Mortality temporal trends and cancer incidence profiles of residents in the petrochemical industrially contaminated town of Gela (Sicily, Italy)

Abstract

Objective. In 2000, a vast area in Gela (Sicily, Italy) was defined as a national priority contaminated site due to pollution from a petrochemical complex. This study is aimed at addressing the influence of the petrochemical complex on the health profile of residents in Gela. Methods. Trend analysis by gender was performed for mortality for all diseases and malignant cancers, in the period 1980-2014 for residents in the municipality of Gela, by directly standardized rates and Joinpoint regressions, using, as a reference population, people resident in the Sicily region. SMRs were computed for 5-year periods in the same timespan. Since the beginning of the period analyzed, the share of population of Gela represents 1.5% of total residents in Sicily. Cancer incidence was analyzed for the period 2007-2012 applying a hierarchical Bayesian model to estimate Standardized Incidence Ratios (SIR). Ranks of these ratios were computed to highlight the most incident diseases affecting the population. Malignant neoplasms of lung, stomach and colon were selected because of a priori interest, as they are associated, in etiological terms, with the main contaminants found in the area. Malignant neoplasms of liver, pancreas and larynx were selected as “control diseases” since they share the same main risk factors (smoke and alcohol consumption) of neoplasms of a priori interest, but are not associated with the priority index contaminants identified in Gela. Results. Mortality rates for all causes combined in both genders in Gela decreased over time, but they were higher than those of the whole Sicilian population. The trend of mortality rates due to all malignant cancers increased in men, especially from 1980 to 1987. This result was confirmed by the Joinpoint regression (annual percentage change (APC) 9.8). SMRs analysis showed significant excesses in mortality due to all diseases for both genders compared to the reference population. Other excesses were observed for mortality due to malignant cancers in men and for circulatory diseases in women. The trend for cancers in women in Gela increased from the mid-nineties but less than in men. SIR estimates were higher than 1 for all the diseases analyzed and in both sexes, and their ranks highlighted that cancer sites of a priori interest hold higher positions than “control diseases”, although credibility intervals overlapped. Conclusions. Results highlight that the health profile of residents in Gela is worse than the one of the reference population. Moreover, cancer incidence is in excess in all the sites analyzed and mortality due to all cancers combined has a trend compatible with a cumulative impact due to petrochemical contamination.