Physical Activity, Sedentary Behavior, and Cancer Incidence in Women

点击翻译按钮获取中文摘要

Abstract In the United States in 2016, an estimated 600,000 people will die from cancer and 1.7 million new cases of cancer will be diagnosed. Considering the life span, almost 40% of women and men will be diagnosed with cancer at some point in their lifetime. Understanding amenable risk factors that contribute to this large public health burden is essential. Observational studies consistently indicate associations between self-reported physical activity and increased risk of many types of cancer, especially breast cancer. However, there is insufficient evidence regarding the amount, intensity, duration, and types of physical activity required to reduce cancer risk, especially for older women. Sedentary behavior may provide a more feasible intervention target, especially for older adults. However, even less is known about sedentary behavior and cancer risk. To date, no prospective studies have examined accelerometry-derived physical activity and sedentary behavior to risk of incident cancer outcomes. Advances have been made in new measurement methods by our team, but have not been applied to health outcomes to assess their value. In this application, we propose to assemble accelerometry-assessed physical activity and sedentary behavior and cancer incident events and deaths from two cohort studies of women: the Women's Health Study (WHS) and the Women's Health Initiative (WHI) Study. We will apply sophisticated, yet directly interpretable, methods to determine which physical activity and sedentary behavior features are most important for reducing cancer risk among more than 22,000 women 63 to 101 years of age. For both cohorts, one-week of accelerometry data were collected in a similar manner during 2011-2014. Follow-up of both cohorts for adjudicated cancer outcomes is planned through 2020 and likely beyond. We propose three aims. Aim 1 will apply novel and translational measures of accelerometry- assessed physical activity and sedentary behavior using latent class analysis, an activity index, and machine learning algorithms of raw accelerometry data for use in Aims 2 and 3. We will then investigate the association of accelerometry-assessed physical activity (Aim 2) and sedentary behavior (Aim 3) to overall and site-specific (e.g., breast, uterine, ovarian) incident and fatal cancer. This cost-efficient study will investigate in detail whether and how patterns of frequency, duration, intensity, bouts, and type of physical activity and sedentary behavior predict cancer outcomes. Identification of new cancer-protective patterns of physical activity and sedentary behavior will provide much-needed evidence to inform physical activity and sedentary behavior guidelines for disease prevention, can be used in interventions to reduce risk, and could revolutionize the monitoring of human responses in physical activity and sedentary behavior interventions.