Investigating Cognitive Deficits as Risk Factors for Developing Eating Disorders During Adolescence

ABSTRACT

Adolescents with eating disorders (ED) suffer from deficits in executive functions and “theory of mind.” It is unclear whether these indicate state or trait characteristics. We examined cognitive functioning in 150 adolescents, comparing those at high risk and those not at risk for ED. Deficits in set shifting and in theory of mind were found in all high-risk groups. Adolescents at high risk for bulimia were found to be higher in impulsivity and in theory of mind deficits, compared to adolescents at high risk for anorexia. These are trait characteristics rather than state vulnerabilities and preexist the development of ED.

Acknowledgments

This research is based on a Ph.D. dissertation by the first author, supervised by the second. Preliminary versions of this paper were presented at the International Conference on Eating Disorders (ICED 2014), New York, USA, March, 2014, and at the International Convention of Psychological Science, Amsterdam, The Netherlands, March, 2015. We thank our anonymous reviewers for their insightful comments on a previous draft.

Notes

1. An anonymous reviewer has questioned this manner of presenting the data, stemming from WCST performance, in Figure 2, writing that “a cumulative function will show an increase even in the absence of learning.” This is true, and is also applicable to the way we present the data stemming from IGT performance in Figure 3. We stress that this is a deliberate data-analytic strategy on our part, to build on monotonic curves—and then to look for an interaction with Group. We are less interested in learning per se, and more interested in seeing how cumulative performance changes over time (here, categories, in Figure 3, blocks) as a function of group affiliation.

2. An anonymous reviewer has commented that our policy of dividing participants into those exhibiting learning, and those not exhibiting learning, and then analyzing learning is somewhat circular. To clarify, then: We have provided a definition of learning in this task, based on comparing RA to RP choices, providing in Figure 3 an example of what we mean here; we subsequently look at the performance of each such group. Our primary interest is, of course, to see whether there is a three-way interaction of Group × Block × Learning. Figure 3c then allows us to inspect the different learning curves (as defined here) pertaining to each of our four groups of participants.