目的:需要新的流行病学方法来减少超低频磁场(ELF-MF)与儿童白血病之间关联的科学不确定性。虽然以前的研究大多集中在电力线上,变压器暴露研究试图通过一项多国研究来评估这种关联,研究对象是居住在内置变压器的建筑物中的儿童.内置变压器以上公寓的ELF-MF可以比同一建筑物中的其他公寓高5倍。这项新颖的研究设计旨在最大程度地纳入高度暴露的儿童,同时最大程度地减少选择偏差的可能性。
方法:我们使用在5个国家收集的基于登记的匹配病例对照数据,评估了住宅靠近变压器与儿童白血病风险之间的关联。曝光是基于受试者公寓相对于变压器的位置,编码为高(高于或靠近变压器),中间(与高级公寓相同的楼层),或未暴露(其他公寓)。使用条件逻辑和混合逻辑回归估计儿童白血病的相对风险(RR),对病例对照集具有随机效应。
结果:各国的数据汇集产生了16例中度病例和3例高度暴露病例。在条件逻辑模型中,中度暴露的RRs为1.0(95%CI:0.5,1.9),高暴露的RRs为1.1(95%CI:0.3,3.8)。在混合逻辑模型中,中间的RRs为1.4(95%CI:0.8,2.5),高的RRs为1.3(95%CI:0.4,4.4)。最具影响力国家的数据显示,中度(8例)和高度(2例)暴露的RR分别为1.1(95%CI:0.5,2.4)和1.7(95%CI:0.4,7.2)。
结论:总体而言,风险升高的证据很弱。然而,小的数字和宽的置信区间排除了强有力的结论,不能排除在电力线研究中观察到的量级的风险.
OBJECTIVE: New epidemiologic approaches are needed to reduce the scientific uncertainty surrounding the association between extremely low frequency magnetic fields (ELF-MF) and childhood leukemia. While most previous studies focused on power lines, the Transformer Exposure study sought to assess this association using a multi-country study of children who had lived in buildings with built-in electrical transformers. ELF-MF in apartments above built-in transformers can be 5 times higher than in other apartments in the same building. This novel study design aimed to maximize the inclusion of highly exposed children while minimising the potential for selection bias.
METHODS: We assessed associations between residential proximity to transformers and risk of childhood leukemia using registry based matched case-control data collected in five countries. Exposure was based on the location of the subject\'s apartment relative to the transformer, coded as high (above or adjacent to transformer), intermediate (same floor as apartments in high category), or unexposed (other apartments). Relative risk (RR) for childhood leukemia was estimated using conditional logistic and mixed logistic regression with a random effect for case-control set.
RESULTS: Data pooling across countries yielded 16 intermediate and 3 highly exposed cases. RRs were 1.0 (95% CI: 0.5, 1.9) for intermediate and 1.1 (95% CI: 0.3, 3.8) for high exposure in the conditional logistic model. In the mixed logistic model, RRs were 1.4 (95% CI: 0.8, 2.5) for intermediate and 1.3 (95% CI: 0.4, 4.4) for high. Data of the most influential country showed RRs of 1.1 (95% CI: 0.5, 2.4) and 1.7 (95% CI: 0.4, 7.2) for intermediate (8 cases) and high (2 cases) exposure.
CONCLUSIONS: Overall, evidence for an elevated risk was weak. However, small numbers and wide confidence intervals preclude strong conclusions and a risk of the magnitude observed in power line studies cannot be excluded.