Longitudinal Predictors of Disordered Eating Among Emerging Adults in the U.S

Purpose: The National Comorbidity Study shows that eating disorders (EDs) affect all race/ethnic groups, ages, and sexes. Prevalence of EDs is 5-10% and is of concern to social workers because EDs are associated with psychosocial problems and role impairment, and are often undertreated. The most frequent ED diagnosis is ED-Not Otherwise Specified, in response to which the new DSM-5 proposes to add a diagnosis for Binge Eating Disorder. While epidemiological studies have reported on prevalence of disordered eating among teen and adult populations, little is known about how adolescent risk and protective factors may impact long-term risk for disordered eating among emerging adults. Human development theory posits, and empirical studies confirm, that the transition to adulthood is a sensitive period for exploring and establishing health habits and beliefs. The purpose of this study is to assess longitudinal risk factors for unsafe and extreme weight loss behavior (EWLB), binge eating, and EDs during the transition to adulthood in the U.S.

Methods: Data were drawn from the National Longitudinal Study of Adolescent Health, the largest nationally representative panel study of adolescents followed into adulthood ever undertaken in the U.S. The analytic sample consisted of participants in Wave1 (1994) and Wave3 (2001; ages 18-26) with valid sampling weights to ensure representativeness (N=15,170). Using STATA survey commands to adjust for the clustered sampling design, we conducted multivariate logistic regression procedures to examine the impact of psychosocial and behavioral risk factors measured at W1 (weight loss behaviors, depression, body image, Body Mass Index z-scores) on three indicators of disordered eating (EWLB, ED diagnosis, binge eating) measured at W3, controlling for age, race/ethnicity, SES, and family structure at W1. All analyses were stratified by sex.

Results:  Among women, EWLB at W1 was associated with more than threefold increases in odds of both ED diagnosis (OR=3.67, p<.01) and EWLB at W3 (OR=3.12, p<.01). Dieting during adolescence was also associated with increased odds of EWLB (OR=1.64, p<.01) among young women, while early depression more than doubled their odds of ED diagnosis (OR=2.13, p<.01) and binge eating at W3 (OR=2.43, p<.01). Among men, depression during their adolescence was associated with an increase in odds of ED (OR=2.88, p<.01) and binge eating (OR=3.49, p<.01) at W3; and body image distortion was associated with a more than fivefold increase in odds of ED at W3 (OR=5.69, p<.01). Adjusting for baseline EWLB, higher BMIz at adolescence was associated with increased odds of onset of EWLB among both men and women at W3 (OR=2.01 and 1.49, respectively, p<.01).

Implications:  Screening for EDs among emerging adults is an important part of comprehensive biopsychosocial assessment, and should occur for both sexes and regardless of weight status. Knowledge of early risk factors may improve detection and thus treatment of EDs, and should also be incorporated into ED prevention and education efforts by Social Workers. Promoting ED screening at this time presents a unique opportunity to address ED under-treatment as more emerging adults gain access to health care under the Affordable Care Act of 2010.