首页> 外文会议>Joint annual meeting of the International Society of Exposure Science and the International Society for Environmental Epidemiology >Correlation patterns between multiple chemicals measured through human biomonitoring to describe changes over time
【24h】

Correlation patterns between multiple chemicals measured through human biomonitoring to describe changes over time

机译:通过人生物监唱测量的多种化学品之间的相关模式来描述随时间的变化

获取原文

摘要

While mixture toxicology and mixture risk assessment have been on the research and policy agenda for decades, assessment of the actual exposure of the population to mixtures of chemicals has received rather little attention. Within the HBM4EU project, individual human biomonitoring (HBM) data of multiple chemicals was analysed to better understand the patterns in correlation and network structures over time. These analyses are informative in understanding the common exposure patterns over time to combinations of chemicals. Three Flemish birth cohorts (FLEHSI-II-III) were included; recruited in ~2003, 2008 and 2013 from the general population. Heavy metals (Cd, Pb), p,p'-DDE, HCB and PCBs (PCB138, PCB153, PCB180) were measured in cord blood at all three points in time. Additionally, in 2008 and 2013 the heavy metals (As, Cu, Mn, Tl, MeHg (in hair)) and PFASs (PFOA, PFOS) were measured. Besides the Flemish data, preliminary comparative results of the cross-sectional population-representative German Environmental Survey for Children and Adolescents (GerES V) will be presented. At first, correlation structures within and across the datasets were assessed visually with Circos plots. Next, partial correlations were calculated to measure strength and direction between two chemicals whilst controlling for the effect of the other chemicals in the data. Based on the partial correlations, the best undirected graph networks were estimated using sparsity parameters. 'Undirected graphs' is an approach to describe the conditional independence among many variables. Within the networks, groups of closely related chemicals were identified (communities), which can provide insight into the existence of groupings. Changes over time were assessed with differentia) network analysis, assessing the similarities and differences between network structures and the communities. Here, differences are changes in correlation strength and direction between chemicals.
机译:虽然混合物毒理学和混合物风险评估几十年来研究和政策议程,但对化学品混合的人口实际暴露的评估已经受到相当一点的关注。在HBM4EU项目中,分析了多种化学品的个人生物监测(HBM)数据,以更好地了解相关性和网络结构的模式随着时间的推移。这些分析在理解普通曝光模式的情况下,随着时间的推移,以时间与化学品的组合的方式理解。包括三个佛兰氏佛佛兰氏佛兰氏佛兰氏岛(Flehsi-II-III);从一般人口招募〜2003年,2008年和2013年。重金属(CD,PB),P,P'-DDE,HCB和PCB(PCB138,PCB153,PCB180)在脐带血上以三个时间点测量。此外,在2008和2013年,测量了重金属(如,Cu,Mn,T1,Mehg(毛发))和PFAS(PFOA,PFOS)。除佛兰芒语言外,还将提出初步人口德国儿童和青少年环境调查的初步比较结果(GERES v)。首先,使用电池图,在视觉上评估数据集内部和跨越数据集的相关结构。接下来,计算部分相关性以测量两种化学品之间的强度和方向,同时控制数据中其他化学品的效果。基于部分相关性,使用稀疏参数估计最佳无向图网络网络。 “无向图”是描述许多变量之间有条件独立性的方法。在网络中,确定了密切相关的化学品组(社区),可以提供对分组存在的洞察。随着时间的推移而改变,评估了不同的网络分析,评估了网络结构与社区之间的相似性和差异。这里,差异是化学物质之间的相关强度和方向的变化。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号