Long-Term Lung Health After Exposure to Chlorine Gas

说明(申请人提供)：氯气是一种潜在的大规模杀伤性武器，是最常用的化学品之一[1-5]。氯化事故很常见，每年有数千人集体暴露[1，2，5-16]。一次重大的氯气泄漏事故可能导致公共健康灾难[6，17]。对肺部的急性健康影响有很好的文献记载[1，7，9，18-22]，但还没有对暴露在单一高浓度氯气事件中的人群进行大规模、系统的长期纵向队列研究。新的科学已经为潜在的长期健康影响提供了证据[24]，并呼吁对刺激性气体暴露者进行至少5年的长期纵向研究[27]。2005年1月6日，南卡罗来纳州格兰尼特维尔棉纺厂发生火车脱轨事故，随后发生54,422公斤氯气泄漏[28-33]。数百人立即患病，数千人可能接触到病毒，其中包括磨坊工人。从1980年到2006年，磨坊工人接受了肺健康监测，看是否患有酒精性肺病。我们有8234名工作人员，他们至少有3年的活动前肺活量测量数据。事故后重返工作岗位的磨工在2005年和2006年都出现了肺功能丧失(FEV1)，这一比例是事故发生前几年观察到的两倍。我们将使用嵌套在前瞻性纵向队列研究中的双盲中断时间序列设计，通过3年的新肺功能测试与3年自然实验中的事件前数据对比，来评估不同的事件前/事件后肺功能下降。(H3)年龄、性别、种族、吸烟、创伤后应激障碍(PTSD)、特应性、肺刺激性暴露职业史、既往存在的肺部疾病和肥胖修改了接触氯气与肺功能下降的关系(H4)肺功能下降与纤维化(限制性肺缺陷)和呼吸道炎症(阻塞性肺缺陷)的增加有关。假设年消耗率为5%，我们将招募6.7亿名工人来研究600人，其中包括来自3个经过验证的羽流模型暴露估计组(室外氯气持续30分钟浓度峰值分别为0.5ppm、2.8ppm和28ppm)的150名工人和150名未暴露的工人。我们将在6-10年后测量肺功能、呼吸道炎症、呼吸道纤维化指标、特应性、创伤后应激障碍和呼吸道反应性。这项研究意义重大，因为这是有史以来暴露在氯气事件中的最大普通美国人。我们的调查人员是强大和有能力的，他们成功地进行了公共健康筛查，验证了羽流模型，并有处理其他灾难的经验。这项研究具有独特的科学影响力，这是因为创新和稳健的研究设计，使个人暴露组双盲，每个参与者作为自己的对照，并混淆暴露前的数据。

DESCRIPTION (provided by applicant): Chlorine gas is a potential weapon of mass destruction and is one of the most commonly used chemicals [1-5]. Chlorination accidents are common, exposing thousands of people collectively each year [1, 2, 5-16]. A large accident releasing chlorine gas could result in a public health disaster [6, 17]. The acute health effects on the lungs are well documented [1, 7, 9, 18-22], but there has been no large systematic long-term longitudinal cohort study of a population exposed to a single high-concentration chlorine gas event. New science has provided evidence for potential long-term health effects [24], and called for long-term longitudinal study of irritant gas exposed persons for at least 5 years [27]. On January 6, 2005 a train derailment and subsequent 54,422 kg chlorine gas spill occurred in the cotton mill town of Graniteville, South Carolina [28-33]. Several hundred people became immediately sick and thousands were potentially exposed, including millworkers. Millworkers had their lung health monitored for byssinosis from 1980 to 2006. We have a cohort of 8,234 millworkers with at least 3 years of pre-event spirometry data. Millworkers who returned to work after the accident had lung function loss (FEV1) in both 2005 and 2006 that was twice the rate observed in the years immediately before the accident. We will assess differential pre/post-event lung function decline in a sample of millworkers with 3 years of new lung function tests contrasted with data from 3 years pre-event in a natural experiment, using a double-blinded interrupted time-series design nested within a prospective longitudinal cohort study. Hypothesis are: (H1) Lung function decline increases with increasing chlorine gas exposure (H2) Airway responsiveness increases with increasing chlorine gas exposure (H3) Age, sex, race, smoking, post traumatic stress disorder (PTSD), atopy, occupational history of lung-irritant exposures, pre-existing lung disease, and obesity modify the association of exposure to chlorine gas with lung function decline (H4) Increased lung function decline is associated with increased fibrosis (restrictive lung defects) and airway inflammation (obstructive lung defects). Assuming a 5 percent annual attrition, we will enroll 670 millworkers to study 600, including 150 millworkers from each of the 3 validated plume model exposure estimate groups (peak sustained 30-minute concentrations of 0.5 ppm, 2.8 ppm, and 28 ppm chlorine gas outdoors) and 150 unexposed millworkers. We will measure lung function, airway inflammation, indicators of airway fibrosis, atopy, PTSD, and airway responsiveness at 6-10 years post- event. This study is significant because it is the largest general US population ever exposed to a chlorine gas event. Our investigators are strong and capable, having succeeded with public health screening, validated plume modeling, and experience with other disasters. This study has the unique opportunity for strong scientific influence due to the innovative and robust study design, having personal exposure group double blinded, each participant serving as their own control, and confounder data available from before the exposure.