Objectives Workers in coke ovens are potentially exposed to polycyclic aromatic hydrocarbons (PAHs), and studies have variously shown increased lung cancer risks. A study of two British company cohorts (NSF & BSC), defined in 1967, reported in 1991 on a 20-year mortality follow-up of over 6,500 workers.

Analyses of mortality risks and exposures showed some statistically significant exposure-response relationships, as did analyses characterising exposure by time worked on coke oven tops. Limitations of the exposures were documented: notably, exposures to BSM were estimated only up to the start of follow-up (1967); and the exposures calculated made no allowance for cancer latency (time elapsing between exposure and increase in risk). The present reanalysis project was designed to bypass those limitations.

Methods Revised exposures to the general class of benzene-soluble materials (BSM) and to the specific marker benzo[a]pyrene (B[a]P) were calculated from work histories up to 1976, as were estimates of time worked on oven tops. Lung cancer mortality was analysed using Cox regression models, with time-dependent exposures lagged by 10 years to allow for latency.

Results Most results were not statistically significant. There were no significant trends with continuous measures of exposure to either BSM or B[a]P or of time worked in ovens job locations. With grouped exposures, at NSF, the highest category of B[a]P exposure had a statistically significant relative risk of 1.51. In BSC, the relative risk coefficient for working 5 or more years there was 1.81, which was statistically significant (but lower than the equivalent value of 2.10 from the previous analyses).

Conclusions These results showed some signs consistent with an effect of coke ovens work on lung cancer risk, but there was little firm evidence of any real effect.