Predicting distress-induced eating with self-reports: mission impossible or a piece of cake?

Comments on the original article, "Assessing yourself as an emotional eater: Mission impossible?" by C. Evers, D. T. D. de Ridder, and M. A. Adriaanse (see record 2009-20990-009). Results of a functional MRI study (Bohon, Stice, & Spoor, 2009) contradict the assertion that it is "impossible" to self-assess emotional eating because the self-report emotional eating scale of the Dutch Eating Behavior Questionnaire (DEBQ-em) predicted important individual differences in reward response during negative moods. Evers et al advance their argument in the context of results of four experiments where self-reported "emotional eaters" (DEBQ-em) did not eat more food during emotional encounters as compared to control conditions or "no emotional eaters." However, the core characteristic of emotional eaters is not that they eat so much during distress (though binge eaters may do), but that they do not show the typical stress response of eating less (the typical stress response being loss of appetite because of physiological effects that mimic satiety) (Gold & Chrousos, 2002). Accordingly, the moderator effect of emotional eating during distress would be that. "No emotional eaters" eat less and "emotional eaters" eat the same or more compared to control conditions. Close inspection of the results of Evers et al reveals that their "no emotional eaters" did not show the typical stress response of eating less. This opens the possibility that the null findings of Evers et al may be simply explained by misclassification of "no emotional eaters" versus "emotional eaters" because of their use of median splits (a procedure notorious for possible misclassification of subjects into distinct groups). (PsycINFO Database Record (c) 2010 APA, all rights reserved).