and while it may go without saying, i ask that you please respect my academic integrity and refrain from recycling this for any other purposes. thanks :)
Anorexia Nervosa in Non-Western Contexts: Implications for Diagnostic Definition
Anorexia nervosa (AN) is a psychological disorder characterized primarily by a body weight at least 15% less than expected, achieved either by means of self-imposed dietary restriction or purging behaviors. Other clinical features of this disorder, as defined by the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) and the International Classification of Diseases (ICD-10) include an absence of menses in women, body image disturbances, and a fear of being or becoming fat (Klein &Walsh, 2003). Complications of this disorder include disruptions in neuropsychological functioning, mood disturbances, diminished gastrointestinal functioning, and cardiac arrythmias, among other consequences. Five to ten percent of patients with AN die as a result of the disorder (Klein & Walsh 2003). This alarming statistic has fueled much research about the nature and etiology of AN in order to understand its features so it can be effectively treated and even perhaps prevented.
Most studies on AN have been conducted in Western societies (Hoek & van Hoeken, 2003). These studies assess prevalence rates, identify subjects with the disorder, examine factors related to their condition, and construct hypotheses of etiology from these data. This whole process hinges on the current definition of AN as defined by the DSM-IV and/or the ICD-10. This definition is hence problematic because it limits the diagnosis of AN to the Western conceptualization of the disorder. To construct a culturally universal description, therefore, cross-cultural studies are necessary to evaluate the current diagnostic definition of AN. However, this inquiry is difficult for a number of reasons. Large-scale epidemiological studies of eating disorders in non-Western countries are rare (Hoek & van Hoeken, 2003). Consequently, smaller studies that assess the prevalence of eating pathology must suffice. Not all of these studies explore etiological factors associated with the identified disorders. Furthermore, given that the prevalence rate for AN is low in Western cultures with a body of large-scale epidemiological research, it is difficult to identify cases of AN in studies in non-Western cultures with smaller samples. Because of the current limits of the empirical literature, evidence of specific and detailed features of AN in these cultures comes from case studies. These case studies can compliment the findings from the existing empirical research and provide a basis for further inquiry and the construction of a cross-culturally valid definition of AN. The present paper reviews recent empirical and case studies of eating disorder symptomatology and AN in female adult populations in Africa and Asia.
Prevalence and Etiology in Non-Western Cultures
There have been several approaches to studying eating disorders in a non-Western context. One approach is comparing measures of eating pathology in a non-Western culture to a Western culture. Another is examining correlations between the extent of exposure to Western culture and eating pathology in non-Western cultures. A third approach investigates the specific epidemiology and etiology of eating pathology in non-Western cultures. Examining the findings from this variety of approaches enables a comprehensive understanding of the prevalence of and factors associated with eating pathology in non-Western cultures.
Western vs. Non-Western
Gupta, Chaturvedi, Chandarana, and Johnson (2001) compared attitudes towards body image in two cultures. Sixty-five women in Canada (mean age 21.4) who were recruited from the community and 47 women at a university in India (mean age 18.7) completed two subscales from the Eating Disorder Inventory (EDI), the Drive for Thinness subscale and the Body Dissatisfaction subscale. They also provided height and weight information from which the researchers calculated body mass index, a ratio of weight to height (BMI; kg/m2). During analysis of the EDI scores, BMI was statistically controlled to avoid any possible confounds. The results indicated no differences between the two samples, indicating similar attitudes were held in both cultures. Scores on the Drive for Thinness subscale were comparable between the two before and after controlling for BMI. However, the Body Dissatisfaction scores were higher in the Canadian sample than in the Indian sample before BMI was controlled. Given this finding, the authors suggested that BMI is a confounding factor and should be controlled in studies examining weight concerns and body image. The study did not assess eating disorder symptomatology or its relationship with the EDI subscales.
Sjostedt, Schumaker, and Nathawat (1998) examined eating pathology and fat phobia in 297 Australian (mean age 23.2, range 17-50) and Indian (mean age 22.2, range 18-36) college students. Subjects completed the 26-item version of the Eating Attitudes Test (EAT-26) and the Goldfarb Fear of Fat Scale (GFFS). Female subjects from India scored significantly higher than female Australian subjects on the EAT-26. There were no differences between the samples on the GFFS. Within the Indian sample, subjects residing in a rural locale scored higher on the EAT-26 than those residing in an urban environment; no differences were found on the GFFS. These results contradicted the authors’ expectations (based on previous studies) that eating pathology would be more prevalent in the Western culture and in urban subjects (who theoretically had a high probability of exposure to Western culture than rural subjects). They were unable to offer any explanation for this trend. Because this finding is subject to replication, it is safer to focus instead on the finding that there were no significant differences between the two cultures on a fear of fatness scale.
Suhail and Nisa (2002) conducted a study to determine the relationship between media exposure to Western culture and eating pathology in Pakistan. They administered a battery of instruments to 111 female postgraduate students (mean age 21, range 18-24). The measures included the EAT-26 and the Body Shape Questionnaire (BSQ), a measure of body dissatisfaction. Exposure to Western culture was measured by asking subjects if and how often, on a 6-point scale, they watched Western and Indian television programs. All questionnaires were delivered in an interview format to ensure comprehension. The results indicated a positive correlation between Western media exposure and the scores on both the EAT-26 and BSQ.
Abdollahi and Mann (2001) also conducted a study to determine the influence of Western culture on eating pathology and body image. Their design was unique in that their subjects were from the same cultural background, but living in different cultures. They compared Iranian women living in America with Iranian women living in Iran, a country where women are required to fully cover their bodies in public and where Western media is prohibited. Fifty-nine female students in Iran and 45 female students of Iranian descent in Los Angeles completed various relevant questionnaires. The Figure Rating Scale (FRS) presented drawings depicting female figures with a range of weights. Subjects picked the figures that they believed represented their current and their ideal bodies. The Eating Disorder Examination-Questionnaire (EDE-Q) assessed eating disorder symptomatology. Subjects provided height and weight information, as well as desired weight. The Iranian women in America had lived there for an average of 14.8 years (range 4-22). The length of residence was considered as a measure of acculturation, along with a self-report of how often the subjects used the Iranian language. In the Iran sample, exposure to Western culture was measured by frequency of seeing various Western media and self-reported interest in Western culture. The results revealed no significant differences between the samples on the EDE-Q or the FRS. Additionally, the extent of acculturation in the Iranian subjects in America and the degree of exposure to Western media in the subjects in Iran were not correlated with measures of eating pathology and body dissatisfaction.
Non-Western Epidemiology and Etiology
Two studies were conducted in Turkey to determine prevalence rates of and factors associated with eating pathology. Uzun, Gulec, Ozsahim, Doruk, Ozdemir, and Caliskan (2006) surveyed 414 female college students in Turkey (mean age 19.9, range 16-24) to determine the prevalence of eating pathology. Subjects completed the 40-item version of the EAT (EAT-40) in Turkish. Those that scored above the clinical cut-off (17.1%) were then interviewed using the Structured Clinical Interview for the DSM-IV (SCID). AN was diagnosed in 0.5% of the total sample, a rate similar to that in Western cultures. The study also examined sociodemographic factors (education, residence, income) and their relationship to levels of eating disorder symptomatology. There were no differences on any of these factors between those who scored above the clinical cutoff and those who scored below it. Etiological factors were not considered in this study, but they were considered by Kugu, Akyuz, Dogan, Ersan, and Izgic (2006). They administered the EAT-40 to 980 college students (50% female) in rural Turkey. A study group with high scores above the clinical cutoff (21 subjects, 18 female; mean age 21.1, range 18-24) was compared with a control group of subjects who scored below the clinical cutoff (21 subjects, 18 female; mean age 20.0, range 18-24) on validated measures of self-esteem, child abuse and neglect, and family functioning. There were no significant differences between the study group and the control group on demographic factors. Subjects in the study group with higher scores of eating pathology were found to have lower self-esteem and family functioning, along with higher levels of sexual and emotional childhood abuse. The authors noted that these factors have also been found to be associated with eating disorders in Western cultures.
Bennett, Sharpe, Freeman, and Carson (2004) investigated the prevalence and features of AN in rural Ghana. They screened 668 female secondary school students (mean age 18; range 15-25) for height and weight information, which they then used to calculate BMI. One hundred subjects had BMIs below 19 and were then individually interviewed to assess eating attitudes and behaviors, as well as physical health. A control group of subjects with BMIs above 19 were also assessed. The researchers administered the EAT and the Bulimic Investigatory Test, Edinburgh (BITE) questionnaires in an interview format to ensure subjects understood the content and terminology. There were 29 subjects who met the weight criterion for AN (BMI below 17.5). Of these, 19 had a medical cause for their low weight (i.e., malaria). The remaining 10, however, admitted to intentional dietary restriction as the reason for being underweight. There were no instances of amenorrhea. None of these subjects indicated any fear of being or becoming fat. Instead, all subjects reported religious reasons for their behaviors. Importantly, the EAT and the BITE did not predict the presence of self-starvation in these subjects. Their scores on these instruments and the scores from subjects with BMIs under 19 and the control group were comparable. The authors indicated that many subjects from the assessment sample reported fasting occasionally for religious reasons (Christian and Muslim). On this point, the authors suggested that such religious fasting in this culture is similar to dieting in Western culture; that is, it is relatively common but does not necessarily constitute pathological eating behavior. Despite the absence of fear of fatness and amenorrhea, the subjects were diagnosed with AN on the basis that they exhibited morbid self-starvation.
S. Lee, A. M. Lee, Ngai, D. T. S. Lee and Wing (2001) examined women with AN in China and their self-reported reasons for refusing food. They administered a battery of instruments to a sample of 48 female patients with AN. The clinical measures included the General Health Questionnaire, the Beck Depression Inventory (BDI), and the Hamilton Depression Rating Scale (HAM-D). The researchers also utilized a rationale for food refusal questionnaire (RFQ), designed for the study. This measure asked subjects to choose one of eight possible reasons for refusing food at different times in the history of their disorder (12 months ago, 3 months ago, and at clinical presentation). Subjects were also interviewed to determine the clinical features of their disorder. If they indicated a fear of being or becoming fat that motivated pathological eating behavior and subsisted despite a low body weight, they were classified as “fat-phobic.” All other subjects were classified as “non-fat-phobic.” There were 32 and 16 subjects in each group, respectively. The two groups did not differ significantly on any of the psychological measures, save the RFQ. On this measure, the fat-phobic subjects rated “fat phobia” as their primary reason for dietary restriction. Other rationales cited by this group were “stomach bloating”, “stomach pain”, and “don’t know”. These reasons were reported to be more common later in the course of the disorders (i.e., more commonly endorsed at the time of clinical presentation than at 12 months previously). In the non-fat-phobic subjects, the primary reason they reported was “stomach bloating”. The other rationales identified were “no hunger”, “no appetite”, and “don’t know”. While no psychological instruments correlated with “fat phobia” as a reason for food refusal, differences in premorbid BMI between the two subject groups were significantly associated with this difference. Fat-phobic subjects had higher premorbid BMIs than did non-fat-phobic subjects. The authors note that AN could not be attributed to depression.
Implications for Diagnostic Definition
The research reviewed indicates that AN exists in non-Western cultures, but does not always share the same etiological factors as AN in Western cultures. Studies comparing Western and non-Western cultures (Gupta et al., 2001; Sjosted et al.,1998) found that factors such as body dissatisfaction, drive for thinness, and fear of fatness did not differ between cultures. Etiological factors that were associated with eating disorder symptomatology in non-Western cultures included religious ascetics (Bennett et al., 2004), physiological complaints (Lee et al., 2001), low self-esteem, childhood abuse and neglect, dysfunctional families (Kugu et al., 2006), body dissatisfaction, and exposure to Western culture (Suhail & Nisa, 2002). Concerning the influence of exposure to Western culture, the studies by Suhail and Nisa (2002) and Abdollahi and Mann (2001) achieved contradicting results; the former found that more exposure to Western media positively correlated with measures of eating pathology body dissatisfaction, whereas the latter did not find a correlation. Further, Abdollahi and Mann (2001) did not find a correlation between the eating pathology and the level of acculturation of subjects of non-Western origin residing in a Western country. This suggests that exposure to Western culture can increase the prevalence of eating pathology and influence the etiological factors associated with disordered eating in Western cultures, but does not necessarily do so.
One of the most important issues in cross-cultural studies is the integrity of the instruments utilized. Most of the authors in this review cited evidence for cross-cultural validation of instruments administered in their studies (Bennett et al., 2004; Suhail & Nisa, 2002; Sjostedt et al., 1998; Abdollahi & Mann, 2001). Another issue to consider is the accuracy of translation of these measures. Abdollahi and Mann (2001) offered the EDE-Q and FRS in both English and Iranian to Iranian subjects, but they found no differences between the responses from the different translations. Uzun et al. (2006) utilized a Turkish translation of the EAT-40 that had been previously validated. Thus, it appears that there were no problems associated with translated versions of the instruments utilized in cross-cultural contexts. Other studies that only administered instruments in English appropriately anticipated comprehension difficulties. In Pakistan, Suhail and Nisa (2002) delivered the English EAT-26 and BSQ in interviews to ensure comprehension, because their pilot study revealed that many subjects did not understand terms such as “binge,” “diet foods,” and “laxatives.” Similarly, Bennett et al. (2004) administered the English EAT-40 in an interview format in English. They also noted that many subjects did not know the definitions of words such as “binge” and “diet.” In sum, the methods of assessment utilized in the reviewed studies retained their integrity when administered in non-Western cultures.
But even given that the instruments proved valid in cross-cultural settings, they are all nonetheless Western measures. Several authors have noted this significant limitation (Bennett et al., 2004; Suhail & Nisa, 2002; Sjostedt et al., 1998). One of the most important findings is from the study by Bennett et al. (2004), which found that the EAT and BITE instruments could not identify the subjects who were ultimately diagnosed with AN by a clinician. Because a fear of fatness is required for the diagnosis of AN, the instruments used to detect eating disorder symptomatology include items related to fat phobia. This is problematic, however, given that a fear of fatness was not found in all cross-cultural studies (Sjosted et al., 1998; Bennett et al., 2004; Lee et al., 2001). The defining feature of AN (i.e., that distinguishes it from other eating disorders) is intentionally attaining or maintaining an unhealthily low body weight as a result of motivating psychological factors to do so. The current Western definition requires fat phobia as one of these psychological factors, but in non-Western cultures, these psychological causes may be different, as evidenced in the studies in this review. There are methodological limitations to these studies, however, because they have all utilized Western instruments to detect and examine the epidemiology and etiology of eating pathology. Given this situation, an examination of case studies can help identify the causes of AN in non-Western cultures by providing clues as to the direction further research should take.
Case Studies of AN in Non-Western Cultures
Abou-Saleh, Younis, and Karim (1998) presented three case studies of women in Arab cultures with AN. The first case was a 21-year-old unmarried female from Oman. She started dieting and excessively exercising in response to weight concerns (i.e., reported that she did not want to be fat like her mother) at the age of 18. Her premorbid BMI was 19.8; her lowest was 14.8. She had dysmenorrhea and amenorrhea for a period of 3 months. She exhibited many symptoms of depression, such as unhappy mood, lethargy, and sleep disturbances, and expressed a belief that she was fat and ugly. The second case was in a 23-year-old female from Oman. At the age of 18, she lost her appetite and experienced significant weight loss. This was involuntary and with no discernable organic cause. Her lowest BMI was 13.1. She was married at the age of 20 and became pregnant shortly thereafter. She reported family pressure to restrict her food intake during pregnancy and was distressed when she put on weight. She had dysmenorrhea and had amenorrhea for four months. She complained of palpitations, especially when she ate. There were no indications of any comorbid psychiatric diagnoses, although there were some in her family (her father had panic and her younger brother was anxious). She had no problems in her marriage. The third case was an 18-year-old female from the United Arab Emirates. She was a nomad found in the desert and brought to the hospital at a weight of 62 lbs (her height was not indicated, so her BMI could not be calculated). As an inpatient, she refused to eat, pulled out her feeding tube, and vomited after force-feeding. She only started eating after eight treatments of electroconvulsive therapy. She was discharged at 79 lbs and maintained this weight at a 6-month follow-up assessment. The authors did not indicate if the patient cited any reason for her behavior. They did note, however, that as a nomad, she arguably had limited exposure to Western culture. The authors suggested that influence of Western culture could account for at least some of the etiological factors in the other two cases.
Chandra, Shah, Shenoy, Kumar, Varghese, et al. (1995) described a case study in India of AN and family functioning with a 20-year-old female. Her eating pathology began at the age of 18, after her parents began to argue frequently. After arguments with the father, her mother would refuse to eat. Soon, she also came to refuse food after arguments with her father. As her condition worsened, she became increasingly disconnected from her parents. At the time of assessment, she felt lonely and depressed. Her BMI was 16, but she continued having normal menstrual cycles. She did not report any fear of becoming or being fat. The authors hypothesized that her AN was a coping mechanism for dealing with family problems.
Khandelwal, Shara, and Saxena (1995) also investigated AN in India with three cases studies. The first case was an 18-year-old female student who started restricting her diet after she was rejected from an academic program. Although her BMI was 13.33 and she had amenorrhea at admission to the hospital two years later, she insisted her eating behavior was normal. She did not have body image disturbance. The second case was an 18-year-old female student who was premorbidly overweight. She began to diet and exercise in an effort to lose weight, but shortly experienced spontaneous vomiting after eating. She consequently restricted her diet further and over 18 months lost 33% of her body weight. While her BMI was 16.65, she still had regular menstrual cycles. She believed that her weight was normal. The third case was a 22-year-old female. For five years, she claimed she was unable to eat much because of discomfort and spontaneous vomiting after meals. She was treated previously for a potential medical cause of these symptoms, but did not improve. At assessment, she had a BMI of 12.06 and had not menstruated for four years. She did not report any body image concerns and believed her eating habits and body weight were normal.
Lee (1995) described two cases of AN and the associated etiological features in China. The first case was a 29-year-old female who cited a loss of appetite as the reason for her weight loss. Her lowest BMI was 14 and she had amenorrhea for seven years. As a child, she had been sexually abused by her father. She had attempted suicide two times and reported feelings of hopelessness, but her score on the BDI did not indicate that she was depressed. Similarly, her score on the EAT was subclinical. The second case was a 31-year-old female. She began reducing her food intake after the end a romantic relationship. Like the first case, she presented herself as extremely hopeless, but her BDI score did not indicate symptoms of severe depression. Her EAT score was also subclinical. She cited a lack of hunger as the cause of her weight loss, although there was no apparent medical cause. Nonetheless, she vehemently opposed all forms of treatment, including tube feedings, psychotherapy, anti-depressants, traditional herbal remedies, and hypnotherapy. Her lowest BMI was 6.3. She died of cardiac arrest.
These case studies do not provide exhaustive representative data of the features and etiology of AN in non-Western cultures. They do, however, illustrate several features found in the existing empirical research, such as family dysfunction (Chandra et al., 1998; Lee, 1995; Kugu et al., 2006), sexual abuse (Lee, 1995; Kugu et al., 2006), physiological complaints (Lee et al., 2001; Abou-Saleh et al., 1998; Khandelwal et al., 1995; Lee, 1995), and premorbid BMI (Lee et al., 2001; Khandelwal et al. 1995). While body image concerns were found in some cases and studies (Lee et al., 2001; Suhail & Nisa, 2002; Khandelwal et al., 1995), they were not present in all of them. Indeed several cases particularly noted the absence of fat phobia (Chandra et al., 1998; Lee, 1995; Khandelwal et al., 1995). There were two instances in these cases in which subjects with AN scored below the clinical cutoff on the EAT (Lee, 1995). This supports the findings from empirical study by Bennett et al. (2004), which also noted this. These case examples support the existing research and serve as foundations upon which future empirical studies can build. Future research on these and other potential etiological factors can help to develop a culturally universal understanding of AN that will in turn improve its treatment.
While the current definition of AN may be culture-bound, the disorder is certainly not. The condition of psychological factors being responsible for pathological eating behaviors resulting in a significantly low body weight found both in Western and non-Western cultures. However, the current diagnostic definition of AN also requires a fear of becoming or being fat. The studies reviewed here strongly suggest that this definition is not appropriate in all cross-cultural contexts and should be revised accordingly. This revision will improve further research on and, ultimately, treatment for this life-threatening disorder. Epidemiological studies will be able to more accurately identify cases of AN in all cultures and, consequently, etiological studies will be better able to examine the features associated with it, which can then improve its treatment. In sum, evidence of AN in non-Western countries in Africa and Asia challenges the current definition of this disorder and suggests the diagnostic criteria be altered to accommodate cross-cultural causes of this psychopathology.
Abdollahi, P. & Mann, T. (2001). Eating disorder symptoms and body image concerns in Iran: Comparisons between Iranian women in Iran and in America. International Journal of Eating Disorders, 30(3), 259-268.
Abou-Saleh, M. T., Younis, Y., & Karim, L. (1998). Anorexia nervosa in an Arab culture. International Journal of Eating Disorders, 23(2), 207-212.
Bennett, D., Sharpe, M., Freeman, C., & Carson, A. (2004). Anorexia nervosa among female secondary school students in Ghana. British Journal of Psychiatry, 185, 312-317.
Chandra, P. S., Shah, A., Shenoy, J., Kumar, U., Varghese, M., Bhatti, R. S., & Channabasavanna, S. M. (1995). Family pathology and anorexia in the Indian context. International Journal of Social Psychiatry, 41, 292-298.
Gupta, M. A., Chaturvedi, S. K., Chandarana, P. C., & Johnson, A. M. (2001). Weight-related body image concerns about 18—24-year-old women in Canada and India: An empirical comparative study. Journal of Psychosomatic Research, 50, 193-198.
Hoek, H. W. & van Hoeken, D. (2003). Review of the prevalence and incidence of eating disorders. International Journal of Eating Disorders, 34(4), 383-396.
Khandelwal, S. K., Sharan, P., & Saxena, S. (1995). Eating disorders: An Indian perspective. International Journal of Social Psychiatry, 41, 132-146.
Klein, D. A. & Walsh, B. T. (2003). Eating disorders. International Review of Psychiatry, 15(3), 205-216.
Kugu, N., Akyuz, G., Dogan, O., Ersan, E., & Izgic, F. (2006). The prevalence of eating disorders among university students and the relationship with some individual characteristics. Australian and New Zealand Journal of Psychiatry, 40, 129-135.
Lee, S. (1995). Self-starvation in context: Towards a culturally sensitive understanding of anorexia nervosa. Social Science Medicine, 41(1), 25-36.
Lee, S., Lee, A. M., Ngai, E., Lee, D. T. S., & Wing, Y. K. (2001). Rationales for food refusal in Chinese patients with anorexia nervosa. International Journal of Eating Disorders, 29(2), 224-229.
Sjostedt, J. P., Schumaker, J. F., & Nathawat, S. S. (1998). Eating disorders among Indian and Australian university students. The Journal of Social Psychology, 138(3), 251-257.
Suhail, K. & Nisa, Z. (2002). Prevalence of eating disorders in Pakistan: Relationship with depression and body shape. Eating and Weight Disorders, 7, 131-138.
Uzun, O., Gulec, N., Ozsahim, A., Doruk, A., Ozdemir, B., & Caliskan, U. (2006). Screening disordered eating attitudes and eating disorders in a sample of Turkish female college students. Comprehensive Psychiatry, 47, 123-126.