Group versus individual exercise interventions for women with breast cancer: a meta-analysis

Abstract

Background. Both during and after treatment, cancer survivors experience declines in physical and psychosocial quality of life (QoL). Prior research indicates that exercise interventions alleviate problems in physical functioning and some aspects of psychological functioning. For survivors seeking social support, exercise programmes that are conducted in group settings may foster optimal QoL improvement (by addressing additional issues related to isolation and social support) over individually based exercise programmes. Methods. We reviewed literature on group cohesion in exercise studies, and conducted a meta-analysis to test the hypothesis that group as compared to individual exercise interventions for breast cancer survivors would show greater improvement in QoL. Results. As currently implemented, group exercise interventions showed no advantage. However, they typically did not provide any evidence that they capitalised upon potentially beneficial group processes. Conclusions. Future exercise intervention studies could investigate the effect on QoL of deliberately using group dynamics processes, such as team-building experiences and group goal setting to foster group cohesion.

Keywords:

• cancer
• oncology
• exercise
• group dynamics
• quality of life
• meta-analysis

Acknowledgements

The authors would like to extend their sincere thanks to the authors of the studies included in the meta-analysis who provided data upon request (Dr. Kerry Courneya, Dr. Melinda Irwin, Dr. Alejandro Lucia, Dr. Don McKenzie and Dr. Bernardine Pinto), to Dr. Barbara Resnick and Dr. Paul Estabrooks for information and suggestions, to Dr. Hans-Joachim Huff and Anna Kuhlen for assistance coding studies written in German and to the members of the Vermont Dragonheart Dragonboat team for inspiring this project. This project was partially supported by a grant from the National Cancer Institute (R01 CA100810) to Anne Moyer.

Notes

1. The total PEDro score as used here is eight, not 10 as the scale is usually used; two of the items were not applicable to behavioural interventions such as physical activity interventions (these were the items assessing the blinding of participants and the blinding of therapists conducting the interventions). Previous meta-analyses have used a score of six out of 10 to distinguish lower quality from higher quality studies (e.g., Orr, Raymond, & Singh, 2008).

2. Kolden et al. (2002) reported three different outcomes related to SF QoL. Two of these showed no significant change over the course of the intervention (LFS, social functioning; FACT, social well-being); one showed significant improvement (CARES, psychosocial, at p<0.05). Because of the number of analyses they ran (presenting an increased possibility of a Type I error) and because two of the SF subscale outcomes were not significant, we chose to categorise this study that reported SF QoL, but did not report an improvement.

3. The other three studies that implied interest in social interaction did not report SF QoL outcomes separately.