National Institute for Public Health and the Environment, Bilthoven, the Netherlands.
Institute for Risk Assessment Sciences, Utrecht University, Utrecht, the Netherlands.
PLoS One. 2019 Oct 14;14(10):e0223601. doi: 10.1371/journal.pone.0223601. eCollection 2019.
In the Netherlands, an association was found between the prevalence of pneumonia and living near goat and poultry farms in 2007-2013. This association then led to regulatory decisions to restrict the building of new goat farms and to reduce emissions of poultry farms. Confirmation of these results, however, is required because the period of previous analyses overlapped a Q-fever epidemic in 2007-2010. To confirm the association, we performed a population-based study during 2014-2016 based on general practitioner (GP) data. Electronic medical records of 90,183 persons were used to analyze the association between pneumonia and the population living in the proximity (within 500-2000 m distance) of goat and poultry farms. Data were analyzed with three types of logistic regression (with and without GP practice as a random intercept and with stratified analyses per GP practice) and a kernel model to discern the influence of different statistical methods on the outcomes. In all regression analyses involving adults, a statistically significant association between pneumonia and residence within 500 meters of goat farms was found (odds ratio [OR] range over all analyses types: 1.33-1.60), with a decreasing OR for increasing distances. In kernel analyses (including all ages), a population-attributable risk between 6.0 and 7.8% was found for a distance of 2000 meters in 2014-2016. The associations were consistent across all years and robust for mutual adjustment for proximity to other animals and for several other sensitivity analyses. However, associations with proximity to poultry farms are not supported by the present study. As the causes of the elevated pneumonia incidence in persons living close to goat farms remain unknown, further research into potential mechanisms is required for adequate prevention.
在 2007-2013 年期间，荷兰发现肺炎的流行与居住在山羊和家禽养殖场附近有关。这一关联促使监管部门做出决定，限制新建山羊养殖场，并减少家禽养殖场的排放。然而，需要对这些结果进行证实，因为之前分析的时间段与 2007-2010 年的 Q 热疫情重叠。为了证实这一关联，我们在 2014-2016 年期间基于全科医生（GP）的数据进行了一项基于人群的研究。我们使用 90183 人的电子病历来分析肺炎与居住在山羊和家禽养殖场附近（距离 500-2000 米范围内）人群之间的关联。数据分析采用了三种类型的逻辑回归（包括和不包括 GP 实践作为随机截距，以及按每个 GP 实践进行分层分析）和核模型，以辨别不同统计方法对结果的影响。在所有涉及成年人的回归分析中，均发现肺炎与居住在距离山羊养殖场 500 米以内之间存在统计学显著关联（所有分析类型的比值比 [OR]范围：1.33-1.60），随着距离的增加 OR 逐渐降低。在核分析（包括所有年龄）中，在 2014-2016 年期间，对于 2000 米的距离，发现人群归因风险为 6.0%-7.8%。这些关联在所有年份都是一致的，并且对于其他动物的接近度以及其他几种敏感性分析的相互调整也是稳健的。然而，本研究并未支持与家禽养殖场接近度相关的关联。由于居住在山羊养殖场附近的人肺炎发病率升高的原因尚不清楚，因此需要进一步研究潜在的机制，以便进行充分的预防。
