0493 Associations of Obstructive Sleep Apnea with Total and Site-specific Cancer Risk in Elderly Women: A Prospective Study

Introduction Chronic intermittent hypoxia resulting from obstructive sleep apnea (OSA) may activate multiple pathways relevant to carcinogenesis (e.g., HIF, MAPK and PI3K pathways) and lead to cancer development. However, several population-based studies reported mixed results on the associations between OSA and cancer risk, with inadequate control of important confounding factors and limited assessment of site-specific cancer risk. Methods We prospectively examined the association between OSA and cancer risk among 62,789 women participating in the Nurses' Health Study who were free of cancer in 2008 (mean age: 73.3 years). Incident cancer diagnoses were collected until 2014 and ascertained against pathology reports. Clinically-diagnosed OSA was self-reported in 2008 and updated in 2012. We used time-dependent Cox regression to estimate hazard ratios (HR) for the associations of OSA with total and site-specific cancer risk. Results During 6 years of follow-up, we documented 3,968 incident cancer diagnoses. In the age-adjusted model, OSA was associated with a 17% (95% CI: 1.02, 1.33) increase in total cancer risk. Adjustment for multiple cancer risk factors (e.g., family history of cancer, BMI, pack-years, alcohol drinking, physical activity, hormone therapy, etc.) attenuated the association between OSA and cancer risk (HR: 1.09; 95% CI: 0.95, 1.25). When examining cancer risk by site, OSA was associated with significantly increased risk for lung cancer (n=484; HR: 1.56; 95% CI: 1.09, 2.24) and possibly increased risk for kidney (n=83; HR: 1.70; 95% CI: 0.81, 3.58) and bladder cancer (n=137; HR: 1.76; 95% CI: 0.94, 3.29). There were no associations with other sites examined including breast, colon/rectum, uterus, ovary, pancreas, and melanoma. Conclusion We observed heterogeneous associations between OSA and site-specific cancer risk. The results need to be confirmed in other studies given the relatively small number of cases for bladder/kidney cancers and the focus on elderly women. Further investigation is required to understand the mechanisms underlying the associations with specific cancer sites, such as variations in the sensitivity of different tissue/cell types to hypoxia. Support (If Any) NIH grants UM1CA186107, P01CA87969, K01HL143034, R35HL135818