Predictors of Adherence to a Home-based Walking Intervention in Breast Cancer Survivors

Abstract

1992 PURPOSE: To describe adherence to a home-based walking intervention among breast cancer survivors using three methods and to evaluate predictors of intervention adherence. METHODS: In this ongoing intervention, participants received an inperson counseling session and up to five telephone counseling calls. The intervention focused on goal setting, reinforcement, social support, safety and self-monitoring (modeled after the Stanford program). Adherence (actual walking/walking goal) was calculated three ways. Adherence was calculated using walking frequency (d/wk) and walking duration (min/wk) based on walking logs. In a sub-sample (n = 6), adherence was calculated using walking frequency (at least 20 min/d) from the MTI Actigraph. Predictors of adherence were evaluated using Spearman correlations (r) and ANOVA, and included: personal (demographic, body weight, physical activity); intervention related (number and length of calls); clinical (time since diagnosis, stage, treatment type); and quality of life (SF-36) variables. RESULTS: To date, 16 of 17 women randomized to intervention have been retained and have provided adherence data. Adherence for each month and overall for the frequency method was 101%, 79%, 66%, and 89%; and for the duration method was 101%, 76%, 63%, and 88%. Actigraph adherence was 83% in month two and 79% in month three. We identified potential predictors of adherence that will be evaluated closely as additional women complete the study. Adherence was somewhat lower for women with less education (< 4 years college: 67%, > 4 years college: 108%, p = 0.11) and higher body weight (r = −0.48, p = 0.06). Lower mental health scores were associated with higher adherence (r = − 0.52, p = 0.04). Differences in adherence were not noted for other variables examined, although women with later stage disease or more aggressive treatment tended to report better adherence to the intervention. CONCLUSIONS: Adherence calculated using the frequency and duration methods provided similar results. Objective measures of adherence were consistent with self-report, supporting the use of walking logs to estimate adherence. Walking adherence was associated with education, body weight, and baseline mental health scores. These data will be used to strengthen and individualize walking interventions among breast cancer survivors. Funding: South Carolina Cancer Center; Vanderbilt-Ingram Cancer Center