Abstract
Sleep disruptions may be risk factors for eating disorder symptoms. However, mean estimates of sleep characteristics may not be ideal metrics, considering many individuals have irregular sleep patterns. Although variability in sleep-wake timing is associated with irregular eating patterns among individuals with eating disorders (Linnaranta et al., 2020), studies of sleep intraindividual variability (IIV) and eating pathology are limited. To fill this gap, we evaluated two indices of sleep IIV from weekdays-to-weekends and their relation to eating disorder risk. The sample was drawn from the 2018-2019 Healthy Minds Study (N = 25,879). We conducted two binary logistic regressions predicting screening positive for an eating disorder from IIV indices, adjusting for depression and mean sleep time. Greater IIV(timing) (i.e., greater shifts in the timing of the sleep midpoint between weekdays and weekends) was associated with greater odds of screening positive for an eating disorder (OR = 1.049). Greater IIV(duration) (i.e., greater shifts in duration of sleep between weekdays and weekends) was associated with lesser odds (OR = 0.967); however, suppressor effects of depression likely explain this finding. Greater mean duration of sleep was associated with lower odds of screening positive for an eating disorder in both models. Results are consistent with theories emphasizing the importance of sleep IIV in conceptualizing psychopathology risk; however, effect sizes were small and may have limited clinical significance. By only asking about mean sleep duration, clinicians may miss other sleep-related predictors of eating disorder psychopathology. Findings support investigating techniques to regularize sleep patterns among people with disordered eating.