Sleep Disruption among Veterans: Implications for Cancer Risk

Veterans Health Administration, Decatur PA

Investigators

Abstract

The prevalence of insomnia and other sleep disorders has been increasing among veterans in the US. Poor sleep has been associated with fatigue, depression, accidents and injuries, suicide, post-traumatic stress disorder, increased disability and mortality rates, decreased quality of life, altered immune function, higher medical care costs, and the development of several major chronic diseases including heart disease, stroke, and diabetes. However, studies of the relationship between sleep disruption and cancer (the second leading cause of mortality in the US) are lacking. Our pilot study results indicate that sleep disorder prevalence among VISN-7 veterans more than doubled between 2000 and 2008. These trends may have important implications for the future development of chronic diseases among veterans. The VA maintains one of the largest electronic medical record systems in the world. The proposed study will utilize pre-existing, de-identified VA data for nearly three million diagnoses over a 10-year span to test the hypothesis that sleep disorders are associated with an increased incidence of breast, colorectal, prostate, or total cancer. The specific aims of this study are to: 1). Characterize trends in sleep disorder prevalence among VISN-7 veterans using descriptive epidemiology, and;2). Conduct a retrospective cohort study to determine whether sleep disorder diagnoses are associated with an increased incidence of breast, colorectal, or prostate cancer, or all cancer combined. The relationship between prior sleep disorder diagnosis and a subsequent cancer diagnosis will be determined via time-dependent Cox proportional hazards regression analyses with adjustment for age, gender, smoking, body mass index, contextual race and socioeconomic status, co-morbid disease, and the total number of concurrent diagnoses. Sleep disorder severity will be characterized by quantifying: a.) the number of times each sleep disorder code appears in the medical record;b.) duration of time since original diagnosis, and c.) the cumulative number of sleep disorder procedures and prescriptions for each patient. Dose-response will then be evaluated by computing hazard ratios among patients grouped into categories of increasing sleep disorder severity. The proposed retrospective cohort design is advantageous in terms of feasibility, cost, and ability to evaluate strength of association, temporal sequence, and dose-response. In follow-up studies, we plan to examine the relationship between sleep disorders and cancer survival. The long term goal of this research is to develop sleep disorder management strategies that reduce cancer morbidity and mortality. If our hypothesis is supported, a Merit Award proposal will be developed to identify sleep disorder therapies that may reduce cancer incidence and enhance survival by measuring inflammatory and interim cancer biomarkers among patients receiving different types of sleep disorder therapy.