H-22B Free Communication/Slide Cancer
The Colorectal Cancer and Home-Based Physical Exercise (CAN-HOPE) trial was a randomized controlled trial (RCT) examining the effects of a 16 week home-based exercise program on fitness and quality of life in colorectal cancer survivors. We previously reported problems with exercise adherence and contamination that negated the effects of the intervention when using intention-to-treat analysis.
This study examined predictors of exercise adherence and contamination in the trial.
We randomly assigned participants in a 2:1 ratio to either an exercise (n = 69) or control group (n = 33). At baseline, all participants completed measures of the theory of planned behavior, the five factor model of personality, past exercise, physical fitness, medical variables, and demographics. We monitored exercise over a 16 week period using weekly self-reports.
Data were available on 93 participants at follow-up (62 exercisers and 31 controls). Participants were 38 to 79 years old, 58% were male, 76% had colon cancer, 21% received radiation therapy, and 65% received chemotherapy. Adherence was good in the exercise group (76%) but contamination was a significant problem in the control group (52%). Hierarchical stepwise regression analyses indicated that the independent predictors of adherence in the exercise group were baseline exercise stage (ß = .35; p < .01), employment status (ß= −.28; p < .05), treatment protocol (ß= −.26; p < .05), and perceived control (ß = .20; p < .06). These variables explained 40% of the variance in exercise during the RCT. The independent predictors of exercise contamination in the control group were pre-existing intention (ß = .36; p < .05) and baseline exercise stage (ß = .30; p < .10), together explaining 30% of the variance in exercise during the RCT.
The predictors of exercise adherence and contamination in an RCT of colorectal cancer survivors are different. These findings may have important implications for maximizing exercise adherence and minimizing contamination in RCTs in this population.