by Melissa Munn-Chernoff, Ph.D.

Eating disorders and substance use disorders frequently co-occur (see previous blog for a review).  Studies have consistently shown that women with a history of bulimia nervosa (BN) or the binge-eating/purging subtype of anorexia nervosa (AN) are more likely to also have an alcohol use disorder and nicotine dependence than women with the restricting subtype of AN and women without an eating disorder. Further, women with the restricting subtype of AN have a higher prevalence of these substance use disorders compared with women without an eating disorder. Having both an eating and substance use disorder can make recovery more challenging,¹ yet treatments for individuals who experience both disorders do not currently exist. Thus, identifying potential mechanisms contributing to this co-occurrence is critical to assist in prevention and treatment.

One potential explanation is that eating disorders and substance use disorders share genetic factors. Indeed, twin studies have found that the strongest genetic associations occur between BN and related symptoms—including binge eating—and problematic alcohol use (see our previous blog). However, most existing research has not assessed a range of eating disorder and substance use traits to determine whether the shared risk is consistent across different types of eating disorders and specific substances. Understanding the specific patterns of association will provide important information regarding the role of genetics for individuals who experience both conditions.

In our recent paper,² we used existing data from multiple genome-wide association studies to assess genetic correlations between each eating disorder and various substance use traits. Overall, we included four eating disorder traits (AN, AN with binge eating, AN without binge eating, and a BN-like factor) and eight substance use traits, including nicotine use, alcohol use, and cannabis use. Positive genetic correlations indicated that the same genes are influencing both disorders in the same direction, whereas negative genetic correlations indicated that the same genes are influencing both disorders in the opposite direction.

Three significant patterns of association emerged. First, there was a significant positive genetic correlation between AN and alcohol use disorder (r_g = 0.18), which is consistent with prior twin studies. Second, there was a significant positive genetic correlation between AN and cannabis initiation (i.e., trying cannabis for the first time) (r_g = 0.23), as well as between AN with binge eating and cannabis initiation (r_g= 0.27). This is a unique finding not previously examined in twin studies. Finally, negative genetic correlations emerged between AN without binge eating and several smoking-related traits (r_gs ranged from -0.19 to -0.23), suggesting that some of the genetic factors influencing risk for both classes of traits are shared but operating in opposite directions. In short, these finding suggest that genes do explain, at least in part, the relation between eating disorders and problematic substance use.

Although these types of genetic studies cannot identify the specific genes influencing both disorders, they do provide initial evidence about patterns of shared genetic factors across different eating disorders and use of specific substances. Ultimately our goal is to use this information and results from larger genetic studies to inform the development of biologically informed prevention and treatment interventions.

References

1. Gregorowski C, Seedat S, Jordaan GP (2013). A clinical approach to the assessment and management of co-morbid eating disorders and substance use disorders. BMC Psychiatry;13:289

Munn-Chernoff MA, Johnson EC, Chou Y-L, Coleman JRI, Thornton LM, Walters RK, … Agrawal A (in press). Shared genetic risk between eating disorder- and substance-use-related phenotypes: Evidence from genome-wide association studies. Addiction Biology. https://doi.org/10.1111/adb.12880.