@ARTICLE{10.3389/fnhum.2021.653108, AUTHOR={Zika, Maya A. and Becker, Linda}, TITLE={Physical Activity as a Treatment for Social Anxiety in Clinical and Non-clinical Populations: A Systematic Review and Three Meta-Analyses for Different Study Designs}, JOURNAL={Frontiers in Human Neuroscience}, VOLUME={15}, YEAR={2021}, URL={https://www.frontiersin.org/articles/10.3389/fnhum.2021.653108}, DOI={10.3389/fnhum.2021.653108}, ISSN={1662-5161}, ABSTRACT={The fear of being in the focus of attention in social situations can develop into a social anxiety disorder (SAD). The classical treatment for SAD is cognitive behavioral therapy, which is in many cases accompanied by drug treatments. A promising alternative treatment is physical activity (PA) interventions, because regular PA has been shown to be suitable for reducing anxiety in general. We conducted a pre-registered systematic review and meta-analysis (PROSPERO registration no. CRD42020191181) as well as two additional searches. Our aim was to investigate whether PA interventions are a suitable treatment for SAD and whether PA is suitable for reducing social anxiety (SA) in general. For studies with randomized controlled trial designs, a not statistically significant effect of medium size toward lower general SA symptomatology was found in the PA group in comparison with the control group (d = −0.24, p = 0.377). For studies with longitudinal designs, significantly lower SA symptoms were found after PA treatments (d = −0.22, p = 0.001). The effect of PA on SA was stronger for adults than for children and adolescents (p = 0.003). For cross-sectional studies, a small negative association between SA symptoms and the amount of PA was found, i.e., lower SA was found for people who were more physically active (r = −0.12, p = 0.003). We conclude that PA is a promising means for the (additional) treatment of SAD or to reduce SA in general in non-clinical samples, but more research in which high-quality studies with randomized controlled trial designs are used is needed. Furthermore, open questions with respect to moderating variables (e.g., age, sex, BMI, type of intervention, stress, amount of regular PA before the intervention, and comorbidities) remain still open.} }