Self-reported and actual physiological responses in social phobia.

OBJECTIVES: The aim of the current study was to compare physiological reactions and self-reports of bodily sensations for social phobics, clinically anxious and non-anxious controls across four tasks.
DESIGN: A within-group design was used in which each participant took part in four tasks. Two were designed to be demanding, either physically (riding an exercise bicycle) or mentally (mental arithmetic task), while two, a mental imagery task (personally relevant situation) and a social conversation, were designed specifically to be anxiety provoking.
METHODS: Of the 54 participants, 18 were generalized social phobics, 18 were clinically anxious but not socially phobic (8 with panic disorder, 6 with generalized anxiety disorder and 4 simple phobics), and 18 were non-anxious. Heart rate, skin conductance, and facial and neck temperatures were recorded continuously during four different tasks and rest periods with corresponding self-report ratings of bodily sensations taken to reflect 13 sampling points.
RESULTS: There were no group differences on any of the physiological measures during any of the four tasks. However, there were a number of between-group differences with regard to ratings of bodily sensations. Both clinical groups had higher ratings of racing heart than the non-anxious control group during the imagery task. In addition, social phobics had significantly higher ratings of racing heart during the social conversation in relation to both comparison groups. With regard to ratings of body heat, the anxious group had greater ratings than the non-anxious controls during the imagery task. Finally with regard to ratings of sweaty hands, both clinical groups had higher ratings than the non-anxious controls during the social conversation. All three groups were generally inaccurate in their ratings of bodily sensations.