A limited amount of research, primarily conducted in Western countries, has suggested that higher socioeconomic status (SES) is associated with higher risk of eating disorders (EDs). However, little is known about this association in Asian countries. We examined the association of SES with disturbed eating behavior (DEB) and related factors in Korean adolescents.
A nationwide online panel survey was conducted in a sample of adolescents (n = 6,943, 49.9% girls). DEB was measured with the 26-item Eating Attitudes Test (EAT-26). Participants who scored ≥20 on the EAT-26 were considered to have DEB. Participants’ SES was determined based on self-reported household economic status.
The prevalence of DEB was 12.7%: 10.5% among boys and 14.8% among girls. Both boys and girls with DEB were more likely to perceive themselves as obese, experience higher levels of stress, and have lower academic achievement. The risk for DEB was significantly higher in boys of higher SES than in those of middle SES (OR = 1.45, 95%CI = 1.05–1.99 for high SES; OR = 5.16, 95%CI: 3.50–7.61 for highest SES). Among girls, higher risk of DEB was associated with the highest and lowest SES (OR = 1.52, 95%CI: 1.13–2.06 for lowest SES; OR = 2.22, 95%CI: 1.34–3.68 for highest SES).
Citation: Lee H-J, Park S, Kim C-i, Choi D-w, Lee JS, Oh SM, et al. (2013) The Association between Disturbed Eating Behavior and Socioeconomic Status: The Online Korean Adolescent Panel Survey (OnKAPS). PLoS ONE 8(3): e57880. https://doi.org/10.1371/journal.pone.0057880
Editor: Stephanie T. Broyles, Pennington Biomedical Research Center, United States of America
Received: July 10, 2012; Accepted: January 29, 2013; Published: March 5, 2013
Copyright: © 2013 Lee et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This research was supported by a grant (09082KFDA904) from Korea Food & Drug Administration in 2009. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have the following competing interest: Doo-won Choi is an employee of Embrain Co.Ltd. There are no patents, products in development or marketed products to declare. This does not alter the authors' adherence to all the PLOS ONE policies on sharing data and materials.
Eating disorders (EDs) such as anorexia nervosa, bulimia nervosa, and binge ED are serious and growing health problems among adolescents . Although a wide range of estimates of the prevalence of EDs in adolescents (3.2%–26.0%) has been reported based on studies using different methods in different settings and in different countries –, it is clear that the incidence and prevalence of EDs in adolescents has been on the increasing during recent decades , . Thus, it is critical that we better understand and identify adolescents at increased risk for EDs for purposes of early detection and appropriate management.
EDs are a multifaceted problem associated with various interacting factors related to adolescence, including biological, psychological, familial, and sociocultural variables. As risk factors in a review, perceived body image is strong risk factor and next being overweight is a well-known factor, and screen time including media and internet use were associated with the risk of eating disorders . Residential region  and physical activity  may also affect the development of eating disorders. Among these factors, the association between socioeconomic status (SES) and EDs is of particular interest given that EDs have been shown to be more prevalent in white females of higher SES in Western countries –. Indeed, several studies have reported such trends not only among adolescent girls but also among adolescent boys; EDs were more prevalent in upper-middle-SES adolescents than in lower-SES adolescents –. However, these studies are limited in that they often failed to adequately sample from lower SES groups – even though higher rates of EDs have been observed in girls with lower SES than in their counterparts with higher SES . Recent large community- or population-based studies have reported inconsistent results , .
Contemporary South Korean culture can be characterized as dynamic, with rapid growth and diversification in areas such as industrialization, urbanization, and economics. Concurrently the prevalence of obesity is increasing, especially with regard to adolescents. Based on Korea National Health and Nutrition Examination data the ratio of Korean adolescents suffering from obesity (BMI of 25 or higher) increased to 11.3% in 2009 up from 5.8% in 1997 . This increase is significant and worrying even though the prevalence of obesity remains lower in Korea than other western countries. These changes clearly warn of the likelihood of an increasing prevalence of eating disorders in Korea. Other Asian countries will likely face these situations in the near future as they too embark on further development and globalization. Therefore the results of this research may suggest public health policy directions, further research and fields of enquiry well beyond Korea’s national boundaries.
Our understanding of the epidemiology of EDs and its associations with SES among adolescents in Asian countries is very limited , . Causal conditions, if (partially) socio-culturally based are also likely to be both more similar between Asia and Korea, and more divergent between Western countries and Korea. The purpose of this study was to estimate the prevalence of disturbed eating behavior (DEB) in adolescents using a nationally nationwide sample of Korean adolescents and to examine associations between SES and DEB.
Materials and Methods
This study used cross-sectional baseline data from the Online Korean Adolescent Panel Survey (OnKAPS). The purpose of OnKAPS was to estimate the number of adolescents at risk for EDs and to identify the risk factors associated with these disorders. The study was initiated in 2009 and follow-up surveys were conducted every 2 years. A nationwide Korean adolescent panel was selected according to the distribution of this population across the 16 regions, adolescent age groups, and genders reported in the 2005 Korean Population Census. Our Candidate panels were members of Embrain Research Co. Potential respondents were invited to take part in the survey, which they could access upon agreement to participate. Of the 16,090 Korean students who received an invitation email, 9,327 contacted the survey webpage. A total of 7,711 individuals completed all survey questions. The sample used for analysis included 6,943 students aged 12–18 years (3,476 boys and 3,467 girls). We excluded 768 students who 1) did not match with basic information such as age or region, or 2) answered with “I don’t know” for our interesting variables in this study, or 3) recorded an extreme duration of web-survey (less than 5 minutes, over than 50 minutes). The response rate was 48%.
This survey was conducted by Embrain Online Research Company (Embrain Co., Ltd, Seoul, Korea) between October and November 2009. Use of this on-line research design was made possible by the 100% rate of Internet adoption in Korea . This study was based on baseline data from the OnKAPS.
This survey was approved by the Korea Food and Drug Administration Institutional Review Board, and all participants and their mother or father provided written informed consent before participating. All informed consent was obtained by internet or email.
OnKAPS questionnaires included questions addressing three domains: 1) socio-demographic characteristics (e.g., age, gender, height, weight, school type, economic status, and academic achievement); 2) eating disturbances and weight perceptions; and 3) dietary habits, physical activity, screen time (TV, video games, Internet), and stress level.
The questionnaire was designed to be answered in 15–20 minutes to ensure high-quality responses. This timeframe is suggested in order to avoid survey fatigue and maintain respondent attention on survey items . The mean survey time was 16.0 minutes, and the questionnaire had been pilot-tested and revised to clarify the meanings of questions.
Eating Attitudes Test-26 (EAT-26)
The Eating Attitudes Test-26 (EAT-26) is the most frequently used tool for identifying adolescents with DEB in a community setting , . The EAT-26 was originally developed in the USA and based on American English. It has been translated into many languages. A validated Korean version of the EAT-26 for adolescents was used in this study . Based on the results of the pilot test, the phrasing and literal translation of 4 questions seemed to be problematic and unclear in our online survey. Therefore we added an explanation with parenthesis to the Korean version of the EAT-26 to clarify the meaning of the following four items: “Display self-control around food,” “Give too much time and thought to food,” “Like my stomach to be empty,” and “Enjoy trying new rich foods”. The EAT-26 incorporates 26 items in a 4-point Likert scale, where the participants indicate the extent of their agreement or disagreement. Answers range from 3: always, 2: usually, 1: often, 0: sometimes/rarely to never (e.g.: ‘I engage in dieting behavior’). The responses on the 26 items are summed at the end and a total score, ranging from 0 (minimum) to 78 (maximum), is extracted. DEB was defined as scores of 20 or more on the EAT-26, which has been shown to be associated with abnormal eating attitudes and behavior –. Cronbach’s alpha of EAT-26 in our sample was 0.85.
SES was determined based on responses to the question, “What is your household’s economic status?” The response categories were highest, high, middle, low, and lowest. To test the reliability of reported SES, we randomly selected 50 participants and asked their parents about their annual household income via phone. We found a significant correlation (r = 0.64, p<0.0001) between reported SES and annual income (Supplementary table1).
Participants’ height and weight and the size of their school-uniform pants were assessed. Body mass index (BMI = weight (kg)/squared height (m2)) was calculated based on self-reported weight and height. Weight status was defined according to data from the 2007 Korean National Growth Charts on BMIs by age (in months) and gender: a) underweight (BMI <5th percentile); b) normal (5th percentile ≤ BMI <85th percentile); or c) overweight (BMI ≥85th percentile) .
Data on age, gender, school type, academic achievement, perceived economic status, stress level, perceived body weight, physical activity, and screen times were collected. The questions were as follows:
“What type of school and class do you attend?”; the response categories were single-gender school, single-gender class in coeducational school, or coeducational class in coeducational school. “How much stress do you usually feel?”; the response categories were highest, high, middle, low, and lowest. “What do you think your weight status is?”; the response categories were very thin, thin, normal, fat, and very fat. “Do you exercise three or more times per week and for 30minutes or more per session?”; the response categories were yes or no. “How often do you devote more than 2 hours per day to screen time (e.g., watching TV, playing video games, connecting to the Internet); the response categories were usually, sometimes, and never. For purposes of analysis, we collapsed the five levels of stress and academic achievement (highest, high, middle, low, and lowest) into three (high, middle, and low). The perceived weight status response data was also collapsed into 3 groups during analysis.
Due to the significant interaction between gender and SES (P for interaction <0.001), analyses were stratified by gender. Descriptive statistics were used to summarize the characteristics of the study sample. Numerical data was presented as mean values and standard error (SE). Proportion and SE of proportion was calculated for categorical variables. To compare the characteristics of Korean adolescents with and without DEB, t-tests and chi-square tests were performed for continuous and categorical variables, respectively.
Logistic regression models were used to examine the association between SES and DEB. Obesity, perceived weight status, screen time, stress, and academic achievement was significantly correlated with eating disturbance (P<0.05). Residential region (i.e. metropolitan, rural) could be associated with the risk of DEB . Physical activity could affect development of eating disorder . Multivariate model (MV) 1 adjusted for demographic and social factors: region (metropolitan area, middle–small city, rural area), school type (coeducation bi-gender classroom, coeducation single gender classroom, single gender school), BMI (underweight, normal, overweight), academic achievement (high, middle, low), stress level (high, middle, low), regular physical activity(≥30 min, 3days or more/week) (yes, no), and screen time in usual days (usually, sometimes, never). MV2 additionally was adjusted for perceived weight status in MV1 to observe the effect of this factor.
The prevalence (%) and 95% Confidence Interval of DEB were calculated according to SES by gender. Additionally, the proportion (%) and 95% Confidence Interval of those who were simultaneously overweight, very stressed, perceived themselves as fat, and had high academic achievement were calculated according to SES by gender, respectively. All analyses were performed using SAS 9.2 (Cary, NC). P-values less than 0.05 (two-tailed test) were considered to indicate statistical significance.
The prevalence of DEB in this national sample of Korean adolescents was 12.7% (95%CI: 11.9–13.4), and it was higher among Korean adolescent girls than boys (14.8% (95%CI: 13.6–16.0) vs. 10.5% (95%CI: 9.5–11.5)).
Table 1 shows the characteristics of the study subjects by gender and DEB. Boys with DEB were more likely to be overweight, attend a coeducational or single-gender class, perceive themselves as fat, experience high levels of stress, have lower academic achievement, and report less screen time. Girls with DEB were more likely to be taller and heavier, perceive themselves as fat, experience high levels of stress, and have lower academic achievement.
We found an inverse association between SES and obesity. The prevalence of overweight individuals was highest in the lowest SES group and lowest in the highest SES group. By contrast, the proportion of individuals with high levels of academic achievement was highest among the highest SES group and lowest among the lowest SES group. The distribution of these factors by SES was similar for both genders, but the pattern was stronger in girls than in boys (Fig. 1). The proportion of overweight individuals, those who perceived themselves as fat, and those with high levels of stress also seemed to be highest in the lowest SES group in both genders. The prevalence of DEB was highest in the highest SES group, followed by the lowest SES group. The distribution of DEB across the five SES groups differed by gender: it was U-shaped in girls and J-shaped in boys (Fig. 1). We found a significant association between SES and DEB in the study sample (Table 2). Boys in the high or highest SES groups were at higher risk for DEB than were those in the middle SES group after adjusting for potential confounders (high vs. middle SES: OR = 1.45, 95%CI = 1.05–1.99; highest vs. middle SES: OR = 5.16, 95% CI = 3.50–7.61). On the other hand, girls in the lowest and highest SES groups had a higher risk of DEB than did those in the middle SES group (lowest vs. middle SES: OR = 1.52, 95%CI = 1.13–2.06; highest vs. middle: OR = 2.22, 95% CI = 1.34–3.68 in MV2).
We calculated the prevalence (%) and 95% Confidence Interval of DEB among those who were overweight, were very stressed, perceived themselves as fat, and had high academic achievement according to SES by gender.
This nationwide cross-sectional online survey of Korean adolescents found an overall prevalence of DEB of 12.7%; the prevalence was higher among girls (14.8%) than among boys (10.5%). U-shaped and J-shaped associations between SES and DEB were observed in Korean girls and boys, respectively.
Studies using the EAT-26 with non-clinical samples found rates of DEB that were higher than or similar to those found in this study , . One study reported that the prevalence of DEB was 9.7% among 1,829 girls aged 12–18 years in Canada , and another study found the prevalence of DEB to be 22% in girls and 2.0% in boys in a sample of 666 students in the US . A more recent national study conducted in US high schools found that the prevalence of ED was 14.4% among girls and 3.8% among boys in a sample of 5,618 students . One recent Asian study found that the prevalence of DEB was 10.4% among 835 Taiwanese females,  and another found a prevalence of 3.9% among boys and 6.5% among girls in a sample of 2,382 secondary school students in Hong Kong . However, it is hard to compare these rates with those observed in the present study because the previous Asian studies were not conducted on a nationwide sample. According to Makino’s review , which considered various studies using different measures of EDs, the prevalence of EDs in non-Western countries appears to be increasing, and those with DEB during adolescence were at increased risk for EDs 10 years later . Moreover, EDs during adolescence may be associated with elevated risk for a broad range of physical and mental health problems in adulthood .
In addition, the prevalence of DEB in boys was high compared to the other study mentioned above. A review suggested adolescent boys with an eating disorders tend to be overweight, have body dissatisfaction (e.g. want to lose body fat, develop a leaner and more muscular body), and may be affected by gay culture’s increased emphasis on physical appearance . The pressure to be thin in Korea may affect boys’ eating behaviors because male K-pop stars are very slim and youthful. We assumed that appearance became a more important issue to boys. It is warranted for future research to study what exactly was associated with the risk of boys in detail.
Previous community- or school-based studies have reported diverse results about the association between SES and EDs , , , , –. Deleel’s study  found no significant differences in DEB across SES groups in a sample of 518 young females. Other studies have found an increased risk of ED in upper-SES individuals , , . On the other hand, several studies ,  found that disordered eating was more prevalent in the lower SES group.
Most recently, a national study among Mexican adolescents found that higher SES was associated with higher risk of DEB . However, our study found that the prevalence of DEB was elevated not only in the highest SES but also in the lowest SES group. A study conducted in one province in South Korea also found that both high and low SES were associated with DEB in girls in the fourth and seventh grades .
We can provide no clear explanation for the U- and J-shaped associations revealed in our results. We observed similar findings in a 2 year follow-up survey as in this study. One possible explanation is that the higher ED risk among those with the highest SES may be due to perfectionism . High academic achievement among those in the highest SES group in our study may support this hypothesis. However, this explanation may not be applicable to the lowest SES group in light of the lower academic achievement of those in this group. Given the high prevalence of overweight individuals and those with negative body images in the lowest SES group, we considered that these factors may be connected to higher risk for EDs –. These patterns were much stronger in girls than in boys. Although these explanations of our results may be plausible, other factors may be involved, and interactions among these factors may exist. Further studies are needed to examine the U-shaped association more clearly.
Our study has several strengths. This is the first nationwide survey on EDs in Korean adolescents. We also used the EAT-26, a highly sensitive and reasonably specific instrument with the ability to identify subjects at high risk for ED and to evaluate the risk for ED , .
This study has also several limitations. First, SES levels were determined from perceived self- reported data. The correlation between income level reported by parents and self-reported SES level reported by adolescent subjects was moderate (r = 0.6, p<0.01) in the sub-sample (n = 50). It was hard to get information concerning household income through asking their parents because of budget problems. The determination and inclusion of more exact household income data is warranted for future studies in order to confirm the results of our study. Second, our cross-sectional study as a baseline for OnKAP survey was not optimal for the investigation of the direction of casual inferences. Third, our study population was drawn from online panels, and data was collected via questionnaires. Respondents in online survey may be more conscious about health and eating disturbance issues, hence the prevalence may be overestimated. However, participants may be more comfortable using an online as opposed to an in-person reporting process. Although almost all Korean middle and high school students use the Internet, non-panel students may differ from panel participants due to the potential selection bias of an on-line panel , . Although we used a validated Korean version of the EAT-26, such bias might have led us to miscalculate the number of young people at risk. Furthermore, an accurate diagnosis would require a clinical evaluation . This study suggests that both the lowest and the highest SES are associated with increased risk for developing an ED and that these patterns are stronger in girls. Further research is necessary to understand why the girls at the extremes of SES display such behavior. Most importantly, more public attention should be devoted to adolescents in both the lowest and the highest SES groups to identify and manage EDs. Based on this outcome, public programs for preventing or treating of ED should focus on controlling obesity as well as correcting perceived weight status for the lower SES groups meanwhile focusing on correction of perceived weight status for the higher SES groups.
Data collection: DWC. Interpretation of data and critical revision of the manuscript for important intellectual content: HJL SWO JSL SMO EC SP CIK DWC HKP KIK. Administrative, technical, or material support: HKP KIK. Conceived and designed the experiments: SWO HJL. Analyzed the data: HJL SP. Wrote the paper: HJL.
- 1. Thompson JK, Smolak L. (2002) Body image, eating disorders, and obesity in youth (2nd Ed.), washington DC, US.
- 2. Tsai MR, Chang YJ, Lien PJ, Wong Y (2011) Survey on eating disorders related thoughts, behaviors and dietary intake in female junior high school students in Taiwan. Asia Pac J Clin Nutr 20: 196–205.
- 3. Tam CKM, Ng CFN, Yu CM, Young BWY (2007) Disordered eating attitudes and behaviours among adolescents in Hong Kong: Prevalence and correlates. Journal of Paediatrics and Child Health 43: 811–817.
- 4. Fan Y, Li Y, Liu A, Hu X, Ma G, et al. (2010) Associations between body mass index, weight control concerns and behaviors, and eating disorder symptoms among non-clinical Chinese adolescents. BMC Public Health 10: 314.
- 5. Chisuwa N, O’Dea JA (2010) Body image and eating disorders amongst Japanese adolescents. A review of the literature. Appetite 54: 5–15.
- 6. Makino M, Tsuboi K, Dennerstein L (2004) Prevalence of eating disorders: a comparison of Western and non-Western countries. MedGenMed 6: 49.
- 7. D’Souza CM, Forman SF, Austin SB (2005) Follow-up evaluation of a high school eating disorders screening program: Knowledge, awareness and self-referral. Journal of Adolescent Health 36: 208–213.
- 8. Jones JM, Bennett S, Olmsted MP, Lawson ML, Rodin G (2001) Disordered eating attitudes and behaviours in teenaged girls: a school-based study. Canadian Medical Association Journal 165: 547–552.
- 9. Haines J, Ziyadeh NJ, Franko DL, McDonald J, Mond JM, et al. (2011) Screening High School Students for Eating Disorders: Validity of Brief Behavioral and Attitudinal Measures. Journal of School Health 81: 530–535.
- 10. Rosen DS (2010) Identification and management of eating disorders in children and adolescents. Pediatrics 126: 1240–1253.
- 11. Dorian BJ GP (1999) The contributions of epidemiologic studies to the etiology and treatment of the eating disorders. Psychiatry Ann 29: 187–192.
- 12. Striegel-Moore Rh Fau - Bulik CM, Bulik CM (2007) Risk factors for eating disorders.
- 13. Preti A, Pinna C, Nocco S, Pilia S, Mulliri E, et al. (2007) Rural/Urban Differences in the Distribution of Eating Disorder Symptoms among Adolescents from Community Samples. Australian and New Zealand Journal of Psychiatry 41: 525–535.
- 14. Davis C, Kennedy SH, Ravelski E, Dionne M (1994) The role of physical activity in the development and maintenance of eating disorders. Psychological Medicine 24: 957–967.
- 15. Hoek HW, van Hoeken D (2003) Review of the prevalence and incidence of eating disorders. Int J Eat Disord 34: 383–396.
- 16. Wardle J, Griffith J (2001) Socioeconomic status and weight control practices in British adults. J Epidemiol Community Health 55: 185–190.
- 17. Story M, French SA, Resnick MD, Blum RW (1995) Ethnic/racial and socioeconomic differences in dieting behaviors and body image perceptions in adolescents. Int J Eat Disord 18: 173–179.
- 18. Palma-Coca O, Hernández-Serrato MI, Villalobos-Hernández A, Unikel-Santoncini C, Olaiz-Fernández G, et al.. (2011) Association of Socioeconomic Status, Problem Behaviors, and Disordered Eating in Mexican Adolescents: Results of the Mexican National Health and Nutrition Survey 2006. The Journal of adolescent health : official publication of the Society for Adolescent Medicine.
- 19. Crisp AH PR, Kalucy RS. (1976) How common is anorexia nervosa? A prevalence study.
- 20. Moya T, Fleitlich-Bilyk B, Goodman R (2006) Brief report: Young people at risk for eating disorders in Southeast Brazil. Journal of Adolescence 29: 313–317.
- 21. Fox KC, James NM (1976) Anorexia nervosa: a study of 44 strictly defined cases. N Z Med J 84: 309–312.
- 22. Kendell RH, DJ; Hailey, A; Babigian, (1973) Kendell RE, Hall DJ, Hailey A, Babigian HM.
- 23. Doyle J, Bryant-Waugh R (2000) Epidemiology.In: Lask, B.; Bryant-Waugh R. (eds.), Anorexia Nervosa and Related Eating Disorders in Childhood and Adolescence. 2nd ed. Psychology Press, East Sussex UK: 41–61.
- 24. Hoek H (1995) The distribution of eating disorders. In: Brownell, KD.; Fairburn, CG., editors. Eating disorders and obesity: A comprehensive handbook. Guilford Press; New York. 207–211.
- 25. Gibbons P (2001) The Relationship Between Eating Disorders and Socioeconomic Status: It’s Not What You Think. Nutrition Noteworthy 4.
- 26. Power Y, Power L, Canadas MB (2008) Low socioeconomic status predicts abnormal eating attitudes in Latin American female adolescents. Eat Disord 16: 136–145.
- 27. Korea CDC (2009) 2009 Korea National Health and Nutritional Examination Survey-result report.
- 28. (2008) Internet in Asia Pacific Region - A comparative report. Asia Pacific Internet Research Alliance (APIRA): http://www.apira.org (last accessed date 2012.12.1).
- 29. Galesic M, Bosnjak M (2009) Effects of Questionnaire Length on Participation and Indicators of Response Quality in a Web Survey. Public Opinion Quarterly 73: 349–360.
- 30. Garner DM OM, Bohr Y, Garfinkel P (1982) The eating attitudes test: psychometric features and clinical correlates. Psychol Med 12: 871–878 http://eat-26.com/Form/index.php.
- 31. Choi JH AD, Nam JH, Jo YK, Choi BY (1998) Reliability and validity of Eating Attitude Test-26 for Korean adolescents. Korean J Child & Adol Psychiatry 9: 91–97.
- 32. Yang SJ, Kim JM, Yoon JS (2010) Disturbed eating attitudes and behaviors in South Korean boys and girls: a school-based cross-sectional study. Yonsei Med J 51: 302–309.
- 33. Moon JS, Lee SY, Nam CM, Choi JM, Choe BK, et al. (2008) 2007 Korean National Growth Charts: review of developmental process and an outlook. Korean J Pediatr 51: 1–25.
- 34. Williams R, Lee C-K, Back K (2012) The prevalence and nature of gambling and problem gambling in South Korea. Social Psychiatry and Psychiatric Epidemiology: 1–14.
- 35. Muise AM, Stein DG, Arbess G (2003) Eating disorders in adolescent boys: a review of the adolescent and young adult literature. The Journal of adolescent health : official publication of the Society for Adolescent Medicine 33: 427–435.
- 36. Miotto P, De Coppi M, Frezza M, Preti A (2003) The spectrum of eating disorders: prevalence in an area of Northeast Italy. Psychiatry Research 119: 145–154.
- 37. Deleel ML, Hughes TL, Miller JA, Hipwell A, Theodore LA (2009) Prevalence of Eating Disturbance and Body Image Dissatisfaction in Young Girls: An Examination of the Variance across Racial and Socioeconomic Groups. Psychology in the schools 46: 767–775.
- 38. Thomas CL, James AC, Bachmann MO (2002) Eating attitudes in English secondary school students: influences of ethnicity, gender, mood, and social class. Int J Eat Disord 31: 92–96.
- 39. Rogers L, Resnick MD, Mitchell JE, Blum RW (1997) The relationship between socioeconomic status and eating-disordered behaviors in a community sample of adolescent girls. International Journal of Eating Disorders 22: 15–23.
- 40. Neumark-Sztainer D, Story M, Falkner NH, Beuhring T, Resnick MD (1999) Sociodemographic and personal characteristics of adolescents engaged in weight loss and weight/muscle gain behaviors: who is doing what? Prev Med 28: 40–50.
- 41. Wassenaar D, le Grange D, Winship J, Lachenicht L (2000) The prevalence of eating disorder pathology in a cross-ethnic population of female students in South Africa. European Eating Disorders Review 8: 225–236.
- 42. Darby A, Hay P, Mond J, Rodgers B, Owen C (2007) Disordered eating behaviours and cognitions in young women with obesity: relationship with psychological status. Int J Obes (Lond) 31: 876–882.
- 43. Haines J N-SD (2006) Prevention of obesity and eating disorders: a consideration of shared risk factors. Health EducRes 21: 770–782.
- 44. D N-S (2009) Preventing obesity and eating disorders in adolescents: what can health providers do? J Adolesc Health 44: 206–213.
- 45. Wong Y, Chang Y-J, Tsai M-R, Liu T-W, Lin W (2011) The Body Image, Weight Satisfaction, and Eating Disorder Tendency of School Children: The 2-Year Follow-up Study. Journal of the American College of Nutrition 30: 126–133.
- 46. Dotti A LR (1998) Validation and reliability of the Italian EAT-26. Eat Weight Disord 3: 188–194.
- 47. Clinton JD (2001) Panel Bias from Attrition and Conditioning: A Case Study of the Knowledge Networks Panel. Stanford.
- 48. Bartels LM (1999) Panel Effects in the American National Election Studies. Political Analysis 8: 1–20.
- 49. Williams P, Hand D, Tarnopolsky A (1982) The problem of screening for uncommon disorders - a comment on the Eating Attitudes Test. Psychol Med 12: 431–434.