Morning and evening light treatment of seasonal affective disorder: response, relapse and prediction.

Patients with seasonal affective disorder were randomly assigned to treatment with light in the morning (9.00-12.00 a.m.; n = 16; ML) or evening (6.00-9.00 p.m.; n = 11; EL). An intensive 24-day assessment procedure revealed the same response rates: 57% for ML, 50% for EL. During the rest of the winter season a relatively low relapse rate of 54% was found. No differences between ML and EL were found in the time course of depressed mood or fatigue. A significant negative correlation was found between diurnal variation during baseline and therapeutic response: the larger the diurnal variation the less the response, indicating a potential negative predictive value for this symptom. There were no significant correlations between baseline fatigue or hypersomnia and response.