(PS5-49) Dismantling Loss of Control Eating Episodes: Impacts of Perceived and Actual Episode Size on Eating Disorder Severity and Treatment Outcomes in Anorexia Nervosa and Bulimia Nervosa

Considerable evidence suggests that the sense of loss of control (LOC) is the most salient facet of binge eating episodes. However, the DSM-5 definition of binge eating also requires the consumption of an objectively large amount of food (OBEs). In contrast, the ICD-11 definition of binge eating has no size requirements and encompasses OBEs and subjectively large LOC episodes (SBEs), where the amount consumed is not objectively large, but still perceived as excessive. In addition, although not assessed by traditional measures nor previously described in the literature, ICD-11 binge eating also includes LOC episodes that are not objectively large and are not perceived as excessive (NSOBEs). Several studies have shown that SBEs, independent of OBEs, are a marker of eating disorder psychopathology and persist post-treatment in bulimia nervosa (BN). However, less is known about changes in OBEs, SBEs, and NSOBEs over treatment and their impact on outcomes in anorexia nervosa (AN) and BN. 

Here, we examined the prevalence and predictive value of LOC episodes in patients with AN (n = 112) and BN (n = 114) who completed the Eating Disorder Examination Questionnaire (EDE-Q) upon admission to residential treatment and at discharge. 

At admission, more BN than AN patients endorsed OBEs (χ² = 126.69, p < 0.001), but  SBEs (p = 0.20) and NSOBEs (p = 0.28) were endorsed at similar rates across groups. More BN patients vs. AN patients endorsed NSOBEs at discharge (χ² = 7.32, p = 0.027). Changes in OBE frequency differed by group, decreasing over treatment in BN, but increasing in AN (Group x Time p < .001). Changes in SBEs also differed, increasing in AN but not in BN (Group x Time p < .001). Both groups showed a small, non-significant increase in NSOBEs (Time p > .40). 

In the full sample, the relationship between clinical severity and total number of LOC episodes perceived as excessive (i.e., OBEs+SBEs) varied by group and timepoint: higher total OBEs+SBEs was associated with higher EDE-Q global scores only in BN, and only at discharge (p = .005). In contrast, higher total non-objectively large LOC episodes (i.e., SBEs+NSOBEs) was associated with more severe global pathology (p = .006) across treatment and groups. 

These findings suggest that LOC episodes have more complex predictive value than previously understood. In BN, but not AN, OBEs decreased with treatment and a higher frequency of perceived-excessive LOC episodes at discharge were linked to persistently high global symptoms at discharge in BN. However, LOC episodes that are not objectively large increased over treatment in AN, and were linked to more severe global symptoms at both timepoints across groups. Notably, the current version of the EDE-Q does not measure SBEs or NSOBEs. Our results suggest that more precise assessment and treatment of LOC eating, across all episode sizes, is needed.  Addressing distorted perceptions of intake and feelings of LOC may enhance the effectiveness of transdiagnostic interventions.