General and Symptom-Specific Illness Duration and Course as Predictors of Symptom Severity in Anorexia Nervosa.

OBJECTIVE: Illness duration has been examined as a predictor of anorexia nervosa (AN) outcomes to mixed results, yet is frequently used to make treatment and prognosis decisions. More specific metrics, such as prior illness course (e.g., continuous vs. intermittent symptoms) and symptom-specific duration or course, may more effectively predict concurrent and longitudinal outcomes. METHOD: Adults with acute or weight-restored AN (N = 75) completed a measure assessing duration and course (e.g., continuous, intermittent) of their eating disorder in general and of specific symptoms (low weight, restrictive eating, weight/shape preoccupation). Hierarchical linear regressions predicted the severity of each symptom (BMI, restrictive eating, shape concern, weight concern) at baseline and one-year follow-up from covariates (Step 1), general illness duration and course (Step 2), and symptom-specific duration and course (Step 3). RESULTS: Cross-sectionally, symptom-specific predictors improved model fit for all outcomes (ps = < 0.001-0.043). A more continuous course of restrictive eating was associated with more frequent baseline restrictive eating (p = 0.002) and longer duration (p = 0.009) and a more continuous course (p < 0.001) of weight/shape preoccupation was associated with higher baseline shape concern. General illness duration and course did not predict symptom severity after accounting for symptom-specific factors. Longitudinally, no measure of illness or symptom duration or course predicted follow-up symptom severity. CONCLUSIONS: Courses of specific symptoms may be more meaningful symptom severity markers in AN than general illness duration or course, particularly for restriction and shape concern. Findings highlight the clinical utility of assessing symptom-specific chronicity and of interrupting eating disorder symptoms in treatment, given the impact of continuous symptom expression.