Measurement of physical activity and sedentary behavior in breast cancer survivors

Background Breast cancer survivors' self-perceived physical activity (PA) and sitting time (ST) may differ significantly from the general population and other survivor groups, so it is important that PA and ST measurement tools are compared within the breast cancer survivor population. Objective To compare accelerometer and self-report estimates of PA and ST in breast cancer survivors. Methods 414 breast cancer survivors (age, 56.8 years [SD, 9.2 years]; BMI, 26.2 kg/m(2) [5.4 kg/m2]) wore an accelerometer for 7 consecutive days and completed a modified Godin Leisure-Time Exercise Questionnaire (GLTEQ), the International Physical Activity Questionnaire (IPAQ), and the Sitting Time Questionnaire (STQ) which all measured hours/minutes of activity/sitting per day. Mean differences and correlations of ST, light PA (LPA; <= 1.5 metabolic equivalents [METs]), and moderate and vigorous PA (MVPA; >= 3 METs) were compared using random-intercept mixed-effects regression models and the Spearman rank correlation coefficient (Spearman's rho [r(s)], where r(s) = 1 means a perfect positive correlation, and r(s) = -1 means a perfect negative correlation). Results Mean daily durations of MVPA were: accelerometer, 20.2 minutes; GLTEQ, 23.6 minutes (P-diff = .02); and IPAQ, 87.4 minutes (P-diff < .001). Correlations between accelerometer-estimated MVPA were moderate for the GLTEQ (r(s) = 0.56) and poor for the IPAQ (r(s) = 0.02). Mean daily durations of LPA were 239.5 minutes for the accelerometer and 15.4 minutes for the GLTEQ (P-diff < .001); the measures were not correlated (r(s) = 0.004). Mean daily durations of ST were: accelerometer, 603.9 minutes; STQ, 611.8 minutes (P-diff = 0.9); and IPAQ, 303.8 minutes (P-diff < 0.001). Correlations with the accelerometer were fair (STQ: r(s) = 0.26; IPAQ: r(s) = 0.30). Differences in estimates varied by disease stage, age, presence of chronic conditions, and race. Limitations Participants were predominantly white, highly educated, and high earners, which reduced generalizability. Conclusions Congruency of measurement was dependent on tool, intensity of activity, and participant characteristics. Target outcome, implementation context, and population should be considered when choosing a measurement for physical activity or sitting time in breast cancer survivors.