectations, attitudes, values 
and behaviors. The individual is seen to demonstrate variable degrees of participation in the 
general acculturation of the group. ?Not every individual enters into, participates in, or changes 
in the same way and there are vast differences in psychological acculturation even between 
individuals within the same acculturating arena ?(Berry, 2003).  
 
Graves (1967; cited by Berry, 2003) also distinguished between acculturation as a ?group 
phenomenon? and ?psychological acculturation? which refers to a change in the functioning of the 
individual who is in culture-contact. Berry (1997) noted that psychological acculturation refers to 
individual changes in identity, values and behaviors that may occur during contact with another 
culture.  Other researchers have distinguished between behavioral acculturation as participation 
in the customs and habits of the culture, and value acculturation where the values of the culture 
are adopted (Mendoza, 1984; cited from Ward et al., 2003). An acculturating individual may also 
differentially integrate specific beliefs and practices. They may, for example, reject religious 
practices, assimilate dress customs and integrate food preferences or they may select traditional 
dress when at home but more contemporary dress when at work. Padilla (1986) proposed that 
acculturation may be seen as two dimensions: cultural awareness, where the individual 
demonstrates knowledge of particular cultural variations and ethnic loyalty, which refers to the 
individual?s preference of cultural orientation.  
 
Ward et al., (2003) have categorized individual responses into behavioral, affective and cognitive 
components, which have been associated with three contemporary approaches: 
 
 
 
 61 
Drawing from Argyle?s (1969; cited from Ward et al., 2003) work on social skills, the behavioral 
dimension highlights the importance of culture-learning in the development of socio-cultural 
adaptation and focuses on social skills such as language competency in predicting effective 
intercultural adaptation. Adaptation is, therefore, achieved by learning the culture-specific skills 
needed to manage everyday social encounters. 
 
Drawing on research on stress and life-events (Lazarus & Folkman, 1984; Holmes & Rahe, 1967; 
cited by Ward et al., 2003), the affective dimension focuses on the emotional components of 
intercultural contact. This approach highlights the significance of cross-cultural transition as a 
series of life-changes that are intrinsically stressful and require appraisal, selection and coping 
strategies to deal with them.  Successful adaptation is seen to be mediated by both individual and 
situational variables and the model identifies affective, psychological (or psychopathological) 
consequences of failing to adjust to the new culture (Ward et al., 2003).     
 
Drawing from literature on self identity and theories of social cognition (Tajfel & Turner, 1986; 
cited by Ward et al., 2003), the cognitive dimension focuses on social cognition and changes in 
cultural identity during the process of acculturation.  Phinney, Chavira and Williamson et al., 
(1992) noted that individuals in bi-cultural contact need to address two important questions. 
?Who am I?? and ?How do members of my group relate to other groups? Following Phinney et al., 
(1992), these questions focus on two, interrelated levels; the first entails the recognition, 
categorization or definition of oneself as a member of an ethno-cultural group and the second 
focuses on inter-group relations, perceptions and attitudes.   
 
In a review of research, Phinney (1990) outlined that ethnic identity may be seen as the ethnic 
component of social identity, which, following Tajfel and Turner (1986), refers to that aspect of 
an individual?s self-concept that is derived from membership in a social group and the emotional 
significance attached to that membership. Phinney (1990) argues that ethnic identity refers to 
that aspect of acculturation which focuses on how the individual relates to their own group and  
 62 
includes self-identification with a social group (or groups), along with a sense of acceptance and 
?belongingness?, pride and positive evaluation of one?s group, and involvement in ethno-cultural 
values and traditions. Phinney (2003) also perceives ethnic identity as a core concept which 
directly links group membership to self-definition and self-esteem such that the effects of culture 
contact may be classified in terms of changes in the individual?s identity when exposed to 
heterogeneous cultural influences. Phinney (1989) also reported a positive relationship between 
level of ethnic identity and psychological adjustment. Acculturation theory has thus focused on 
identity changes that occur in response to intercultural or interethnic contact and constructs of 
cultural identity and acculturation have been used interchangeably as if they are synonymous; 
such that one may be measured from the other (Ward et al., 2003).  
 
2.4.1.4         Directionality 
 
Contemporary models (Berry, 1997, 2003) describe acculturation as a bi-directional process 
which may result in changes in either or both groups and whereby the individual needs to 
negotiate between the ?old? and ?new? socio-cultural realities in terms of a relative balance 
between the two. Acculturating individuals must therefore negotiate between these two disparate 
realities and the tension between them is potentially stressful and may result in acculturative 
stress (Berry, 1997). Differentiation and integration are seen as the basic mechanisms involved in 
the process of acculturation (Berry, 1976) which requires the generation of novel behaviors. 
?Acculturation thus basically entails learning to cope with a new cultural situation? (Sam & 
Oppedal, 2002) which may require the integration of discrepant or even mutually-opposing 
fields or scripts into a novel, yet compatible formulation. Directionality may thus be particularly 
important when considering the possible associations between socio-cultural transition and 
eating disorders, where women may be more at risk as they attempt to ?juggle conflicting 
demands and expectations of two different cultural worlds? (Katzman & Lee, 1997).   
 
 
 63 
Escobar and Vega (2000) argued that conflict predictably occurs when two or more cultural 
groups come into direct contact with one another and experience change, and that it is how 
groups and individuals deal with this conflict that continues to be an important research 
question.  These authors concluded that acculturation refers to both process and outcome as 
acculturation cannot be understood as a simple reaction to changes in the cultural context, but as 
an active dealing with these challenges. Acculturation has thus been defined as a major life-
 change event that is characterized by stress, demands cognitive appraisal of the situation and 
requires adaptive coping strategies.   Acculturation is thus seen as a more complex process than 
the ?simple? transmission of western ideals and may be accompanied by stress that is deeply 
rooted within the process of acculturation itself (Berry, 1997). 
 
Berry?s (1997, 2003) bi-directional conceptualization of acculturation asserts that acculturation 
needs to be described in terms of two cultural orientations; namely an individual?s relationship 
towards a new culture whilst, at the same time, allowing for a relationship towards their original 
culture. Berry (1997, 2003) described four different modes or strategies of acculturation available 
to individuals facing these acculturative challenges, namely: assimilation, when individuals reject 
their original cultural identity and identify with the new culture; separation,  when individuals 
maintain their traditional cultural heritage and reject the new culture; marginalization,  when 
individuals  reject both groups or vacillate between the two cultures, feeling  not entirely ?at 
home? in either; referred to as the ?marginal syndrome? (Park, 1928, cited in Ward et al., 2003); 
and integration, when the individual has positive attitudes towards both groups and seeks an 
integration of their own cultural heritage within the larger, super-ordinate, social framework.  
 
Traditionally, integration has been associated with reduced levels of acculturative stress 
(Krishnan & Berry, 1992; Schmitz, 1992), and has been seen as the most psychologically adaptive 
outcome of bi-cultural acculturation. In contrast, marginalization has been associated with the 
most stressful outcome (Berry & Sam, 1996) and the most likely to result in chronically 
dysfunctional and deviant behaviors (Schmitz, 1992). Strategies of assimilation and separation  
 64 
are seen to lie between these extremes and may be differentially stressful depending upon the 
context. Separation has also been positively associated with drug and alcohol addiction (Schmitz, 
1992) and chronic stress (Berry & Kim, 1988).  
 
At face value, the strategy of integration sounds like a very compatible solution. As Sam and 
Oppedal (2002) pointed out, however, there are no guidelines that discuss specifically what 
happens during this bi-cultural process, and how an adaptive outcome is achieved; particularly in 
children and adolescents who are growing up in bi-cultural environments with two very different 
or mutually exclusive cultural scripts. If the individual identifies with the new society, this may 
lead to a weakening or abandonment of ties with their culture of origin; however, if they resist 
adopting the practices of the new or dominant society, they risk marginalization. Adaptation 
requires the internalization of different cultural scripts that guide behavior under different 
cultural circumstances; and an important part of this includes learning when it is appropriate to 
switch between different scripts.  
 
Rodriguez et al., (2002) also identify acculturation as a bi-directional process where the 
individual feels both a push towards acculturating to the ?new? culture and a ?pull? towards 
identification with their original ethnic group. These opposing pressures are conceptualized as 
pressures to assimilate to the new culture (Pressure to Acculturate) and to retain traditional 
orientation (Pressure against Acculturation) which are identified as potentially stressful 
(acculturative stress). Following Berry (1997), failure to integrate these opposing cultural 
expectations into an integrated or adaptive multicultural identity may result in chronic 
acculturative stress. Identification with two different cultures is, however, likely to be 
problematic for identity formation in ethnic group members where there is conflict between the 
attitudes, values and behaviors of the two groups (cited from Phinney, 1990). ?The issue is 
whether individuals must choose between two conflicting identities or can establish a bicultural 
ethnic identity, and if so, whether this is adaptive? (Phinney, 1990). 
  
 65 
Sodowsky and Lai (1997) propose that of living with two sets of opposing cultural expectations 
leads to a sense of inefficacy, identity crisis, a personal sense of inferiority as a member of one?s 
own cultural group, lack of ethnic ego differentiation due to feeling marginalized from both 
cultural groups and feelings of anger and guilt toward one or both cultural groups. Extreme 
bicultural conflict has also been associated with chronic acculturative stress which may be 
manifested by feelings of emotional turmoil and alienation (Sue & Sue, 1990) cultural 
marginality (Sodowsky, Kwan & Pannu, 1995), poor self-concept (Padilla et al., 1985), depression 
(Draguns, 1996; cited from Ward et al., 2003) and anxiety (Sue, 1996).  
 
Keifer (1974; cited from Roysircar-Sodowsky & Maestas, 2000) identified three types of bi-
 cultural conflict experienced by Asian Americans. Cultural alienation: where disruption of 
familial cultural patterns is associated with a sense of personal discontinuity and poor self-image. 
Cultural confusion: where an individual is unable to identify with a definite norm within a 
context of multiple norms. Cultural conflict: where an individual?s values are perceived to be 
incompatible with a given social situation. Kim (2000) also identified identity crisis that occurs 
in later-generations as ?a situation where an individual perceives certain aspects of themselves 
that are simultaneously rejected?. Sue and Sue (1990) call on the concept of the ?marginal person? 
to describe the predicament where the individual is ?living between the margin of two different 
cultural traditions? and suffers identity crisis.  
 
Research has demonstrated inconsistent findings and even inconsistent data within the same 
study. Some studies suggest that identification with both referent cultures (integration) is 
associated with positive psychological outcomes such as self-esteem (Phinney, Chavira & 
Williamson, 1992) and less acculturative stress (Gil, Vega & Dimas, 1994). Ying (1995), however, 
found that while a bicultural orientation predicted better mental health than separation, 
separation predicted lower negative affect and higher life satisfaction than assimilation in the 
same study. Furnham and Li (1993) found that greater use of English and involvement in British 
society was associated with greater symptoms of frustration, inadequacy and depression amongst 
 66 
 second-generation Chinese immigrants in Britain; while first-generation subjects who were 
more separated from British society, experienced greater psychological symptoms.   Mena, 
Padilla and Maldonado (1987) demonstrated a strong inverse relationship between stress levels 
and ethnic loyalty, endorsing the view that links to one?s ethnic community may be crucial for 
mental health. Ward and Kennedy (1994) also found that immigrant identification with host 
culture was associated with enhanced psychological adjustment, when identification with culture 
of origin was strong, but with decrements in psychological wellbeing when identification with co-
 nationals was weak; supporting suggestions that continued identification with culture of origin 
may act as a protective buffer against negative effects of identifying with a new culture. Phinney 
(1989) also found a positive relationship between level of ethnic identity and psychological 
adjustment, specifically measures of ego identity, sense of mastery, social and peer relations in 
minority groups in the United States of America. Phinney (1990) argued that the relationship 
between ethnic identity and self-esteem may be moderated by the emotional salience of group 
membership for the individual and where the individual consciously perceives ethnicity or 
culture as a salient feature of his identity. Research has suggested that both cognitive and 
affective components of identity are more strongly aroused in minority groups when under 
perceived threat and they may experience a stronger need for in-group identification than 
members of a privileged majority.  Rotheram-Borus (1993; cited from Ward et al., 2003) found, 
for example that almost half of the minority students attending an integrated school in New York  
identified themselves as bi-cultural; while over 70% of minority students in a less ethnically 
balanced school, exhibited racial tension and identified themselves with their heritage group.  
 
A meta-analytic study (Rogler, Cortez & Malgady, 1991) found an equal number of studies that 
support a positive and negative relationship between identification with the ?new? culture and 
mental health.  Sayegh and Lasry (1993) concluded that acculturation styles had no effect on 
indices of acculturative stress, although individuals who adopted the strategy of assimilation 
reported less discrimination than those in the ethnocentric and integration groups.  
 
 67 
Oh, Koeske and Sales (2002) found that while level of acculturation (language use and social 
association) was related to lower levels of acculturative stress amongst Korean immigrants to the 
United States of America, those reporting abandonment of original ethnic identity, traditions and 
values (cultural identity) scored higher on depression. The authors concluded that attempts to 
operate in two different cultures, carries the risk of increased stress and depression.   
 
Ward and Rana-Deuba (1999) also found that level of acculturation is differentially related to the 
psychological and socio-cultural aspects of adaptation such that strong identification with culture 
of origin was associated with significantly lower psychological distress while strong association 
with the host culture was linked to fewer social difficulties.  Sojourners who adopted a strategy of 
integration, experienced significantly less psychological distress than others, while those who 
preferred assimilation reported less social difficulty. Assimilation was predictive of psychological 
stress, separation was predictive of psychosomatic stress; and integration was predictive of 
overall stress.  
 
Rudmin (2003) criticizes the validity of the constructs of integration and marginalization and 
concludes that there is no robust evidence that biculturalism is most adaptive. Indeed, it appears 
that theoretical understanding of bicultural acculturation is still evolving and demonstrates 
confusion regarding exactly what the most adaptive outcome may be or exactly how the 
individual may achieve this.   
 
Flannery, Reise and Yu (2001) introduce tri-directional model that argues that bidirectional 
change in both groups may result of in the development of a new culture with its own distinctive 
identity. This process of ethno-genesis introduces a third cultural dimension which is seen to be 
particularly pertinent in third-generation immigrants; while unidirectional and bidirectional 
models may be more appropriate for first generation and second generations.  
 
 
 68 
2.4.2           Stages of Acculturation  
 
Berry and Kim (1988) also argue that individuals that are exposed to another, usually dominant 
culture pass through predictable stages of acculturation. Early stages require relatively little 
adjustment while later stages may be accompanied by uncertainty and anxiety. In the phase of 
conflict, the individual is confronted with discrepancies between the demands of their own 
traditional system and those of the new culture and feels unable to find an effective resolution. 
During the stage of crisis the individual tests different acculturative strategies in order to find an 
adequate coping strategy. These latter stages have a high degree of uncertainty and are highly 
stressful. Once an adaptive strategy is found, this stress is reduced. If an adaptive strategy is not 
found; acculturative stress may become chronic (Berry & Kim, 1988).  
 
Helms (1990), argues that black-Americans are influenced by competing Afro-American and 
?white? American ethno-cultural groups and that during the process of racial identification, they 
pass through a series of stages: Pre-encounter, where ?white?-American is idealized and they 
reject their own race. Encounter, where the individual encounters lack of acceptance in the 
?white? world and struggles to find a new identity. Immersion/Emersion, where the individual 
rejects the ?white? world and immerses himself into the Black world view and Internalization, 
where the individual acknowledges himself as primarily Black but uses both groups to shape a 
perspective that best fits his life circumstances. Thompson, Anders0n and Bakeman (2000) 
found that racial identity may play a mediational role between racial socialization and 
acculturative stress and that both immersion and internalization predicted acculturative stress. 
 
Phinney (1990) noted that ?it is how conflicts between two distinct cultures are accommodated by 
the individual that is part of the process of ethnic identity formation? and it is the psychological 
consequences of this process that determines the individual?s ultimate psychological adjustment 
within a multicultural society.   
 
 69 
2.4.3           Acculturative Stress and Psychopathology 
 
Recognition that acculturation may be stressful can be traced back to reports of transient 
psychotic reactions that were identified in the period of industrialization and mass-urbanization 
that took place in 19th century Europe (Jilek, 2001); and were described as ?folie hysterique? in 
Paris (Morel, 1860; cited from Jilek, 2001) and ?amentia transitonia? in Vienna (Meynert 1889; 
cited from Jilek, 2001). The term ?emotional psychosis? was also used to describe transient 
psychotic reactions experienced by rural Italian migrants under acculturation pressure in Swiss 
cities (Jilek, 2001).  
 
An association between migration and mental health was identified early in the 20th century 
(Chun, Organista & Marin, 2003) and by the mid 1970?s, research held an implicit assumption 
that migration was inextricably linked with an increased risk of developing a wide range of 
physical and mental pathologies (Chun et al., 2003). In early years, this increased psychiatric 
morbidity in migrants was attributed to the belief that people who were dysfunctional were more 
likely to emigrate. The idea that it might be migration that is stressful was first introduced in the 
literature by Park (1928; cited from Berry, 1997) and Stonequist (1937; cited from Ward et al., 
2003) whose illustration of the ?Marginal Man? was an attempt to comprehend the social and 
psychological conditions of persons caught between two cultural systems; being ?poised in 
psychological uncertainty between two worlds? (Berry, 1976). The term ?culture shock? was also 
coined by Oberg (1960; cited from Ward et al., 2003) to describe emotional reactions associated 
with intercultural contact. Eisler (2001) notes that, for many, society in transition may be 
compared to the ?culture shock? experienced during migration, only, instead of moving from one 
culture to another, they have the experience of the culture migrating around them. Social 
marginalization incurred during imposed westernization has indeed been associated with alcohol 
abuse and juvenile suicide in many of the aboriginal First Nations of North America (Jilek, 1974 
cited from Jilek, 2001) and in some regions of South Pacific (Jilek, 1987; cited from Jilek, 2001). 
Cases of transient psychoses in Africans have also been attributed to stress associated with the  
 70 
process of acculturation and marginalization due to rapid socio-cultural change in Africa (Jilek & 
Jilek-Aall, 1970: cited from Jilek, 2001). Jilek (2001) also identifies ?anomie depression? among 
indigenous peoples of North America encountering westernization. Nigerian students have also 
been found to suffer psychosomatic symptoms of stress or ?brain fag? when exposed to western 
educational systems which are different from traditional African ways of acquiring knowledge 
(Prince, 1985).  
 
By 1987, Berry and his colleagues had defined acculturative stress as ?a reduction in the health 
status of individuals that may include physical, psychological and social symptoms that occurs 
when an individual?s adaptive resources are insufficient to support an adjustment to a new 
cultural environment (Berry & Annis, 1974, Berry, 1997). To qualify as acculturative stress, these 
symptoms needed to be related in a systematic way to known features or changes involved in the 
acculturation process (Berry, Kim, Minde & Mok, 1987). These authors noted that this stress 
tends to manifest as a particular set of stress responses such as anxiety, confusion, depression, 
feelings of marginality and alienation, psychosomatic symptoms and identity confusion; which 
may be identified as acculturative stress. Schmitz (1992) proposed that acculturative stress may 
also manifest as feelings of distress, homesickness, depressive reactions, psychosomatic 
complaints and vulnerability to disease, psychological maladjustment and psychopathic 
behavior, including aggressiveness, norm violence, drug, and alcohol abuse. Acculturative stress 
has also been empirically associated with general distress (Neff & Hoppe, 1993), depression and 
suicidal ideation (Hovey & King, 1996), substance and alcohol abuse (Zimmerman & Sodowsky, 
1993), wife assault (Jasinski, 1998) delinquent and criminal activity (Vega, Gil, Warheit, et al., 
1993, cited from Ward et al., 2003) and symptoms of eating disorder (Perez et al., 2002).  
 
Berry and Annis (1974) argued that mental health problems are not inevitable and depend upon 
a variety of group and individual moderating factors or variables which enter into the 
acculturation process. By the late 1980?s, theoretical focus thus shifted away from the prevailing 
clinical model, with its view of ?culture shock? as a medical problem (Ward et al., 2003) and  
 71 
contemporary theory focuses on acculturative stress as a potentially adaptive ongoing, dynamic 
process rather than a static association between migration and pathology; which has allowed for 
the investigation of acculturation in a wider range of circumstances, including sedentary and 
indigenous communities (Ward et al., 2003). Contemporary approaches have continued to 
regard intercultural contact as a stressful life-event and to describe the adverse psychological 
consequences of failing to adjust (Ward et al., 2003); but discuss intercultural contact in terms of 
coping with stress, which is seen to be mediated and moderated by characteristics of the 
individual and the situation. Contemporary approaches thus propose a bidirectional model of 
acculturation (Ward et al., 2003; Berry, 2003) where pressures emanate from both culture of 
origin and culture of settlement and successful adaptation is seen as dependent upon a number 
of situational and individual variables (Figure 1). 
 
 
 SOCIETY OF ORIGIN                                                  SOCIETY OF SETTLEMENT 
                        Social Factors                                                                   Social Factors 
                        Political Factors                                                                Political Factors 
                        Economic Factors                                                             Economic Factors 
                        Cultural Factors                                                                Cultural Factors 
 
 
SOCIETAL LEVEL VARIABLES 
 
 
 
              CROSS-CULTURAL                 STRESS AND                 RESPONSES                      OUTCOME 
                     TRANSITION                          SKILLS                         Affective                            Psychological 
                 Life Changes                                DEFICITS                      Behavioral                          Sociocultural 
                 Intercultural Contact                                                             Cognitive 
 
 
 
INDIVIDUAL LEVEL VARIABLES 
 
 
                                        Characteristics of the Person             Characteristics of the Situation 
Personality                                            Length of cultural contact 
 Language fluency                                 Amount of intra and inter- 
                                            Training & experience                          group contact 
                                            Cultural Identity                                   Quality of intra and inter- 
                                            Acculturation strategies                        group contact 
                                            Values                                                   Amount of life changes 
                                            Reasons for migration                           Social Support 
 
                Figure 1: Model of Acculturation and Stress (Ward et al., 2003), 
 72 
 2.4.3.1         The Type of Acculturating Group  
 
Berry (1997) argued that acculturating groups vary in terms of their degree of voluntariness, 
mobility and permanence such that while some have voluntarily migrated, others such as 
indigenous people have had the new culture involuntarily imposed upon them. In a comparative 
study of acculturative stress, Berry et al., (1987) found that ?native? (indigenous) peoples scored 
the highest levels of acculturative stress, followed by refugees, sojourners and then voluntary 
immigrants; supporting the prediction that individuals who are involuntarily exposed to cultural 
change suffer more acculturative stress than those doing so voluntarily and that indigenous or 
?native? populations encountering enforced westernization may experience the highest levels of 
acculturative stress. These findings also suggest that acculturative stress, as experienced by 
indigenous non-western populations encountering westernization, may not necessarily follow the 
patterns identified in most research which has been conducted in minority immigrant 
populations. Theoretical expectations based upon immigrant samples thus needs to be 
extrapolated with caution to the majority black population in South Africa; where westernization 
has been imposed on a predominantly non-western, indigenous, majority population. 
 
The few research samples of indigenous people have demonstrated a negative association 
between stress and level of education, wage employment, ownership of western goods, English 
literacy and media use (Berry, Wintrob & Sindell et al., 1982). Prior intercultural experiences, 
participation in the larger society and having lived in an urban (usually multicultural) as opposed 
to a rural setting (usually uni-cultural), have also been found to be predictive of lower levels of 
stress (Berry, 1976). Berry and Kim (1988) contend that pre-acculturation contact and experience 
may provide information about the ?other? which may lead to a sense of ?know thy enemy? and 
control. These findings suggest that high levels of acculturative stress may be experienced by 
black South Africans, who, until the dissolution of apartheid, were largely kept apart from 
contact with white, western South Africans by segregationist national policies and were relatively 
excluded from western economic and educational opportunities. 
 73 
2.4.3.2         The Nature of Inter-group Contact 
 
Berry (2003) argues that the process and outcome of acculturation may be influenced by national 
policies which may influence the quality of inter-group contact, including: Melting Pot (voluntary 
assimilation; Pressure Cooker (forced assimilation), Segregation or Ethnocide. Segregation refers 
to the deliberate exclusion of certain groups from mainstream positions, institutions and 
territories by discriminatory policies or social norms; or where a group seeks to separate 
themselves into distinct states, reserves, cultural enclaves or special institutions.  
Multiculturalism refers to the policy where cultural diversity is accepted and respected by the 
larger society (Berry, 2003) and may be associated with less stress than Assimilationist societies, 
where there are pressures to conform to a single cultural standard (Murphy, 1977). The most 
notorious example of Segregation in modern times has been the policy of apartheid which existed 
in South Africa in the fifty years prior to democracy in 1994. During this time, a white, western 
colonial power dominated the black African indigenous majority with a Segregationist policy 
which ensured that different races and cultures were kept apart in almost every aspect of human 
interaction. Following Berry (2003), this policy may influence the process and outcome of 
acculturation in post-apartheid South Africa where, although the black majority power has 
adopted a policy of Multiculturalism, many relics of apartheid appear to remain.  
 
Triandis et al., (1986) found that the more power the immigrant group has in its new setting, the 
less the new group will assimilate to the norms of the larger or new group. Relative status and 
power or numerical balance (majority-minority) of the acculturating groups may therefore also 
moderate the experience of acculturation and acculturative stress (Ward et al., 2003; Padilla & 
Perez, 2003). Sheer numbers of people in the group of origin (Berry, 2003), vitality (Phinney, 
2003) and demographic density (Berry & Annis, 1974) of the ethnic community may also 
increase the preference for cultural maintenance. Following these authors, majority black South 
Africans may not choose to assimilate ?white? minority western culture in post-apartheid South 
Africa and may prefer to maintain their traditional cultural heritage. 
 74 
 
Studies have consistently demonstrated that the greater cultural difference/dissimilarity between 
groups, the more difficult the adaptation (Berry, 1997). Ward et al., (2003) note that 
distinguishing physical characteristics of the participant cultures may also moderate 
acculturative stress, whereby the greater the physical differences between participants, the more 
difficulties they will experience (Furnham & Bochner, 1982; Berry, 1980). Nguyen, Messe & 
Stollak (1999) found that high levels of psychological distress amongst traditional Vietnamese 
youth living in a predominantly white American environment. Chataway and Berry (1989; cited 
from Ward et al., 2003) also found that Chinese students in an English-speaking Canadian 
university reported more anxiety than French-speaking Canadian-born groups. These findings 
support suggestions that the greater the differences between groups, the greater the degree of 
acculturative stress and suggest that higher levels of acculturative stress may be experienced by a 
black minority group in a ?white? majority South African setting.  
 
2.4.3.3         Duration of Acculturation 
 
Berry and Kim (1988) argue that early phases of acculturation (pre-contact and initial contact) 
may require relatively little adjustment while later phases may be accompanied by uncertainty 
and anxiety before a suitable adjustment is established. Schmitz (1994) demonstrated that 
acculturative stress was positively correlated with the stages of conflict and crisis. Persistent 
acculturative stress has also been found to be the strongest predictor of poor long-term mental 
health (Nicholson, 1997; Organista et al., 2003). 
 
Rodriguez et al., (2002) argue that acculturation is an ongoing process and that while recent 
immigrants may experience greater pressures to assimilate to the new culture (Pressures to 
Acculturate), later generations may experience more pressures to maintain their culture of origin 
(Pressures against Acculturation) as they become more distanced from their traditional cultural 
heritage. These authors found that Pressure to Acculturate was significantly associated with  
 75 
psychological distress but, contrary to expectation, was not correlated with any age, duration of 
stay, generation, or level of acculturation; suggesting that regardless of level of acculturation, 
these immigrant subjects continued to perceive that their cultural values conflicted with those of 
mainstream American culture and continued to experience Pressures to Acculturate and 
acculturative stress. Pressure against Acculturation was perceived to be most stressful by later 
generations and by individuals who were less competent in the traditional language and 
identified more strongly with Anglo-American culture. Pressure to Acculturate was also 
significantly associated with Pressure against Acculturation, and, according to these authors, was 
?most likely due to the continued push-pull effect? between these two opposing pressures, 
emanating from two disparate cultures.  This may persist even into second generation 
immigrants where children may be exposed to the new culture at school, but may live in a 
traditional family setting at home, where customs, values, roles, food, child-rearing and 
disciplinary practices completely conflict with traditions and practices of the host culture. This 
may lead conflicts between the first and second generation immigrants.  Padilla, Alvarez, and 
Lindholm (1986) found that second generation subjects experienced significantly more stress 
than later generations; with third and later generations experienced the least stress. 
 
Padilla and Perez (2003) note that while social awareness of the culture of origin may decline 
with generation, ethnic loyalty, which depends on self-ascribed ethnicity, may remain 
consistently high from the first to the fourth generation. Atkinson, Morten and Sue (1983: cited 
from Ward et al., 2003) describe a pattern that typically develops over successive generations. 
The first generation is often separatist, retaining a strong sense of identity with heritage culture, 
the second generation more often assimilate with the host culture, and the third generation often 
emerges with a renewed interest in and identification with, ethnic customs, values and practices. 
South Africa is in its first generation of racial integration, suggesting that most black South 
Africans may maintain identification with their ethnic group and adopt a separatist position.   
  
 
 76 
Ward et al., (2003) distinguish between socio-cultural (social competency such as language skill), 
and psychological levels of adaptation. The former is seen to follow a learning curve that 
improves rapidly in the earliest stages of acculturation, reaches a plateau, then stabilizes; (Ward 
& Kennedy, 1996b). In contrast, psychological distress tends to fluctuate over time with a 
tendency to peak in the earliest stages of transition (Ward & Kennedy, 1992). Padilla and Perez 
(2003) also argue that social competence may precede changes in social identification. These 
findings suggest that Pressures to Acculturate may be not be associated with duration of  
intercultural experience in black South Africans while Pressures against Acculturation may be 
associated with increasing duration of  intercultural contact and westernization. Socio-cultural 
adaptation such as language competency may also precede psychological adaptation. 
  
2.4.3.4         Characteristics of the Acculturating Individual  
 
Younger migrants have been found to be more malleable than older ones and tend to suffer less 
acculturative stress (Rogler et al., 1991; Marin, Sabogal & Vanoss et al., 1987; Rodriguez et al., 
2002) than those who emigrated after 14 years (Padilla et al., 1986); and adolescent migrants are 
known to be particularly at risk (Sam & Berry, 1995). Adolescents tend to experiment with their 
identities and may be simultaneously influenced by peer pressure (Pressure to Acculturate) and 
familial conflict (Pressure against Acculturation) over changing attitudes and behaviors. Conflicts 
between the demands of parents and peers may be maximal at this period and transition from 
childhood to adulthood may be compounded by cultural transitions (Berry, 1997).  
 
Developmental issues of identity also come to the fore at adolescence and may be complicated by 
issues of ethnic identity (Phinney, 1989). Phinney (1990) outlined a number of stages in the 
development of ethnic identity, namely the initial stage of ?diffusion? where ethnicity is not seen 
as a salient issue and has not been adequately explored; the stage of ?awareness?, which occurs as 
a result of experiences which force the individual to examine issues of ethnicity and identity; and 
the stage of ?acceptance and internalization? with the achievement of a clear sense of ethnic  
 77 
identity. There is evidence to suggest that these stages may proceed in an age-related pattern 
where a greater proportion of pupils in grade 10 (adolescents 15 -17 years) are actively engaged in 
an identity search than adolescents in grade 8 (13 -15 year-olds) and that a coherent sense of 
ethnic identity is more frequently internalized by older adolescents than by high school students 
(Phinney, 1989 & 1990). Cross cultural transitions made in later life present different challenges. 
A strong sense of identity with heritage culture may be well established and resistant to change. 
Separatist strategies adopted by parents may also significantly influence attitudes towards 
change in their offspring and thereby influence adolescent ethnic identity and acculturation 
patterns (Sam, 1995). Women also tend to report greater acculturative stress than men 
(Ghaffarian, 1987) and retain a stronger sense of identity with culture of origin (Liebkind, 1996). 
These findings suggest that black South African female adolescents may be most at risk for 
intergenerational conflict and the development of identity confusion and acculturative stress. 
 
Family dysfunction, ineffective social support and low self-esteem also been associated with 
acculturative stress (Hovey, 2000) while family closeness and support may provide a buffer 
against acculturative stressors (Rogler et al., 1991); suggesting that black South African 
adolescents may be most at risk when traditionally supportive and protective family structures 
have been eroded.   Lazarus and Folkman (1984) contend that a life change is only a stressor if it 
is interpreted as such by the individual experiencing it. Individual differences in appraisal of 
these experiences may thus be influenced by differences in personality and cognitive processing.  
Following Berry (1997, 2003), acculturation requires the ability to successfully integrate 
disparate cultural orientations, suggesting that various cognitive and personality dimensions 
may moderate the level of acculturative stress (Schmitz, 1992). Field-dependent cognitive styles 
and an external locus of control (Berry et al., 1987) have been shown to mediate more stressful 
reactions to situations. Neuro-cognitive rigidity may also promote dichotomous thinking and 
difficulty with the cognitive integration of conflicting ideas ((Tchanturia et al., 2004; Katzman et 
al., 2004), suggesting that South African adolescents who are unable to integrate opposing 
cultural scripts may be most at risk. 
 78 
2.4.4           Measurement of Acculturative Stress 
 
The construct of acculturative stress is still in a process of evolution.  Many conceptualizations of 
acculturative stress have thus been proposed and acculturative stress has been operationalized in 
terms of a wide range of measures including, specific stressors and difficulties associated with the 
process of acculturation; the potentially stressful reactions to these stressors; and  mode of 
acculturation. Research on acculturative stress has thus been confounded by lack of 
methodological uniformity in directly assessing acculturative stress (Rodriguez et al., 2002).  
 
2.4.4.1         Stressors associated with the Acculturation Process  
 
?Difficulties and stressors arising as part of the process of adaptation and acculturation to a new 
culture constitute the construct of acculturative stress? (Joiner & Walker, 2002). Following this 
conceptual focus, most of these measures have focused on the process of migration and have thus 
included a wide variety of different stressors pertinent to migrants, ranging from separation from 
family, unemployment and lack of housing to minority status as likely sources of acculturative 
stress (Rodriguez et al., 2002).  
 
The Social, Attitudinal, Familial and Environmental Acculturative Stress Scale (SAFE: Padilla, 
Wagatsuma & Lindholm, 1985) is an example of this type of measure. Joiner and Walker (2002) 
found that acculturative stress (as measured by the SAFE-24) was significantly related to 
depressive and anxiety symptoms and after controlling for general life stress. The authors 
concluded that there is something distinctive about acculturative stress that may be 
distinguished from general life stressors?; thereby supporting the validity of the construct of 
acculturative stress and the SAFE-24 as a measure of this construct. Rodriguez et al., (2002) 
argue that scales such as the SAFE have confounded the measurement of acculturative stress by 
incorporating stressors such as minority status; which, they believe, represents another form of 
life stress particular to migrants that is not necessarily part of the acculturation process. These  
 79 
considerations suggest that there is an urgent need for measures of acculturative stress that 
identify stressors and patterns of acculturation in other types of acculturating groups. The 
Multidimensional Acculturative stress Inventory (MASI: Rodriguez et al., 2002) focuses directly 
on acculturative pressures rather than general life stressors or stressors specific to migrants 
suggesting, that the MASI may be more appropriately used in other types of acculturating 
groups.  Rodriguez et al., (2002) also argue that measures such as the SAFE-24 ask whether the 
subject has experienced these social stressors, but do not assess subjective distress regarding 
these circumstances. The MASI provides a scoring system that distinguishes between the 
experience of a stressor and whether this was indeed perceived as stressful by the respondent.     
 
2.4.4.2         Stressful Reactions to the Process of Acculturation 
 
The notion of stress adopted by Berry et al., (1987) may be traced to the original work of Hans 
Selye (1936; cited in Berry, 1976) who argued that the General Adaptation Syndrome is a 
common reaction to stressors in the environment and may be clearly exhibited in psychological 
and psychosomatic symptoms (Berry, 1976). Following this conceptual focus, acculturative stress 
has been operationalized and measured using scales that assess general stress such as the 
General Health Questionnaire (Goldberg, 1992), psychological and psychosomatic symptoms, 
anxiety, depression, general clinical psychopathology, psychopathic behavior or other personality 
deviations; and many different, dysfunctional or deviant behaviors have been seen as indicators 
of acculturative stress. To date there is thus no uniform conceptualization of exactly what these 
stressful reactions to acculturation include.  
 
Joiner and Walker (2002) argue that studies which measure acculturative stress in terms of 
general psychological and somatic stress, fail to distinguish between stress that may be due to the 
acculturative process and general life stressors such as job-related problems and marital conflict.  
The Multidimensional Acculturative Stress Inventory (MASI; Rodriguez et al., 2002), focuses on  
 
 80 
stressors that may be seen as particular to acculturating individuals, such as Pressure to 
Acculturate from the new culture and Pressures against Acculturation from the culture of origin.  
The respondent is also asked whether the particular stressor has been experienced and is then 
asked to rate this experience in terms of levels of subjective stress; thereby providing a measure 
of subjective stress that is more particular to acculturation than indices of general psychological 
and psychosomatic stress.  
 
2.4.4.3         Measures of Acculturation  
 
Landrine and Klonoff (1996) demonstrated that only acculturation accounted for a statistically 
significant amount of the variance in psychiatric symptoms among African-Americans and Dona 
and Berry (1995) found that mode of acculturation was the best predictor of psychological and 
somatic stress in Central American refugees. Following research findings such as these, different 
levels or strategies of acculturation are believed to be predictive of acculturative stress such that 
measures of acculturation are seen as relevant correlates of psychological adjustment (Sands & 
Berry, 1993).  
 
Rodriguez et al., (2002) note that using acculturation as a measure, infers the presence of 
acculturative stress from a ?theoretical pathway? that is presumed to connect level of 
acculturation and psychological adjustment variables (Rodriguez et al., 2002). Measures which 
use traditional models of acculturation also assume that ?level of acculturative stress decreases as 
level of acculturation increases? (Rodriguez et al., 2002) which assumes that movement towards 
the dominant culture represents adaptation and mental health; while movement towards culture 
of origin is related to pathology. Organista et al., (2003) suggest that reliance on a ?simple? model 
of acculturative stress that equates increasing acculturation with increasing mental health is an 
example of the conceptual difficulties which have slowed progress towards an understanding of 
acculturation and mental health.    
 
 81 
Rodriguez et al., (2002) suggest that, as measures of acculturation have evolved, from uni-
 dimensional and unidirectional indices to complex multidimensional and bidirectional measures,  
measures of acculturative stress need to follow this progression by incorporating bidirectional 
axes that represent both ?original? and ?new? cultures.  The Multidimensional Acculturative Stress 
Inventory (MASI: Rodriguez et al., 2002) identifies two different dimensions which represent 
pressures emanating from both home (original) and host (new) cultures as simultaneous and 
continuous acculturative pressures towards and against acculturation; suggesting that this 
instrument may provide a more comprehensive construct of acculturative stress than earlier 
acculturation models.  
 
The MASI (Rodriguez et al., 2002) also provides a bi-directional structure of acculturative 
pressures which may accommodate different levels of acculturation and different types of 
acculturating groups; where the ?new? reality may conflict with the ?old?. This structure could thus 
be infused with items particular to the context whether it is opposing ?home? and ?host? cultures, 
traditional versus modern expectations; rural versus urban expectations, or western versus non-
 western cultures.  
 
 
 
 
 
 
 
 
 
 
 
 
 82 
2.4.5           Acculturative Stress and Eating Disorders  
 
Stress encountered during the process of socio-cultural transition has been suggested as a risk 
factor in the development of eating disorders in non-western population groups, particularly 
those encountering rapid socio-cultural transition (Nasser & Di Nicola, 2001) that occurs in 
trans-cultural migration (Perez et al., 2002; Davis and Katzman, 1997) and in rapidly 
westernizing or modernizing populations (Rathner et al., 1995; Ruggiero et al., 2000).   
 
Unfortunately, most studies have focused on developing an association between level of 
assimilation to western ideals of thinness and the development of eating disorders (acculturation 
hypothesis) and there are few studies which have attempted to systematically clarify any 
association between acculturative stress and eating disorder. Those studies which do indicate the 
presence of acculturative stress, have generally been retrospective and speculative, particularly to 
explain unexpected findings where eating disorders were not associated with the assimilation of 
western ideals of thinness.  
 
Those studies that have indicated an association between acculturative stress and eating disorder 
have also used a variety of indices of acculturative stress, including level and mode of 
acculturation (Lake et al., 2000), general psychological distress (Furukawa, 1994); subscales 
measuring more general pathology on a screening test for eating disorder (Davis & Katzman, 
1999) and the presence of intergenerational conflict (Mumford et al., 1991); and to date, the 
author is aware of only one study which has systematically set out to explore the relationship 
between eating disorders and acculturative stress, using a specific measure of acculturative stress 
(Perez et al., 2002).  
 
 
 
 
 83 
Perez et al., (2002) explored the relationship between acculturative stress, as measured by the 
Societal, Attitudinal, Familial, and Environmental Acculturative Stress Scale (SAFE; Padilla et 
al., 1985), body image (using the Body Image Scale; Stunkard & Messick, 1983), and body 
dissatisfaction and bulimic symptoms (using scales on the Eating Disorders Inventory: EDI, 
Garner et al., 1983) in black, white and Hispanic-American female university students.  
 
Results of this study demonstrated a significant interaction between acculturative stress and 
body dissatisfaction in predicting bulimic symptoms in black minority subjects. A similar 
interaction was demonstrated between body image perception and acculturative stress in 
predicting bulimic symptoms. Those who were high in acculturative stress demonstrated 
significant correlations between body dissatisfaction and bulimic symptoms, where increasing 
body dissatisfaction was associated with increasing bulimic symptoms; while those who were low 
in acculturative stress, did not demonstrate a significant correlation between these variables. 
Perez et al., (2002) concluded that body dissatisfaction and bulimic symptoms may be 
moderated by acculturative stress. These authors argue that the finding that body dissatisfaction 
and bulimic symptoms are uncorrelated in a group of women is in itself, important as it suggests 
that the lack of acculturative stress may buffer them against bulimic symptoms even in the 
presence of body dissatisfaction. Minority women also reported higher levels of acculturative 
stress than white women, and the authors concluded that acculturative stress may mediate the 
development of eating disorders in minority women in westernizing populations.  
 
Unfortunately, the SAFE focuses on the acculturative pressures faced by migrants as they 
attempt to integrate into a new, dominant culture and the research was conducted in a minority, 
immigrant sample. Findings may thus not reflect the experiences of other types of acculturating 
groups, such as majority, indigenous populations undergoing periods of modernization or 
encountering contact with western cultural influences.  
 
 
 84 
2.4.5.1         The Two-world Hypothesis 
 
Feminist writers have increasingly attempted to understand the increased prevalence of eating 
disorder that appears to occur at times of personal and social transition, including adolescence, 
industrialization, gender role transition, culture clash and generational disparity. The two-world 
hypothesis proposes that eating disorder may represent an attempt by women to cope with 
?juggling? two sets of conflicting socio-cultural realities and role expectations that occurs in any 
period of socio-cultural transition (Katzman & Lee, 1997).  
 
Katzman and Lee (1997) argue that when women change social class, context, class, countries or 
gender boundaries, they encounter feelings of dislocation and disconnection from familiar 
reference groups, and may use symptoms of eating disorder to cope with these feelings.  These 
authors continue to argue that when women are confronted with pressures to conform to 
conflicting ideals, they may also feel oppression and powerlessness and may employ self-
 starvation as an instrument to achieve autonomy, self-determination and control. Eating 
disorder symptoms are thus seen to provide a form of protest through conformity; and popular 
ideals of thinness are seen to sanction self-starvation and become the social ?coloring? of distress, 
rather than the cause (Russell & Treasure, 1989). 
 
Unfortunately, to date, there are no research studies which have attempted to systematically 
measure an association between socio-cultural transition and eating disorders using an 
established construct and measure of socio-cultural change; and researchers have ?only loosely 
described stressful factors that may mediate the relationship between socio-cultural transition 
and eating disorder? (Humphrey & Ricciardelli, 2003). These have included factors such as 
culture-clash (Katzman & Lee, 1997) inter-generational conflict (Mumford et al., 1991) and 
identity confusion (Nasser & Di Nicola, 2001).  
 
 
 85 
2.4.5.1.2       Culture Clash 
 
Katzman et al., (2004) found that, women who presented with anorexia on the island of Curacoa, 
were those who were ?straddling? several worlds simultaneously. They were all of mixed race, 
aspiring to fit into the white elite subgroup, whilst distancing themselves from the black 
majority; yet unable to find a sense of belonging or acceptance in the white elite group. These 
anorexic women also reported more anxiety than controls, suggesting that cultural and role 
transition may be accompanied by acculturation stress in vulnerable individuals.  
 
Davis and Katzman (1997) found that Chinese-American immigrants, who were highly 
assimilated to American culture, scored significantly higher levels of ineffectiveness maturity 
fears and perfectionism on the EDI, than those who were less acculturated. The authors 
concluded that immigrants may overcorrect perceived deficits which, in women may be their 
body shape, in an attempt to assimilate to the dominant culture and overcome feelings of 
ineffectiveness that may accompany the experience of ?living between cultures?.   Katzman and 
Lee (1997) concluded that women may thus employ self-starvation as a means of coping with the 
conflicting demands of ?straddling two worlds? that occurs in any period of personal and social 
transition and propose that the symptom of weight phobia should be changed to ?no control 
phobia? as a clue to better understanding the syndrome.  
 
Soh, Surgenor, Touyz et al., (2007) compared indices of control between European Australians 
and Chinese Singaporean females, with and without eating disorders; and found that control 
profiles differed across cultures for those without eating disorder, while the profiles of those with 
disorder were similar across cultures. The authors concluded that eating disorders are associated 
with distorted control indices irrespective of culture, but that the form, directionality, and degree 
of control, is culture dependent. This supports feminist hypotheses which speculate that eating 
disorders are universally an issue of control for women and that the particular symptom of 
distress or powerlessness may be ?colored? by culturally particular symbols or metaphors.  
 86 
2.4.5.1.3       Intergenerational Conflict 
 
Several studies have found that British-Asian adolescents from more traditional families have a 
greater risk of eating disorders than those in more western-orientated families; and concluded 
that children growing up within two different cultures may experience cultural confusion which 
may increase the risk of eating disorder (Bryant-Waugh and Lask, 1991; Mumford & Whitehouse, 
1988; Mumford et al., 1991; Hill & Bhatti, 1995). These girls were found to face real conflicts in 
their attempt to adjust to a culture that is different from their parents and a ?no-win? situation 
with their parents over cultural orientation. ? If I do as my parents ask, I will lose out on my own 
life and if I do as I please, I will probably lose my family? (Mumford et al., 1991; pg 227). These 
researchers concluded that the most ?traditional girls may be facing the greatest internal conflict 
around issues of identity as they grow up with two sets of cultural values and that the greater the 
difference between these two cultures, the greater the internal conflicts and anxieties which 
arise?. Hill and Bhatti (1995) found that traditional Asian girls in Britain also experienced lack of 
acceptance by British society and conflict with their parents when they attempted to identify with 
the ?new? culture.  
 
El-Islam et al., (1986) found that the greater the cultural differences between parents and 
children, the higher the general psychological distress on the GHQ-28, endorsed by both parents 
and children. The authors concluded that intergenerational conflict of attitudes and values may 
be associated with psychiatric symptoms. McCourt and Waller (1995) also found that risk for 
developing eating disorder could be partially explained by high levels of perceived maternal 
control and intra-familial conflict in British-Asian adolescents. The authors concluded that 
parental traditional values may conflict with the child?s attempt to identify with western culture; 
which may be less centered on obedience to parents. Katzman and Leung (1996; cited from 
Katzman, et al., 2004) reported that Chinese girls in Hong Kong whose values challenged the 
more traditional roles of their mother were more likely to report increased eating disturbance. 
Furnham and Hussain (1999) also demonstrated an association between eating pathology and 
 87 
 levels of conflict between parents and Asian females over going out and choice of friends. Other 
studies have found an association between risk of eating disorder and parental overprotection 
(Humphrey & Ricciardelli, 2003) over-controlling and over-protective mothers (Katzman et al., 
2004) and perceived maternal control (Ahmed, Waller & Verduyn, 1994; McCourt and Waller 
(1995) concluded that conflict between desire to identify with western culture and traditional 
values enforced by overprotective or controlling parents may result in perceived loss of autonomy 
and control such that disordered eating may be employed by girls as a means of gaining a greater 
sense of control over their lives.  
 
2.4.5.1.4       Identity Confusion 
 
Nasser and Di Nicola, (2001) argue that a significant proportion of eating disorders reported 
among non-western immigrant populations seem to highlight the presence of social change and 
confusion over identity. Nasser (1997) argues that eating disorders may be seen as symptomatic 
of underlying human distress, particularly the impact of cultural transition on the definition of 
self and identity. Rapid social changes threaten the identity and obliterate the boundaries, 
leaving the individual open, vulnerable and lacking a distinctive sense of self; which forces the 
individual to search for concrete and explicit references and may lead to a reformulation of the 
self through the framework of the body (Skarderud & Nasser, 2007). 
 
Postmodern American ideology also supports the notion that identity is something that is created 
and achieved rather than inherent or ascribed. The means of projecting personal identity have 
also gradually shifted from mind and character (internal attributes) to an increasingly visual and 
consumeristic (external) focus, and young  women are socialized to believe in the illusion that the 
?self? can be shaped or reshaped by drawing on cultural symbols to construct, understand and 
represent who they are (Becker, 2004).   Markets and their media are thus promoting the 
recreation of a new identity through the manipulation of the body and ?plastic bodies are creating 
plastic selves? (Rogers, 1999). Nasser and Di Nicola (2001) also speculate about a process that  
 88 
they name,  respectively, as  ?body regulation? and ?body modification? where the body is 
transformed into an identity as a method of coping with these cultural changes. Di Nicola (1990a 
& 1990b) suggests that bodily control may be seen as an expression of autonomy and definition 
within an unstable cultural milieu. The individual may thus use the concrete medium of the body 
to construct a self that is coherent and stable yet also distinct from others. Katzman et al., (2004) 
also found that anorexic women were those who struggled to develop a coherent sense of self, 
and that they ?wore their dilemma with their weight? (Katzman et al., 2004, pg 480). The 
increasing use of body building, tattooing and body-piercing (Skarderud & Nasser, 2007) and the 
revival of veiling in Muslim women (Nasser & Di Nicola, 2001) have also been seen as examples 
of body symbolism and the social construction of personal identity.  
 
Becker (2004) demonstrated that during a period of rapid social change, young women in Fiji 
lacked traditional role models for how to successfully maneuver themselves and drew heavily on 
role models from imported media to define themselves. Self identity was thus constructed by the 
manipulation of cultural symbols, including the reconfiguration of their bodies. Becker argued 
that Fijian youth needed to craft an identity which adopted western values about productivity 
and efficiency in the workplace while simultaneously ?selling their Fijian-ness?. Self-presentation 
thus needed to be carefully constructed to integrate these dual identities into a ?hybrid? identity.  
 
In contrast, Anderson-Fye (2004) found that Belizean girls were almost completely resistant to 
eating disorders. The author found that these girls were more concerned with body shape than 
body weight and that their culturally ideal, ?Coca-Cola? body shape, offered a protective effect 
from eating disorder. Belizean culture also placed greater emphasis on personality (internal) 
rather than bodily characteristics in the construction of self-definition and  promoted a 
psychology of self-care and protection (?never leave yourself?) which also seemed to buffer these 
girls from body anxiety; suggesting that cultures which focus on internal correlates of identity 
may demonstrate less risk for eating disorder.  
 
 89 
Studies of eating disorders around the globe have also converged in defining adolescence and 
young women as the segment of the population most at risk for the development of an eating 
disorder (Pike & Borovoy, 2004). Hilde Bruch (1978) first suggested that conflicting demands on 
contemporary young women creates severe identity confusion and that these kind of conflicts 
may be implicated in the increased incidence of eating disorders at adolescence. Adolescence has 
also been seen as the developmental period during which the construction of a coherent sense of 
ethnic (Phinney, 1990) and personal identity (Gowers and Shore, 2001) occurs. Tajfel and Turner 
(1986) and Ward et al., (2003) see ethnic or cultural identity as a core concept that links group 
membership to self-definition and self-esteem; suggesting that the construction of personal 
identity and self-esteem may be particularly difficult for adolescents exposed to heterogeneous or 
conflicting cultural influences (Becker, 2004). Endler and Parker (1990) proposed that 
acculturative stress may also be moderated if flexible cognitive strategies are applied, suggesting 
that those individuals who are unable able to successfully integrate conflicting cultural demands 
into a strong ?hybrid? or integrated cultural identity may experience greater risk of acculturative 
stress and eating disorder.  
 
These considerations suggest that the above conceptualizations of intercultural contact and 
socio-cultural  transition may be appropriately defined and measured within the constructs and 
framework of contemporary multidimensional and bidirectional models of acculturative stress 
(Berry, 1997; 2003) which define and measure acculturative stress in terms of tensions between 
opposing cultural pressures towards acculturation from the ?new? culture and against this 
acculturation from culture of origin (Rodriguez et al., 2002).   
 
 
 
 
 
 
 90 
2.5             Acculturative Stress in South Africa  
 
Following Berry?s (1997; 2003) multidimensional and bidirectional model of acculturation and 
stress, post-apartheid South Africa may be seen to have encountered many acculturative 
challenges, in every dimension of acculturation; and particularly for the previously oppressed 
and disadvantaged black South Africans.  
 
Following the abolition of apartheid, many black South Africans have migrated from small 
tradition-directed, pastoral rural societies into modern mass urban societies; such that, by 1996, 
55.4% of the population was urbanized (Central Statistical Service, 1997; cited from Szabo & Le 
Grange, 2001). Urbanization has been associated with increased socioeconomic opportunities 
and upward social mobility for many previously disadvantaged South Africans. For others, the 
expectation of economic opportunity has only served to widen the gap between expectations of 
affluence and an all-too-often bleak socio-economic reality. Jilek (2001) noted the consequences 
of this transformation on the mental health of individuals, including cultural identity confusion, 
anomie and deprivation; suggesting that previously rural black South Africans may be 
particularly at risk of conflict between modern ideas and traditional, rural values which, 
following Berry (1997, 2003) may promote tension and conflict between two disparate socio-
 cultural realties and acculturative stress. Ivan Eisler (2007) argued that freedom from oppression 
also brings an increase in responsibilities and anxieties and reported that the break up of 
communism in Eastern Europe, was accompanied by an increase in unemployment, poverty, 
crime, corruption, substance abuse, mental disorders, suicide and eating disorders. In post-
 apartheid South African society, many black South Africans also have to cope with their new 
social realities and responsibilities within the context of widespread destruction of black family 
relations. South Africans may thus be vulnerable to symptoms of acculturative stress as also 
reflected in the dramatic increase of suicidal behavior (Wasseneaar, Van der Veen & Pillay, 
1998), gangsterism, substance abuse, antisocial behavior (Thiel, 1997) and an emerging ethnic 
separatism (Stevens & Lockhat, 1997).   
 91 
Berry (1997) noted that changes to the ecological dimension may have profound effects on many 
other dimensions of human acculturation. Urbanization may, for example, may also promote 
changes in the structure and functioning of families and the roles and expectations of women; 
and the breakdown of traditionally protective family structures (Szabo and Le Grange, 2001). 
Black South African women have experienced a marked emancipation in recent years; not only 
have black women entered the male-dominated political and economic arena, but they are 
embracing western feminist ideologies and expectations, which are often in conflict with those of 
a more traditional father, spouse or work place; resulting in feelings of oppression and 
frustration with continued inequalities and psychological conflict. Black South African women 
may thus be particularly at risk for acculturative stress within the changing socio-political 
landscape of post-apartheid South Africa. 
 
Since the abolition of apartheid, black and white South Africans are also now engaging in inter-
 cultural contact in every aspect or level of human activity and sharing their diverse cultural 
backgrounds. The dimension of western versus non-western has been loosely equated with the 
dimension of individualism-collectivism (Bochner, 1994) which provide highly disparate and 
often mutually exclusive expectations regarding central issues such as loyalty and commitment to 
the community versus competition and achievement, which may promote a clash of cultural 
expectations during intercultural contact between white and black South Africans. 
  
Naidoo and Mahabeer (2006) found that black female South African university students in 
KwaZulu-Natal, valued contemporary western feminist ideas about personal growth, marriage, 
achievement and success for women and desired western, individualistic educational and career 
opportunities; while they also wanted to retain core, ancestral, collectivist family and social 
values (Triandis, 1990). Naidoo and Mahabeer (2006) concluded that this ?dual (integrationist) 
pattern heralds a shift that promises to bridge the ethno-gender gap?. 
 
 
 92 
Following Rodriguez et al., (2002), however, level of adaptation or mental health cannot be 
automatically assumed from mode of acculturation and attitudes of integration do not 
automatically assume a compatible and successful resolution between conflicting socio-cultural 
expectations (Sam & Oppedal, 2002). Following feminist hypotheses relating to eating disorders 
(Katzman & Lee, 1997), individualistic western values about women, also, typically, conflict with 
collectivist values about family, marriage, career and children which, prioritize the needs of the 
family or group above those of the individual; suggesting that many of these young female South 
Africans may be attempting to ?juggle? conflicting desires and values which may place them at 
particular risk for the development of eating disorders. Indeed, Naidoo and Mahabeer (2006) 
also found that black South African males favored separation and collectivism, which supports 
suggestions that many young black South African women may be faced with conflict between 
individualistic feminist aspirations and collectivist family and marital expectations. 
 
The greater degree of daily contact between cultures has, therefore, called the need for successful 
acculturative strategies at the level of the group; which require the balance of these conflicting 
cultural scripts into an integrated cultural definition (Phinney 1990; Ward et al., 2003). The new 
South Africa has adopted an official Multicultural policy which should assist the development of 
an adaptive integrated or super-ordinate South African ?culture? and identity by providing 
opportunities for contact and a milieu which accepts and respects all cultural groups. Szabo and 
Le Grange (2001) noted, however, that South Africans are ?faced with the burden of creating a 
new national identity? (Szabo  &  Le Grange, 2001; pg 28), suggesting that this fledging nation 
has yet to find an adequate, integrated, South African identity, with which ?new? South Africans 
could identify. There have been attempts to define an integrated South Africa as the ?Rainbow 
Nation? but this label does not appear to provide an adequate national identity. Many South 
Africans may thus be caught in a struggle between the ?desire to become more western versus a 
new-found pride in being African? (Szabo & Le Grange, 2001); and unable to effectively integrate 
these different cultural scripts into a coherent cultural identity.   
 
 93 
 
Rodriguez et al., (2002) identified acculturative stress as the ?push and pull? between pressure to 
assimilate to the dominant culture and pressures by the culture of origin against this process. 
Considering that the central conflict involved in eating disorder may be an identity struggle, then 
the struggle in South Africa could be defined in terms of the ?push and pull? of conflicting 
?western? and ?black African? cultures (Szabo & Le Grange, 2001). It appears evident that, in the 
absence of an adequate national identity, adolescents have created negative labels such as 
?coconut? or ?oreo? (which means black on the outside but white on the inside) for those blacks 
who have become westernized; suggesting that black adolescents may be experiencing  tension 
between  opposing cultural pressures and difficulty constructing a coherent, integrated identity. 
 
Following Berry (1997), the apartheid policies of segregation and racial discrimination, may also 
affect the degree of acculturative stress encountered during transition, as continued 
discrimination and prejudice may result in greater stress (Pernice & Brook, 1966). Apartheid 
policies of segregation not only kept cultures apart but institutionalized minimal intercultural 
contact, suggesting that lack of prior intercultural contact may foster suspicion, discrimination, 
prejudice and acculturative stress in the ?new? South Africa. Pettigrew and Tropp (2006) argue 
that greater intercultural contact, under optimal conditions, is associated with less inter-group 
prejudice, but that these effects are significantly weaker in minority status groups than in 
majority groups; suggesting that black females who are in a minority position in ?white? schools 
may well experience increased levels of acculturative stress, as South Africa struggles to break 
free from prior discriminatory policies. Although acculturative stressors are likely to be 
experienced by all South Africans, it is thus likely that black South Africans, particularly in 
minority situations, are exposed to greater cultural disparity and acculturative stress than white 
South Africans. The apparent increased risk of eating disorder amongst black schoolgirls may 
thus reflect the effects of acculturative stress which may accompany the transitional process in 
this country.  
 
 94 
Basic tenets of individualism and collectivism also offer profoundly different expectations 
regarding the socialization of children. Collectivism promotes authority, dependency, obedience 
and compliance, while individualism promotes autonomy, self-reliance and assertiveness. 
Tension between these opposing forces may promote conflict between cultures that may be 
expressed in different expectations between adolescents and authority-figures such as teachers 
and in intergenerational conflict which have also been associated with acculturative stress and 
eating disorders (McCourt & Waller, 1995).   
 
Following Erikson, (1963), Stevens and Lockhat (1997) noted that adolescence is a time of 
identity development versus role confusion and that factors that were instrumental in the 
shaping of black adolescent identity during the era of apartheid are fundamentally different in 
contemporary South Africa. Black adolescents now have to attempt to negotiate the apparently 
contradictory expectations of pre and post-apartheid South Africa which may contribute towards 
role confusion rather than identity integration; suggesting that they may be particularly at risk 
for the development of eating disorders. These authors continue to note that with an increasing 
presence of western ideology and economic structures, a ?Coca-Cola? culture has emerged among 
black South African adolescents, which represents a radical ideological shift from African 
collectivism to a more western individualism and competitiveness. This identity allows them to 
cope in their new social milieu but at the same time marginalizes and alienates them from 
traditional realities and community support. Following Bulhan (1980), previously oppressed 
adolescents may thus also experience ?psychological tension? as they straddle the two different 
worlds of pre and post apartheid, that are alien to one another. Following Berry (2003) and Ward 
et al., (2003), these observations suggest that black South Africans adolescents may be 
particularly at risk for the experience of acculturative stress at every level of acculturation. The 
author, did not, however, find any evidence of studies that attempt to formally measure 
acculturative stress in relation to eating disorders in the South African context and the following 
studies only speculated on the possibility of acculturative stress as a factor to explain discrepant 
findings. 
 95 
 
Marais, Wassenaar & Kramers, (2003) demonstrated significant positive correlations between 
the acculturation strategy of assimilation and all the subscale scores on the Eating Disorders 
Inventory (EDI; Garner et al, 1983) in black South African college women. These authors 
concluded that increasing contact with western expectations is associated with greater eating 
pathology ?particularly in those individuals who are highly assimilated into and seek affirmation 
from Western culture?. The same study, however, also demonstrated 100% positive correlations 
between EDI subscales and the acculturation strategy of rejection (separation) in black females; 
which challenges this link between culture and eating disorders. The authors speculated that this 
kind of association may be mediated by acculturative stress, particularly as all of the general 
psychological subscales of the EDI (ineffectiveness, maturity fears, interpersonal distrust, 
perfectionism and interoceptive awareness) were more evident in these studies than subscales 
measuring the cardinal ?western? indices of an eating disorder; and were higher than those of 
Canadian female adolescents (Leichner, 1986), suggesting the presence of conflict and distress.  
 
Berry (1997) noted that separation is more likely to be accompanied by acculturative stress than 
strategies of integration and assimilation, but that assimilation and separation may be 
differentially stressful, depending on the context. Following Berry, the findings of the above 
research study could suggest that assimilation and separation were equally associated with 
distress in this black South African context. Following Davis and Katzman (1997) and Katzman 
and Lee (1997), these findings could also suggest that black South African females may be 
suffering a sense of ineffectiveness and distress as they attempt to ?straddle? two disparate 
cultural worlds.  Following Rodriguez et al., (2002), these black South African women may well 
feel caught between pressures to assimilate to white western cultural ideals and pressures against 
this assimilation (for separation). Marais et al., (2003) and Kramers (2000) also demonstrated 
100% negative correlations between EDI subscales and the strategy of integration in black 
women in KwaZulu-Natal, South Africa; suggesting that eating disorders may be reduced when 
these conflicting pressures were integrated into a compatible formulation.  
 96 
The same study also demonstrated an increased risk of eating disturbances (bulimia) in black 
South African men, although these men scored normally on EDI subscale scores for body 
dissatisfaction, and drive for thinness. Kramers (2000) also found that assimilation did not 
correlate with drive for thinness or body dissatisfaction in black South African women; 
suggesting that factors other than the acculturation of western ideals of thinness are needed to 
explain the increased risk of bulimia found in these groups. The authors also speculated that the 
increased prevalence of bulimia, in the absence of drive for thinness, may reflect a cultural 
practice of purging and the social sanction of large appetites in men.  
 
These findings also call into question the validity of the EDI as an instrument measuring eating 
disorders in South Africa (Edwards & Moldan, 2004). They also caution that rating scales do not 
equal disorder and that community samples are only a proxy for clinical data in terms of 
potential risk.  At most, data using the EDI on community samples in the South African context, 
suggests that some black South African samples may suffer more general psychological conflicts 
possibly associated with acculturation; which may render them more vulnerable to a wide range 
of psychological problems, including eating disorders, than their white South African 
counterparts. This stress may well reflect the stress of acculturation known as acculturative 
stress. The current research study therefore aimed to systematically explore any associations 
between eating disorders and acculturative stress as the ?push and pull? of opposing pressures for 
and against acculturation, as articulated by Rodriguez et al (2002). 
 
During data collection for the current research study, the author came across a school which had 
written a paper entitled ?The experiences of black adolescent females in multi-cultural schools 
and their impact on identity development? (Danville Park Girls; High, 2005) which reinforces the 
possibility that black South African schoolgirls may be experiencing opposing acculturative 
stressors, intergenerational conflict and identity confusion. The material for this paper was 
collected by the school counselor from two Focus Groups of black learners in an urban 
multicultural school and was delivered to a conference of school head teachers.  
 97 
The paper noted that, at primary school level, girls of different race groups are well integrated in 
multicultural schools. Once they graduate to high school, however, there appears to be a trend 
amongst black learners to start associating more strongly with their own cultural group. This 
trend is believed to be due to an increasing awareness of cultural difference within the 
traditionally ?white? school milieu and an increased sense of identification with black peers. The 
paper highlighted that, in addition to the normal difficulties of adolescent individuation, these 
girls experience culturally-related difficulties which make it more difficult for them to forge a 
coherent identity. Specifically, these girls complained of conflicting values and expectations from 
their ?white? school environment on one hand and their ?black? friends, family and community, on 
the other.  At school they are addressed by their English name and are expected to speak English. 
If any Zulu language ?slips out? in this environment, they are accused of being insolent or 
aggressive. At home they are addressed by their Zulu name and expected to speak Zulu. If 
English words ?slip out?, they are accused of abandoning their culture or being a ?sell out? who 
doesn?t even know her own language; or being cheeky or thinking they are ?better? than their 
black community. Some girls had even experienced violent reactions from the black community 
for speaking English.   
 
At school they are also encouraged to think independently, to express their opinions assertively 
and to value their own opinions. While in the black community they are expected to value the 
opinions of their elders and avoid conflict.  In the black community, they are accused of being 
disrespectful and cheeky, if they voice their opinion or make eye contact with their elders, while 
at school they are accused of being disrespectful if they don?t make eye contact with an authority 
figure. Many girls found these mixed expectations confusing and some were even further 
confused by conflicting parental messages. On one hand, their parents want their daughters to go 
to multi-cultural schools because they want them to experience all the things that they never had 
the opportunity of doing during the years of apartheid.  ?Yet, on the other hand they want their 
daughters to hold onto the traditions and values of their culture at home, which may be in direct 
contrast to the traditions and values of their schools. ?  
 98 
These different sets of expectations also make it difficult for these girls to feel a sense of 
belonging and acceptance by any group. They felt a sense of ?being different from? their white 
peers at school, who do not always know about, understand or even respect their difficulties.  Yet, 
on the other hand, they do not always feel a sense of belonging to the black community as there 
are aspects of their own cultural practices and beliefs that they do not fully understand or believe 
in. They are also frequently discriminated against and accused of being a ?coconut? by the black 
community. ?Coconut? is a negative term which refers to the fact that the black girl is seen to be 
?white? on the inside but black on the outside, indicating that she has abandoned her black 
culture and trying to become ?white? (those who try to assimilate into white culture).  There is 
extreme pressure to avoid being called a ?coconut? and the girls have to go to great lengths to 
choose the ?right? clothing and styles which may be seen as ?white enough? to be acceptable by 
white culture but not ?too white? to be accepted by their black peers.  On civvies day, they have to 
make sure that they wear clothes with a Levi or Nike label which tend to be favored by black 
learners. They would also be labeled as a coconut if they socialize with white peers outside school 
or reside in white suburbs as opposed to black townships. These conflicting expectations made 
many of these girls feel like they are continually playing a role where they have to switch roles 
from one setting to another, and where incomplete switching, results negative consequences. 
This also results in the lack of a sense of ?who they really are? and lack of a sense of ?belonging? in 
any group. ?It can become very difficult to answer the question of who am I?? ?It can feel as if we 
are living three or four different lives and at times not living any of them particularly 
successfully?.  ?We do not fit in completely at home or school?.  
 
These observations and related research data suggest that stressors associated with the process of 
westernization may be a significant factor associated with risk for eating disorder, particularly for 
adolescents, in the rapidly changing socio-political landscape of black South African women. It 
may be speculated that post-apartheid black South African women may encounter stress 
associated with westernization and a clash between cultures as non-western collectivist cultural 
traditions come into daily contact with western, individualistic cultural influences. 
 99 
2.6             Study Objectives 
 
The current research study thus aimed to systematically measure individual levels of 
acculturative stress, using the bi-cultural index designed by Rodriguez et al., (2002) in the high 
school context in South Africa; and to identify any relationships between this acculturative stress 
and eating disorder in this group. In order to achieve this research outcome, this study needed to 
fulfill the following objectives:  
 
? To modify and validate the Multidimensional Acculturative Stress Inventory (MASI: 
Rodriguez et al., 2002) as the Multidimensional Acculturative Stress Inventory-Revised 
(MASI-R) for use in the South African context of this research study. 
 
? To identify and measure the presence of acculturative stress in a sample of Black, female 
adolescents in KwaZulu-Natal, South Africa, as measured by the MASI-R.  
 
? To identify relationships between acculturative stress as measured by the MASI-R, 
dysfunctional eating attitudes and behaviors and eating disorders in this sample. 
 
It is hoped that this research may clarify the link between acculturative stress and eating disorder 
and contribute towards a better understanding of eating disorders. This may inform strategies 
aimed at identifying populations at risk and implementing treatment and prevention programs. 
 
Research that identifies acculturative stress in South Africa may also contribute towards a better 
understanding of multicultural integration and inform strategies aimed at identifying 
populations at risk and implementing preventive programs. Research that identifies acculturative 
stress in South Africa may also have implications for the prevention and treatment of other 
aspects of mental health such as depression, anxiety, suicidal behavior, substance abuse, 
deviance, and violence. Research that validates an instrument for measurement of acculturative 
stress within a South African sample may also be useful towards future research in this area.  
 100 
2.6.1           Research Hypotheses 
 
Ho: Acculturative stress as measured by the Multidimensional Acculturative Stress Inventory-
 Revised (MASI-R) will be unrelated to eating dysfunctional eating attitudes and behaviors (as 
measured by the Eating Attitudes Test: EAT26: Garner, Olmsted, Bohr & Garfinkel, 1982) or 
eating disorders (as measured by the Eating Disorders Diagnostic Scale; EDDS: Stice, Telch & 
Rizvi, 2000) in a sample of black schoolgirls in KwaZulu-Natal, South Africa. 
 
H1: Acculturative stress, as measured by the Multidimensional Acculturative Stress Inventory?
 Revised (MASI-R) will be significantly related to dysfunctional eating attitudes and behaviors (as 
measured by the EAT26: Garner et al., 1982) and eating disorders (as measured by the EDDS: 
Stice et al., 2000) in a sample of black schoolgirls in KwaZulu- Natal, South Africa. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 101