Background
Despite being characterized primarily by disturbances in eating behavior, relatively little is known about specific eating behaviors in anorexia nervosa (AN) and how they relate to different emotional, behavioral, and environmental features.
Methods
Women with AN (n = 118) completed a 2-week ecological momentary assessment (EMA) protocol during which they reported on daily eating- and mood-related patterns. Latent profile analysis was used to identify classes of eating episodes based on the presence or absence of the following indicators: loss of control; overeating; eating by oneself; food avoidance; and dietary restraint.
Results
The best-fitting model supported a 5-class solution: avoidant eating; solitary eating; binge eating; restrictive eating; and loss of control eating. The loss of control and binge eating classes were characterized by high levels of concurrent negative affect and a greater likelihood of engaging in compensatory behaviors. The restrictive eating class was associated with the greatest number of concurrently-reported stressful events, while the avoidant and solitary eating episode classes were characterized by relatively few accompanying stressful events. Body checking was least likely to occur in conjunction with restrictive eating behaviors.
Conclusions
Results support the presence of discrete types of eating episodes in AN that are associated with varying degrees of negative affect, stress, and behavioral features of eating disorders. Loss of control and dietary restriction may serve distinct functional purposes in AN, as highlighted by their differing associations with negative affect and stress. Clinical interventions for AN may benefit from targeting functional aspects of eating behavior among those with the disorder.
Anorexia nervosa (AN) is a serious psychiatric illness associated with significant medical and psychosocial comorbidities (Hudson et al., 2007 and Pomeroy and Mitchell, 2002). AN is characterized primarily by disturbances in eating behavior, particularly restriction of energy intake relative to one's energy needs (American Psychiatric Association, 2013), yet relatively little is known about specific eating patterns in AN. With regard to typical eating behaviors in AN, evidence suggests that individuals with the disorder tend to consume fewer kilocalories and less fat than healthy controls when under observation in controlled laboratory conditions (Fernstrom et al., 1994, Gwirtsman et al., 1989, Hadigan et al., 2000, Mayer et al., 2012 and Sysko et al., 2005). However, results from one study that collected data via daily dietary recall based on ecological momentary assessment, or EMA (which addresses concerns about the artificial nature of the laboratory setting by collecting data in “real time” in the natural environment; Shiffman et al., 2008), suggest that the mean daily caloric intake of individuals with AN may be closer to nutritional recommendations than one might expect based on laboratory data (Burd et al., 2009). The extent to which this discrepancy is due to over-reporting in EMA, reactivity associated with laboratory conditions that results in reduced energy consumption, or some combination thereof is unclear. Nevertheless, given that existing psychological treatments of AN have thus far shown limited efficacy (Wilson et al., 2007), a better understanding of the context and associated features of eating episodes in AN could inform the development of more effective interventions for the disorder.
It has been proposed that eating disorder behaviors in AN are learned habits that become well-entrenched since they are persistently reinforced over time (Walsh, 2013). One hypothesized means of reinforcement may be via a reduction in negative affect that occurs subsequent to the behaviors. Self-report questionnaire data suggest that individuals with AN have difficulties tolerating negative emotions (Hambrook et al., 2011, Harrison et al., 2009 and Wildes et al., 2010), and that such difficulties may play a role in the occurrence of eating disorder behaviors (Espeset et al., 2012 and Racine and Wildes, 2014). Similarly, both laboratory data (Steinglass et al., 2010 and Wildes et al., 2012) and previous EMA data reported by our group (Engel et al., 2005, Engel et al., 2013 and Lavender et al., 2013a) have indicated that negative affect (and anxiety in particular) is associated with subsequent eating disorder cognitions and behaviors, including dietary restriction, binge eating and purging, and body checking. In spite of this apparent link between negative affect and eating patterns, it is currently unclear whether distinct emotional or behavioral cues are associated with different types of eating behaviors in AN.
The purpose of the current study was twofold: 1) to identify classes of eating episodes reported in the natural environment by women with AN using an empirical classification approach; and 2) to examine the emotional and behavioral context in which these classes of eating episodes occur. A secondary aim was to examine the extent to which AN diagnostic subtypes (i.e., restricting type vs. binge eating/purging type) differ with respect to self-reported frequencies of different classes of eating episodes. We hypothesized that distinct classes of eating episodes would be identified, characterized by varying combinations of loss of control while eating, overeating, eating by oneself, avoiding certain foods, and restricting food intake. In particular, we expected that classes of eating episodes involving loss of control and/or overeating would be associated with high levels of negative affect, consistent with the previous literature (Haedt-Matt and Keel, 2011). Conversely, classes of eating episodes involving restricted eating or food avoidance were expected to be associated with an increased likelihood of engaging in body checking and related behaviors (Lavender et al., 2013b), which may function as a method of reaffirming the effectiveness of restrictive behaviors. Finally, we expected that individuals with AN binge/purge subtype would be more likely to endorse the classes of eating episodes characterized by loss of control and/or overeating than individuals with AN restricting subtype.
