Carcinogenic effects of diesel emissions and lung cancer: the epidemiologic evidence is not causal.

The effects of diesel engine exhaust in lung carcinogenesis have been evaluated by several scientific organizations and government agencies. This complex issue has required a multidisciplinary approach including atmospheric measurements, toxicology, chemical carcinogenesis, epidemiology, and risk assessment. One important aspect of the epidemiological studies that deserves further attention is the confounding effects of cigarette smoking. Only some epidemiological studies have statistically adjusted for cigarette smoking, usually by years of smoking, cigarettes per day, or pack-years. Some studies obtained smoking information from proxy interviews. However, differences in "tar" intake, interpuff interval, depth of inhalation, and other smoking behavior patterns were not evaluated. These smoking parameters are rarely collected for occupational data analysis, yet the inability to adjust statistically for such parameters may result in a small degree of residual confounding. Because the highest odds ratios for lung cancer associated with diesel engine exhaust are usually less than 2 or 1.5, possible residual confounding effects of smoking may have resulted in spurious associations.