Taking the next step: a systematic review and meta-analysis of physical activity and behavior change interventions in recent post-treatment breast cancer survivors

Abstract

Research has shown that recent post-treatment breast cancer survivors face significant challenges around physical activity as they transition to recovery. This review examined randomized controlled trials targeting physical activity behavior change in breast cancer survivors <5 years post-treatment and described (1) characteristics of interventions for breast cancer survivors as well as (2) effect size estimates for these studies. A systematic search was conducted following PRISMA guidelines with Medline, PubMed, PsycINFO, CINAHL, and Scopus databases. Data were abstracted for primary intervention strategies and other details (e.g., setting, duration, theory use). A subgroup analysis was conducted to assess intensity of exercise supervision/monitoring and intervention effectiveness. The search produced 14 unique behavior intervention trials from the US and abroad published 2005-2013. The mean sample size was 153 participants per study. All interventions included moderate-intensity activities plus various behavioral change strategies. Most interventions were partially or entirely home based. The overall standardized mean difference was 0.47 (0.23, 0.67) with p < 0.001. Most interventions were effective in producing short-term behavior changes in physical activity, but varied greatly relative to intervention strategies and intensity of supervision/monitoring. Highly structured interventions tended to produce larger behavior change effects overall, but many larger effect sizes came from interventions supported by phone counseling or e-mail. We observed that 'more' may not be better in terms of direct supervision/monitoring in physical activity behavior interventions. This may be important in exploring less resource-intensive options for effective behavior change strategies for recent post-treatment survivors.

Figure 1. Search Terms Used with Medline Database

Figure 1. Search Terms Used with Medline Database

Figure 2. PRISMA-based Study Selection Flowchart

Figure 3. Forest plot for effects of physical activity interventions on behavior change

Studies (n=14) measured changes in physical activity behavior comparing results from control to treatment groups. The size of the shaded boxes around each study represents the relative weight assigned to the study in calculating an overall estimate of standardized mean difference (SMD), which in this case was calculated using Hedge's g to reduce positive bias. Larger boxes indicate that greater weight was assigned to each study. Lines going through each box represent the confidence interval for each estimate.