Is It Really All about Self-Esteem? A Study of Eating Disorders Among Chilean Female Adolescents

Background: One of the most common factors discussed in relation to the development of eating disorders among adolescents is that of self-esteem. Decades of research have found that adolescents, particularly females, with low self-esteem are at increased risk of eating disorders. However, recent research and theory have begun to challenge the primacy of self-esteem in the development of eating disorders. Despite increasing evidence that eating disorders affect individuals from various racial, ethnic, and socioeconomic backgrounds, much of the eating disorder literature continues to be limited to the study of White relatively affluent female populations. The purpose of the following study was to examine the relationship between self-esteem and eating disorders among Chilean female adolescents and explore other potential etiological factors that may explain the presence of these problems.

Methods: Data are from a NIDA-funded study of drug use and health behaviors among a community-sample of 1,021 adolescents (M=14.2 years, 49% female) from municipalities of mid- to low- socioeconomic status. Of the total sample, 420 females were included in the analysis. Youth completed 2-hr interviewer-administered questionnaires with comprehensive questions on substance use, individual, peer, familial, and contextual factors. The dependent variable, ever being diagnosed with an eating disorder (anorexia or bulimia nervosa), was coded as “1” for affirmative responses and “0” for negative. Predictor variables included physical appearance self-esteem, family involvement, symptoms of anxiety and depression, and substance use (ever use). Bivariate logistic regression models were utilized to test the independent relationship between ever being diagnosed with an eating disorder and each predictor. Multivariate logistic regression analysis was used to determine which factors, when considered simultaneously, were related to being diagnosed with an eating disorder. All analyses controlled for age of the youth and family SES.

Results: Among this community-sample of Chilean female adolescents, 3.8% had ever been diagnosed with an eating disorder, a rate comparable to that of White U.S. populations. As expected, participants with lower physical appearance esteem, higher symptoms of depression, higher symptoms of anxiety, less family involvement, and who tried cigarettes were more likely to have been diagnosed with an eating disorder. However, when all factors were considered simultaneously, only associations with symptoms of anxiety (OR=1.21, 95% CI=1.00-1.46) and cigarette use (OR=3.55, 95% CI=1.01-12.53) remained significant. Physical appearance esteem was no longer associated with being diagnosed with an eating disorder.

Implications: These findings support that eating disorders do not discriminate based on race/ethnicity or SES. Specifically, Chilean female adolescents of mid-to-low SES are an at-risk population for eating disorders. That physical appearance self-esteem was no longer associated with an eating disorder when all factors were considered supports recent theory that suggests self-esteem may not play as an important role in the development and persistence of eating disorders as once believed. This research supports that eating disorders, albeit maladaptive, may be a means to cope with negative affect, specifically anxiety. These findings may help to inform prevention and intervention programs for eating disorders that target Chilean female adolescents specifically, but possibly other populations as well.