Extreme temperature, humidity, air pollution and spontaneous preterm birth

项目摘要 该项目将扩展估算每日温度、湿度和细颗粒物空气的方法 通过利用美国宇航局和美国地质勘探局新的卫星测量， 在美国东北部的社区水平上产生估计的暴露。这种详细的曝光 记录可用于健康研究，以进一步考虑极端天气和空气污染物对健康的危害。 这些新的每日模型将在2007-2023年以<1公里的分辨率开发，因此包括数十亿个 点日温度、湿度和颗粒物估计值。虽然众所周知， 物质会增加早产的风险，但对极端天气事件（包括 温度和湿度）和空气污染有助于自发性早产的发生。 病理生理学在整个一生中建立&在怀孕期间，但早产和/或破裂的发生 是急性的。通过这种方式，我们调查了潜在的环境触发因素，并问：“为什么是今天？”的 这些新的暴露模型的流行病学应用将通过测试以下因素的关联来证明： 温度、湿度和颗粒物，以及自发性早产和住宅的数据集 地址从一个全面的纽约全州行政数据库。我们将使用地址级 自发性早产前7天内的暴露量与相同固定两个- 17年自发性早产的比较（时间分层病例交叉设计） 全州范围内的出生-使其成为自发性早产的最大环境流行病学研究之一 出生超过17万例。鉴于病例数量庞大，人口变化 整个纽约州的特点，极端天气，空气污染模式和 自发性早产将进一步测试性别、种族、胎龄 城市/农村和邻里贫困。随访分析将使用仔细表型分析的自发性 费城大型医院系统的早产病例。我们的流行病学模型将改进之前的 研究极端气温和空气污染对早产的严重影响， 通过生态时间序列模型进行暴露分配，病例交叉的时间分层过长 比较，或过于广泛的结果定义，包括医学指示的早产。鉴于 全球卫星遥感覆盖，我们的方法-可以产生每日暴露估计， 具有高度空间分辨率（<1 km），可捕获家庭地址处的环境暴露-广泛适用 为了更好地了解极端天气和空气污染作为慢性和急性应激源的作用， 使用大数据注册的健康结果。

PROJECT SUMMARY This project will extend methods for estimating daily temperature, humidity, and fine particulate matter air pollution at high resolution across large regions by utilizing NASA and new USGS satellite measurements to generate estimated exposures at the neighborhood level across the Northeast US. This detailed exposure record can be used in health studies to further consider extreme weather and air pollutant as a health risk. These new daily models will be developed for 2007-2023 at a <1 km resolution and thus include billions of point-day temperature, humidity, and particulate matter estimates. While it is well established that particulate matter increases the risk of preterm birth, less is known about how extreme weather events (including temperature and humidity) and air pollution contribute to the onset of spontaneous preterm birth. Pathophysiology builds throughout a lifetime & during pregnancy, yet the onset of preterm labor and/or rupture of membranes is acute. In this way, we investigate potential environmental triggers to ask, “Why today?”. The epidemiologic application of these novel exposure models will be demonstrated by testing the association of temperature, humidity, and particulate matter with a dataset of spontaneous preterm births and residential addresses from a comprehensive New York statewide administrative database. We will use address-level exposures in the 7 days prior to a spontaneous preterm delivery with matched days in the same fixed two- week period as a comparison (time stratified case-crossover design) for 17 years of spontaneous preterm births statewide – making this among the largest environmental epidemiology studies of spontaneous preterm birth with more than 170,000 cases. Given the large number of cases and the variation in population characteristics across New York State, the association between extreme weather, air pollution patterns and spontaneous preterm birth will be further tested for effect modification by sex, race, gestational age, urban/rural, and neighborhood deprivation. A follow-up analysis will use carefully phenotyped spontaneous preterm cases from large hospital systems in Philadelphia. Our epidemiologic models will improve on prior work on the acute impacts of extreme temperature and air pollution on preterm birth that have relied on coarse exposure assignments via ecological time series models, inappropriately long time strata for case-crossover comparison, or overly broad outcome definitions that included medically-indicated preterm births. Given the global coverage of satellite remote sensing, our approach – which can generate daily exposure estimates that are highly spatially resolved (<1 km) to capture ambient exposure at the home address – is broadly applicable to better understand the role of extreme weather and air pollution as a stressor in both chronic and acute health outcomes using big data registries.