论文部分内容阅读
[背景]已证实住宅杀虫剂暴露与成人和儿童的不良健康结局相关。高质量的暴露评估是确认这些关联的关键。以往的流行病学研究采用一次测量地毯灰尘中杀虫剂浓度来描述个体长期暴露的平均值。如果浓度随着时间的推移而变化,那么这种方法会导致对暴露明显的错误分类,削弱对风险的估计。[目的]评估地毯灰尘样本中杀虫剂浓度的可重复性和流行病学研究中依赖于一个样本的潜在衰减偏倚。[方法]在2003—2005年期间,从加州Fresno县的21个家庭中,重复收集地毯灰尘样本(中位数=3,范围1~7)。使用气相色谱-质谱法分析灰尘中的13种杀虫剂;采用混合效应模型估计家庭之间和家庭内的方差。计算每一种杀虫剂的组内相关系数(ICC),并估计收集单次灰尘样品的假设病例对照研究衰减回归系数。[结果]ICC的中位数为0.73(范围0.37~0.95),表明大多数杀虫剂家庭之间的变异性高于家庭内部的变异性。对于13种化合物中进行评估的7种,与使用单一灰尘样本相关的预期衰减偏倚估计为≤30%。[结论]对于研究的几种杀虫剂,使用一个灰尘样本代表约为2年的暴露期,其比值比不会大幅衰减。还需要进一步研究,以确定上述研究结果是否可以代表更长的暴露时间并且适用于其他杀虫剂。
[Background] Residential pesticide exposure has been shown to be associated with adverse health outcomes in adults and children. High quality exposure assessment is the key to identifying these relationships. Previous epidemiological studies used a single measurement of pesticide concentrations in carpet dust to describe the average long-term exposure of individuals. If the concentration changes over time, then this approach can lead to a clear misclassification of exposure and diminish the estimation of risk. [Objective] To evaluate the repeatability of insecticide concentrations in carpet dust samples and the potential attenuation bias in a epidemiological study that relies on one sample. [Method] During 2003-2005, carpet dust samples were collected repeatedly from 21 households in Fresno County, California (median = 3, range 1 to 7). Thirteen insecticides in dust were analyzed using gas chromatography-mass spectrometry; and mixed-effects models were used to estimate variance within and among households. The intraclass correlation coefficient (ICC) was calculated for each pesticide and the hypothetical case-control study decaying regression coefficients were estimated for collecting single dust samples. [Results] The median ICC was 0.73 (range 0.37 to 0.95), indicating that variability among most pesticide families was higher than within the household. For seven of the 13 compounds evaluated, the expected attenuation bias associated with using a single dust sample was estimated to be ≤30%. [Conclusions] For several pesticides studied, the use of a dust sample represents an exposure period of about 2 years, with a ratio that does not significantly decrease. Further research is needed to determine if the above findings represent longer exposure times and apply to other pesticides.