Wong O, Trent L, Harris F
Tulane University Medical Center, New Orleans, Louisiana, USA.
Occup Environ Med. 1999 Apr;56(4):217-21. doi: 10.1136/oem.56.4.217.
This nested case-control study was based on data in a cohort study of more than 18,000 petroleum distribution workers exposed to gasoline, which contains about 2%-3% benzene. Risks of leukaemia, acute myeloid leukaemia, multiple myeloma, and kidney cancer were examined relative to exposure to gasoline.
For each case, up to five individually matched controls were selected. Analyses based on the Mantel-Haenszel procedure as well as univariate and multivariate conditional logistic regression were performed for each disease category. Jobs with similar exposures were grouped into homogeneous categories for analysis. Several quantitative indices of exposure to gasoline were used in the analyses: duration of exposure, cumulative exposure, frequency of peak exposure, and time of first exposure.
No increased risks for the four cancers were found for any job category. Analyses with logistic regression models based on duration of exposure, cumulative exposure, and frequency of peak exposure did not show any increased risk or exposure-effect relation. Time of first exposure to gasoline was also found to be unrelated to the four diseases under investigation.
Exposure to gasoline or benzene at the concentrations experienced by this cohort of distribution workers is not a risk factor for leukaemia (all cell types), acute myeloid leukaemia, multiple myeloma, or kidney cancer.
这项巢式病例对照研究基于一项对18000多名接触汽油（含约2%-3%苯）的石油分销工人的队列研究数据。研究了相对于接触汽油而言，患白血病、急性髓细胞白血病、多发性骨髓瘤和肾癌的风险。
对于每例病例，选取多达5名个体匹配的对照。对每个疾病类别进行基于Mantel-Haenszel程序以及单变量和多变量条件逻辑回归的分析。将具有相似接触情况的工作归为同类进行分析。分析中使用了几种汽油接触的定量指标：接触持续时间、累积接触量、峰值接触频率和首次接触时间。
在任何工作类别中均未发现这四种癌症的风险增加。基于接触持续时间、累积接触量和峰值接触频率的逻辑回归模型分析未显示任何风险增加或接触-效应关系。首次接触汽油的时间也与所研究的四种疾病无关。
该队列的分销工人所接触的汽油或苯浓度并非白血病（所有细胞类型）、急性髓细胞白血病、多发性骨髓瘤或肾癌的风险因素。
