No cases were observed in the reference category for soft tissue sarcoma, therefore rate ratios could not be calculated. On the other hand, four incident cases were found in the third category of heavy metals among women (RR = 4.71, not statistically significant, n.s.). When results for men and women were combined, six deceased cases were found and an excess in the highest category of exposure (RR = 16.54; 95%CI = 1.72-159.07) was detected.
When all analyses were repeated considering exposure to NO2 as a potential confounder, none of the rate ratios estimates changed substantially and the overall findings were confirmed (data not shown).
We performed an additional analysis of mortality considering only residents who were present in the study area in 1990. This sub-cohort represents the 78.8% of the total cohort (91.3% of the total person-years). The results were substantially similar with wider confidence intervals although statistical significance was reached for liver cancer among women at the 4th level of exposure (RR = 7.39; 95%CI = 1.11-49.10, based on 10 cases).