The role of affect in the maintenance of anorexia nervosa: evidence from a naturalistic assessment of momentary behaviors and emotion

The current study examines the relationship of affect and eating disorder behavior in anorexia nervosa (AN) using ecological momentary assessment. Participants were 118 adult females recruited at three sites from eating disorder treatment centers and community advertisements. All participants met full Diagnostic and statistical manual of mental disorders (4th ed.) criteria or subthreshold criteria for AN. Participants were provided handheld computers and asked to report positive affect, negative affect, loss of control (LOC) eating, purging, exercise, drinking fluids to curb appetite, and weighing one's self multiple times per day as well as dietary restriction once daily over a 2-week interval. Mixed-effects models were used to examine the extent to which affective states predict dietary restriction. In addition, we used two analytic approaches to compare affect before and after other eating disorder behaviors. We found that higher daily ratings of negative affect were associated with a greater likelihood of dietary restriction on subsequent days. When examining the single rating immediately before and after behaviors, we found that negative affect increased significantly after LOC eating, purging, the combination of LOC and eating/purging, and weighing of one's self. Using this same analytic approach, we also found negative affect to decrease significantly after the consumption of fluids to curb appetite and exercise. When examining the covariation of AN behaviors and negative affect assessed multiple times in the hours and minutes before the behaviors, we found negative affect significantly increased before LOC eating, purging, the combination of LOC eating/and purging, and weighing behavior. Negative affect also significantly decreased after the occurrence of these behaviors. These findings are consistent with the idea that that negative affect is potentially a critical maintenance mechanism of some AN symptoms, but that the analytic approach used to examine affect and behavior may have significant implications on the interpretation of findings.