BACKGROUND: The behavioral symptoms of Generalized Anxiety Disorder (GAD) are not well characterized. This study examines behavioral symptoms in patients with GAD compared to healthy participants, their change during behavioral therapy, and their role for predicting short- and long-term outcome.

METHODS: Secondary data analysis of 56 patients with DSM-IV GAD from a randomized controlled trial testing worry exposure (n = 29) and applied relaxation (n = 27), compared to 33 demographically matched healthy participants. Participants reported on attempts to control or prevent worry, specifically cognitive and behavioral avoidance, safety behavior, and reassurance, along with other GAD symptoms. The Hamilton Anxiety Scale served as immediate (post therapy) and the Penn State Worry Questionnaire as immediate and long-term (6-/12-month follow-up) treatment outcome measure.

RESULTS: GAD patients engage significantly more in attempts to control or prevent worry as reflected in cognitive and behavioral avoidance, safety behavior, and reassurance seeking than healthy comparison participants. Behavior therapy significantly reduces these behavioral strategies without substantial indication of differential effects of treatment type. However, only patients remitting from GAD reach the low symptom level of healthy participants. The initial level of behavioral symptoms is irrelevant for immediate treatment success, but higher degrees of cognitive and behavioral avoidance and safety behavior at the end of treatment predict worse long-term outcome.

CONCLUSIONS: Behavioral symptoms appear to be relevant features in GAD that improve with successful treatment. Further research is warranted to examine whether inclusion of behavioral symptoms in the definition of GAD would have beneficial effects on diagnostic recognition and treatment.