Author(s): CONNOR, M., HARRISON, O., HARGREAVES, E., Institution: UNIVERSITY OF OTAGO, Country: NEW ZEALAND, Abstract-ID: 2597

Introduction
Research has shown that exercise can effectively reduce anxiety, however the mechanisms underlying this effect are ambiguous. Exercise may provide a setting for individuals to habituate to and reappraise physical sensations that are common to both anxiety and exercise, such as elevated heart rate and breathing. As such, anxiety sensitivity (AS; fear of anxiety symptoms) and subjective interoceptive attention (SIA; one’s ability to mindfully alter attention towards body symptoms) have been identified as potential mediators of this anxiolytic effect. This study compares the effects of an 8-week self-selected intensity aerobic exercise intervention with a stretching-based condition on anxiety, AS and SIA. It also investigates the mediatory effect of AS and SIA on anxiety to determine whether exposure to symptoms common across exercise and anxiety and/or how these symptoms are perceived and appraised underpin exercise-based anxiety reductions .
Methods
Inactive adults (n=64, mean age±SD=28.2±8.7) with high trait anxiety (≥40 on the Spielberger Trait Anxiety Index; STAI) were randomised to participate in aerobic- or static stretching-based exercise of a self-selected intensity, 3 times a week for 8 weeks. Trait anxiety (STAI-T), AS (Anxiety Sensitivity Index; ASI-3), and SIA (Multidimensional Assessment of Interoceptive Awareness; MAIA) were measured pre- and post-intervention. Exercising heart rate (%HRpeak) was recorded during one exercise session weekly.
Results
Linear mixed effects models showed that trait anxiety and AS significantly decreased from pre- to post-intervention (STAI-T: p<0.001, d=0.54; Total ASI-3: p=0.001, d=0.40), with no difference between conditions. For the MAIA, significant pre- to post-intervention improvements were found for the Not Worrying (p=0.005, d=-0.34), Self-Regulation (p=0.005, d=-0.41), Body Listening (p=0.012, d=-0.28) and Trusting (p=0.01, d=-0.34) subscales, with no differences between conditions. The aerobic group selected to exercise at a moderate intensity (65.8-71.4%HRpeak) which was significantly higher than the stretching group (45.7-49.4%HRpeak, p<0.001).
Structural equation modelling of the total ASI score and those from the significant MAIA subscales showed that change in ASI partially mediated reductions in anxiety for the aerobic group only (bindirect=-0.25, p=0.04, R2=0.73). Conversely, changes to trait anxiety in the stretching group were associated with (although not mediated by) changes to the analysed MAIA subscales.
Conclusion
Both self-selected intensity static stretching and aerobic exercise improved anxiety, with AS mediating changes in the aerobic group, and SIA influencing changes in the stretching group. These results suggest that while both modalities are effective in reducing anxiety, the mechanisms to explain the exercise effect differ.