Lung cancer mortality in a cohort of workers in a petrochemical plant: occupational or residential risk?

The purpose of the present study is to examine the role of environmental (non occupational) exposures in lung cancer risk among petrochemical workers at a large petrochemical plant built on the Sicilian coast in the immediate vicinity of the town of Gela, Italy in 1960. The cohort included workers employed in the Gela petrochemical plant in 1960-1993. We looked at mortality rates for the period 1960-2002. An internal comparison was performed between two categories of workers with different likelihood of residence in Gela during the period of employment. The rate ratio of mortality from lung cancer comparing "probable residents" with "possible non residents," adjusted for age, calendar period, andjob classification (only blue collar, only white collar and both), was 1.66 (90% Confidence Interval 1.07-2.58). Although the information collected is quite sparse and no inferences can be made about risk sources, the results show a possible excess of residential/environmental risk from lung cancer mortality for those workers more likely to have been residents in Gela.

Gela area is an Italian polluted site qualifying for remediation because of widespread contamination from a petrochemical complex. This study investigates mortality and morbidity of the cohort of employees in Gela petrochemical plant with the aim of disentangling the health effect of work and residence. Work experience was classified in terms of job title, while an ad hoc mobility model was applied to define qualitative categories of residence in Gela as probable residents and probable commuters. Mortality rate ratio for lung cancer was 1.60 (90% CI 1.01-2.53) in workers probable resindents compared to probable commuters. For the same comparison, Hospital Discharge Prevalence Ratio for COPD was 1.39 (0.94-2.07). The crude categories of work and residence limits the interpretation of the causal nature of the study results. Despite several limitations, the results for respiratory pathologies are compatible with an etiological role of the documented contamination.