Factors associated with substance use in adolescents with eating disorders

Abstract

Purpose: To examine the prevalence and potential risk factors associated with substance use in adolescents with eating disorders (EDs).

Methods: This cross-sectional study included 290 adolescents, aged 12-18 years, who presented for an initial ED evaluation at The Eating Disorders Program at the University of Chicago Medicine between 2001 and 2012. Several factors including DSM-5 diagnosis, diagnostic scores, and demographic characteristics were examined. Multinomial logistic regression was used to test associations between several factors and patterns of drug use for alcohol, cannabis, tobacco, and any other substance.

Results: Lifetime prevalence of any substance use was found to be 24.6% in those with anorexia nervosa, 48.7% in bulimia nervosa (BN), and 28.6% in ED not otherwise specified. Regular substance use (monthly, daily, and bingeing behaviors) or a substance use disorder was found in 27.9% of all patients. Older age was the only factor associated with regular use of any substance in the final multinomial model. Older age and non-white race was associated with greater alcohol and cannabis use. Although binge-purge frequency and bulimia nervosa diagnosis were associated with regular substance use in bivariate analyses, gender, race, and age were more robustly associated with substance use in the final multinomial models.

Conclusions: Co-morbid substance use in adolescents with EDs is an important issue. Interventions targeting high-risk groups reporting regular substance use or substance use disorders are needed.

Keywords: Adolescents; Alcohol; Anorexia nervosa; Bulimia nervosa; Cannabis; DSM-5; Eating disorders; Substance use; Substance use disorder; Tobacco.

Figure 1. The proportion of Eating Disordered Adolescents with Regular Substance Use out of N = 290 by Diagnosis

NB: chi-square tests examining differences between diagnostic groups for regular use versus no use were non-significant.