Obstructive sleep apnoea and the incidence and mortality of cancer: a meta-analysis

As there are conflicting reports regarding the association between obstructive sleep apnoea (OSA) and cancer incidence and mortality, a meta-analysis was performed to evaluate whether OSA is independently associated with cancer incidence and mortality. Pubmed, EMBASE and Web of Science were searched up until November 2014. Studies that assessed OSA and the future risk of cancer incidence or mortality were included. Pooled hazard ratios (HR) and corresponding 95% confidence intervals (CI) were calculated. Subgroup analysis was conducted based on the polysomnographic variable, apnoea-hypopnoea index. Six studies, which involved 114 105 participants, were pooled in this meta-analysis. Fixed-effects analysis showed the pooled adjusted HR of cancer incidence as 0.91 (95% CI, 0.74-1.13; P = 0.408) for mild OSA, 1.07 (95% CI, 0.86-1.33; P = 0.552) for moderate OSA and 1.03 (95% CI, 0.85-1.26; P = 0.743) for severe OSA. Random-effects analysis demonstrated neither mild OSA (adjusted HR, 0.79; 95% CI, 0.46-1.34; P = 0.381), moderate OSA (adjusted HR, 1.92; 95% CI, 0.63-5.88; P = 0.251) nor severe OSA (adjusted HR, 2.09; 95% CI, 0.45-9.81; P = 0.349) correlated with cancer mortality. This meta-analysis indicates that OSA is not independently associated with cancer incidence and mortality according to currently available data. Additional experimental and human research is required to determine the exact association between OSA and cancer.