Examination of the Recommendation that Anxiety Measured in Association with Exercise Should Focus upon Cognitive Anxiety

1123 In their seminal article, Ekkekakis, Hall, and Petruzzello (1999) concluded that “Researchers now venturing into studies of the exercise-affect relationship…. should devise experimental protocols and measurement approaches that specifically target the cognitive antecedents of anxiety.” Furthermore, they urged doubtful researchers to reexamine their own data to verify for themselves the need to separate activation anxiety items from cognitive anxiety items. PURPOSE: The purpose of this study was to reexamine two data sets to determine if differential results are obtained when looking at total anxiety scores, somatic anxiety scores, and cognitive anxiety scores. Consistent with Ekkekakis et al., it was hypothesized that exercise groups in contrast to a control group, or higher intensity exercise in contrast to moderate intensity exercise would yield higher somatic anxiety and lower cognitive anxiety scores following anxiety. METHOD: Focusing only upon the independent variables of time and intensity, selected somatic and cognitive anxiety scores were analyzed. Four activation items (calm, relaxed, steady, at ease) and three cognitive items (worrying, worried, nervous) from Spielberger's 10-item SAI served as dependent variables. RESULTS: The original total anxiety score results from Cox, Thomas and Davis (2000) revealed a significant time effect with no interaction between time and intensity (no control group). In the reanalysis, a significant time effect with no interaction between time and intensity was again observed for both somatic and cognitive anxiety; with no hypothesized differentiation between exercise intensity groups for either somatic or cognitive anxiety. The original total anxiety score results from Cox, Thomas, Hinton and Donahue (in press) revealed a significant time effect and a significant interaction between time and intensity effect. In the reanalysis, it was observed that the activation anxiety of the exercise groups was elevated above the control group immediately following exercise; while in the case of cognitive anxiety, no differences were observed between exercise groups immediately following exercise. CONCLUSIONS: Very little support was found for the Ekkekakis et al. position that researchers should focus only upon cognitive antecedents of anxiety in exercise related investigations. It does not appear that the positive effects of exercise on cognitive anxiety are cancelled out by infiated somatic anxiety. On the contrary, it appears that total unidimensional anxiety scores provide the most appealing and consistent explanation of the effect of exercise on state anxiety.