To explore associations between residential exposure to specific pesticides and birth outcomes using individual level exposure and pregnancy/birth data.
From all 2009-2013 singleton births in the Dutch birth registry, we selected mothers > 16 years old living in non-urban areas, who had complete address history and changed addresses at most once during pregnancy (N = 339,947). We estimated amount (kg) of 139 active ingredients (AI) used within buffers of 50, 100, 250 and 500 m around each mother\'s home during pregnancy. We used generalized linear models to investigate associations between 12 AIs with evidence of reproductive toxicity and gestational age (GA), birth weight (BW), perinatal mortality, child́s sex, prematurity, low birth weight (LBW), small for gestational age (SGA) and large for gestational age (LGA), adjusting for individual and area-level confounders. For the remainder 127 AIs, we used minimax concave penalty with a stability selection step to identify those that could be related to birth outcomes.
Regression analyses showed that maternal residential exposure to fluroxypyr-meptyl was associated with longer GA, glufosinate-ammonium with higher risk of LBW, linuron with higher BW and higher odds of LGA, thiacloprid with lower odds of perinatal mortality and vinclozolin with longer GA. Variable selection analysis revealed that picoxystrobin was associated with higher odds of LGA. We found no evidence of associations with other AIs. Sensitivity and additional analysis supported these results except for thiacloprid.
In this exploratory study, pregnant women residing near crops where fluroxypyr-meptyl, glufosinate-ammonium, linuron, vinclozolin and picoxystrobin were applied had higher risk for certain potentially adverse birth outcomes. Our findings provide leads for confirmatory investigations on these compounds and/or compounds with similar modes of action.
目的:使用个体水平暴露和妊娠/分娩数据,探讨住宅暴露于特定农药与分娩结局之间的关系。
方法:从荷兰出生登记处的所有2009-2013年单胎出生,我们选择了生活在非城市地区的16岁以上的母亲,谁有完整的地址历史,并且在怀孕期间最多更改一次地址(N=339,947)。我们估计了在怀孕期间每个母亲家周围50、100、250和500m的缓冲液中使用的139种活性成分(AI)的量(kg)。我们使用广义线性模型来研究12个AI与生殖毒性证据和胎龄(GA)之间的关联。出生体重(BW),围产期死亡率,儿童性,早产,低出生体重(LBW),胎龄小(SGA)和胎龄大(LGA),调整个人和地区层面的混杂因素。对于其余的127个AI,我们使用带有稳定性选择步骤的minimax凹惩罚来识别可能与分娩结局相关的因素.
结果:回归分析显示,母亲居住暴露于氟氧吡唑与较长时间的GA有关,草铵膦具有较高的LBW风险,具有较高BW和较高LGA几率的linuron,噻虫啉具有较低的围产期死亡率,而长春唑林具有较长的GA。变量选择分析显示,picoxstrobin与较高的LGA几率相关。我们没有发现与其他AI有关联的证据。除噻虫啉外,敏感性和其他分析支持这些结果。
结论:在这项探索性研究中,孕妇居住在fluroxypyr-meptyl附近的作物,草铵膦,linuron,应用长春唑林和picoxystrobin的某些潜在不良分娩结局的风险较高.我们的发现为对这些化合物和/或具有相似作用方式的化合物的确证研究提供了线索。