PREDICTORS OF RESPONSE AND NONRESPONSE TO LIGHT TREATMENT FOR WINTER DEPRESSION

Objective: The authors' goal was to determine whether the pattern and severity of depressive symptoms predict response to light treatment fo r seasonal affective disorder. Method: Subjects with winter depression (N=103) were given bright light treatment. Seventy-one were classifie d as responders, 15 as nonresponders, and 17 as partial responders. Us ing depression rating scale data and correlational and multivariate an alysis, the authors sought predictors of response in baseline symptom and scale scores. Results: Responders were characterized by atypical s ymptoms, especially hypersomnia, afternoon or evening slump, reverse d iurnal variation (evenings worse), and carbohydrate craving. By contra st, nonresponders were characterized mainly by melancholic symptoms: r etardation, suicidality, depersonalization, typical diurnal variation (morning worse), anxiety, early and late insomnia, appetite loss, and guilt. The ratio of atypical to classical symptoms of depression, rath er than severity per se, best predicted treatment outcome for the grou p as a whole. Pretreatment expectations were positively correlated wit h improvement on the Hamilton Depression Rating Scale but not on a sup plementary scale of atypical symptoms. Conclusions: Light-responsive s easonal affective disorder is distinguished by a dominant atypical sym ptom profile closely associated with depressed mood. Nonresponders for m a clinically distinct group with melancholic features. The patient's symptom profile, therefore, should be considered when diagnosing seas onal affective disorder and selecting treatment.