Zhang Hui, García Rodríguez Luis A, Hernández-Díaz Sonia
Department Epidemiology, Harvard School of Public Health, 677 Huntington Avenue, Boston, MA 02115, USA.
Cancer Epidemiol Biomarkers Prev. 2008 Jun;17(6):1308-15. doi: 10.1158/1055-9965.EPI-07-2817. Epub 2008 Jun 10.
Antibiotic use has been associated with an increased risk of cancer in epidemiologic studies. We evaluated the association between antibiotic use and the risk of primary lung cancer by conducting a prospective case-control study nested in a cohort of subjects who were 40 to 84 years old in 1995 to 2004, with at least 2 years of enrollment in The Health Improvement Network. There were 4,336 cases of primary lung cancer. A random sample of 10,000 controls was frequency matched to the cases for age, sex, and calendar year of diagnosis. Antibiotic exposure was measured by the total number of antibiotic prescriptions and the cumulative number of days on antibiotics since enrollment. We discounted exposure 1 year before the date of cancer diagnosis. Relative risks (RR) and 95% confidence intervals (95% CI) were estimated using conditional logistic regression. Smoking, chronic obstructive pulmonary disease, respiratory infections, body mass index, and health care utilization were introduced in the model. Compared with subjects with no prescription of antibiotics before the index date, the crude RR of lung cancer was 2.52 (95% CI, 2.25-2.83) among those who received 10 or more prescriptions. The corresponding RR was 1.31 (95% CI, 1.10-1.57) upon adjustment for confounders. We used directed acyclic graphs to illustrate that the observed higher risk of lung cancer among antibiotic users may be due to the increased frequency of infections in patients with subclinical cancer and to shared causes between cancer and infections. Current evidence is insufficient to support or refute a carcinogenic effect of antibiotics.
在流行病学研究中，抗生素的使用与患癌风险增加有关。我们通过开展一项前瞻性病例对照研究，评估了抗生素使用与原发性肺癌风险之间的关联。该研究嵌套于一个队列中，队列中的受试者在1995年至2004年期间年龄为40至84岁，且至少在健康改善网络登记了2年。共有4336例原发性肺癌病例。随机抽取10000名对照，按年龄、性别和诊断的日历年与病例进行频数匹配。抗生素暴露通过自登记以来的抗生素处方总数和使用抗生素的累计天数来衡量。我们排除了癌症诊断日期前1年的暴露情况。使用条件逻辑回归估计相对风险（RR）和95%置信区间（95%CI）。模型中纳入了吸烟、慢性阻塞性肺疾病、呼吸道感染、体重指数和医疗保健利用情况。与索引日期前未开具抗生素处方的受试者相比，接受10次或更多处方的受试者患肺癌的粗RR为2.52（95%CI，2.25 - 2.83）。调整混杂因素后，相应的RR为1.31（95%CI，1.10 - 1.57）。我们使用有向无环图来说明，抗生素使用者中观察到的较高肺癌风险可能是由于亚临床癌症患者感染频率增加以及癌症和感染之间的共同病因。目前的证据不足以支持或反驳抗生素的致癌作用。
