Investigation of the possible association between exposure to diesel exhaust particulates in British coalminers and lung cancer

The general aims of this study were firstly to retrospectively assess underground exposures to respirable diesel exhaust paniculate for a cohort of British coalminers and secondly to investigate the association between these exposures and death due to lung cancer. The approach for the second of these aims was essentially a repetition of the exposure-response calculations of Miller et al (1997), which were based on a less comprehensive analysis of underground respirable diesel exhaust paniculate exposures.The cohort comprised 18166 colliers who were employed in 10 former British Coal mines between the early 1950’s to the mid 1980’s. Six of these collieries used diesel locomotives, and for this study, these locomotives are the only sources of diesel exhaust particulate that are considered.All of the collieries involved participated in the (then) National Coal Board’s “”Pneumoconiosis Field Research (PFR)”” study which combined routine monitoring of dust exposure concentration with regular occupational health surveys. In this report, as in all previous reports covering PFR studies, the collieries involved in the study are coded by letter. The ten collieries involved were coded as collieries C, F, P, V, K, Q, T, W, X and Y. Of these, the last 6 were the collieries in which diesel locomotives were used.In addition to the respirable particulate mass measurements (and in many cases compositional analysis) that were reported in the regular PFR survey reports, measurements of nitrous fume concentrations (individual estimates of both nitrogen oxide (NO) and nitrogen dioxide (NO2) collectively referred to as “”NOX””) were also carried out for 5 of the 6 diesel-using collieries listed above. These measurements were confined (approximately) to the period between 1978 and 1983.These respirable particulate mass and nitrous fume data have been used separately to derive two independent estimates of the respirable diesel exhaust particulate mass exposure concentrations for underground worker.Where the starting point was the PFR respirable particulate mass measurements, indirect estimations were made of the fraction of the total respirable particulate mass exposure concentration for underground workers that arose from the exhaust of the diesel locomotive. Central to these estimations were the shift-average respirable diesel exhaust particulate exposure concentrations for the locomotive drivers.Where the starting point was the NOX data, estimation of respirable diesel exhaust particulate mass concentrations were based on an estimated ratios between the concentrations of NOX and respirable particulate in the diesel exhaust emission.Estimates based on PFR respirable particulate mass measurements were carried out for each of the (approximately 5 year) intervals between individual PFR medical surveys, but the final outcome of these analyses was to recommend that single (time independent) estimates be adopted for each seam or colliery.The analyses that were based on NOX data also led to definition of single estimates, for each seam or colliery, of either the shift average respirable diesel exhaust particulate concentrations (where travelling distance is seam or colliery specific) or single coefficients of travelling distance (where travelling distances varied among different coal faces in the same seam or colliery).For four of the five participating collieries for which NOX measurements were reported, the two sets of estimates compared well. For these collieries, this report recommends that the mean of the two corresponding estimates of the shift average respirable diesel exhaust particulate concentrations be adopted.The exception was Colliery K, where agreement between the corresponding estimates was poor. For this particular colliery, suspicion of poor control over NOX measurements led to these being ignored, and the estimate that was finally adopted was that derived from the PFR dust data.The shift average respirable diesel exhaust particulate concentrations that were estimated in this project were then used in an investigation of the relationship between cumulative exposure and death due to lung cancer. The methods of statistical analyses used for this exercise were identical to a separate study carried out by the IOM on behalf of British Coal, to look more broadly at the relationships between underground working in British coalmines and cause of death (Miller et al, 1997). In both cases, exposure-response relationships were investigated using Cox regression models for survival time data, taking account of the time-dependent nature of the cumulative exposures. The underlying time axis on which risk sets were formed was the time on study after cohort entry, and terms were included for significant confounders, notably age, smoking habit and cohort entry date. Having adjusted for these confounders the additional effect of exposure variables, both unlagged and lagged by 15 years was investigated.The main differences between the two studies were that (i) whereas Miller et al (1997) looked at exposure-mortality relationships for a range of exposure material and diseases, the study reported here relates solely to diesel exhaust particulate and lung cancer, and (ii) whereas Miller et al (1997) used cumulative travelling time on diesel locomotives as a surrogate for cumulative exposure, this study used detailed estimates of shift-average respirable diesel exhaust particulate exposure concentrations, together with details of work histories, to provide detailed cumulative individual exposures.The principal finding of this study was weak evidence of an association between lung cancer mortality and exposure to respirable diesel exhaust particulate. However, this association was dependent on the inclusion of a colliery (Colliery Q) for which exposures were estimated to be very much higher than the other collieries, and where mortality was only slightly higher than regional background rates. No association was found among men with different exposures working in the same collieries.The findings reported here are in general agreement with those of Miller et al (1997). This outcome is not surprising in that (a) differences between seams or between collieries in the estimated shift-average exposure concentrations were not large, and (ii) the same work-history information that was used by Miller et al. to estimate cumulative travelling time on diesel locomotives was used for the study reported here to convert from shift average exposure concentrations to cumulative exposures.The estimated cumulative exposures for underground workers in this study were generally below 1 g.hr.m-3 but for diesel drivers, higher exposures up to a maximum of about 11 g.hr.m-3 were estimated. It may be reasonably assumed that the findings of this study will apply equally well to other non-underground workers such as truck drivers if the average cumulative exposure concentrations are of the same order or less. Reference is also made to the possible implications of the findings of this study to the relationship between exposures to diesel exhaust paniculate in the urban environment and development of lung cancer. “”