Abstract Physical activity guidelines state that adults should engage in at least 150 minutes of moderate or 75 minutes of vigorous intensity activity per week and limit time spent sitting. Guidelines are based largely on self-reported data of usual amounts of physical activity and sitting time, as self-report is a feasible and cost-effective method for assessment in large epidemiologic studies. However, self-reported survey measures largely focus on moderate-to-vigorous intensity physical activity (MVPA) and total sitting time, and they are subject to various sources of error. The American Cancer Society’s Cancer Prevention Study-3 (CPS-3) is a large, nationwide, prospective cohort of approximately 304,000 participants recruited from 2006-2013. Among 751 randomly selected CPS-3 participants, seven-day accelerometer data were collected at two-time points approximately 6 months apart in 2015/2016 to assess convergent validity of the CPS-3 physical activity survey data. The average number of wear days was excellent (6.6). Vigorous activity (r=0.75) was reported more accurately than light activity (r=0.31); sedentary time also showed reasonable validity (r=0.63). While CPS-3 and other cohort surveys have acceptable validity and are suitable for categorizing participants according to overall activity level, they do not allow for detailed examination of bouts, patterns, or intensity of different activities, especially in terms of measuring light-intensity activity where most active time is spent. Thus, some large cohorts have collected or are collecting accelerometer data on subpopulations to use for calibrating self-reported measures and for examining objective physical activity in relation to health outcomes. We are initiating collection of accelerometer data from a subcohort of at least 20,000 participants. We aim to oversample for men and racial/ethnic minorities as most previous accelerometry studies included predominantly white women. Participants will be sent an accelerometer (ActiGraph GT3x) with a brief survey/wear-time log and asked to wear the device for a maximum of seven days during all waking hours. Through a series of prediction models, the CPS-3 survey will be trained against a smaller subset of accelerometer data to improve the interpretability, accuracy, and precision of CPS-3 sedentary time and PA survey data. Associations of accelerometer-based physical activity with various outcomes will also be conducted. Through the quantifiable assessment of activity at different intensities, bout lengths, and patterns, we will be well positioned to understand dose-response of physical activity and sedentary time in relation to health and better inform public health guidelines. CPS-3 will not only add to the growing number of cohorts with device-based measures but also can uniquely contribute to gaps in the research with its younger age range, inclusion of men and women, and geographic as well as racial/ethnic diversity. This abstract is also being presented as Poster A33. Citation Format: Alpa V. Patel, Erika Rees-Punia, Susan M. Gapstur. Device-based, objective measurement of physical activity and sedentary behavior in a large U.S. cohort [abstract]. In: Proceedings of the AACR Special Conference on Modernizing Population Sciences in the Digital Age; 2019 Feb 19-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(9 Suppl):Abstract nr PR07.
Read full abstract