Burch James B, Delage Alexandria F, Zhang Hongmei, McLain Alexander C, Ray Meredith A, Miller Austin, Adams Swann A, Hébert James R
Department of Epidemiology, School of Population Health, Virginia Commonwealth University, Richmond, VA, United States.
Department of Epidemiology and Biostatistics, Arnold School of Public Health, University of South Carolina, Columbia, SC, United States.
Front Oncol. 2024 Feb 26;14:1336487. doi: 10.3389/fonc.2024.1336487. eCollection 2024.
Sleep disruption affects biological processes that facilitate carcinogenesis. This retrospective cohort study used de-identified data from the Veterans Administration (VA) electronic medical record system to test the hypothesis that patients with diagnosed sleep disorders had an increased risk of prostate, breast, colorectal, or other cancers (1999-2010, N=663,869). This study builds upon existing evidence by examining whether patients with more severe or longer-duration diagnoses were at a greater risk of these cancers relative to those with a less severe or shorter duration sleep disorder.
Incident cancer cases were identified in the VA Tumor Registry and sleep disorders were defined by International Classification of Sleep Disorder codes. Analyses were performed using extended Cox regression with sleep disorder diagnosis as a time-varying covariate.
Sleep disorders were present among 56,055 eligible patients (8% of the study population); sleep apnea (46%) and insomnia (40%) were the most common diagnoses. There were 18,181 cancer diagnoses (41% prostate, 12% colorectal, 1% female breast, 46% other). The hazard ratio (HR) for a cancer diagnosis was 1.45 (95% confidence interval [CI]: 1.37, 1.54) among those with any sleep disorder, after adjustment for age, sex, state of residence, and marital status. Risks increased with increasing sleep disorder duration (short [<1-2 years] HR: 1.04 [CI: 1.03-1.06], medium [>2-5 years] 1.23 [1.16-1.32]; long [>5-12 years] 1.52 [1.34-1.73]). Risks also increased with increasing sleep disorder severity using cumulative sleep disorder treatments as a surrogate exposure; African Americans with more severe disorders had greater risks relative to those with fewer treatments and other race groups. Results among patients with only sleep apnea, insomnia, or another sleep disorder were similar to those for all sleep disorders combined.
The findings are consistent with other studies indicating that sleep disruption is a cancer risk factor. Optimal sleep and appropriate sleep disorder management are modifiable risk factors that may facilitate cancer prevention.
睡眠中断会影响促进癌症发生的生物过程。这项回顾性队列研究使用了来自退伍军人事务部（VA）电子病历系统的去识别化数据，以检验以下假设：被诊断患有睡眠障碍的患者患前列腺癌、乳腺癌、结直肠癌或其他癌症的风险增加（1999 - 2010年，N = 663,869）。本研究通过考察诊断更为严重或持续时间更长的患者相对于诊断较轻或持续时间较短的睡眠障碍患者，是否患这些癌症的风险更高，从而在现有证据的基础上进行拓展。
在VA肿瘤登记处识别出癌症发病病例，并根据国际睡眠障碍分类编码定义睡眠障碍。分析采用扩展的Cox回归，将睡眠障碍诊断作为一个随时间变化的协变量。
56,055名符合条件的患者存在睡眠障碍（占研究人群的8%）；睡眠呼吸暂停（46%）和失眠（40%）是最常见的诊断。共有18,181例癌症诊断（41%为前列腺癌，12%为结直肠癌，1%为女性乳腺癌，46%为其他癌症）。在调整年龄、性别、居住州和婚姻状况后，患有任何睡眠障碍的患者被诊断为癌症的风险比（HR）为1.45（95%置信区间[CI]：1.37, 1.54）。风险随着睡眠障碍持续时间的增加而增加（短[<1 - 2年] HR：1.04 [CI：1.03 - 1.06]，中[>2 - 5年] 1.23 [1.16 - 1.32]；长[>5 - 12年] 1.52 [1.34 - 1.73]）。使用累积睡眠障碍治疗作为替代暴露指标，风险也随着睡眠障碍严重程度的增加而增加；患有更严重障碍的非裔美国人相对于接受治疗较少的患者和其他种族群体风险更高。仅患有睡眠呼吸暂停、失眠或其他睡眠障碍的患者的结果与所有睡眠障碍合并患者的结果相似。
这些发现与其他表明睡眠中断是癌症风险因素的研究一致。最佳睡眠和适当的睡眠障碍管理是可改变的风险因素，可能有助于癌症预防。
