Purpose: We investigated the risk of death for hematological malignancies (HMs) in the area surrounding an Italian petrochemical refinery, where atmospheric concentrations of benzene (known carcinogen) had not been adequately monitored in the past.
Methods: We performed a population-based case-control study, using conditional logistic regression to estimate odds ratios (ORs) of HM death, with 95 % confidence intervals (CIs), and p trends, in relation to tertiles of time-weighted average residential proximity to the refinery. We identified 177 HM deaths and 349 sex- and age-matched controls from municipal files. Confounding factors were investigated from interviews with consenting relatives for 109 HM deaths and 178 matched controls.
Results: For males and females combined, risk of HM death was unrelated to residential proximity. For females, ORs of HM death by increasing tertiles of proximity were 1, 2.74 (95 % CI 1.48-5.09, significant) and 1.49 (95 % CI 0.76-2.92) (p trend 0.184). For the subgroup of persons who plausibly spent most of their time at home (long-term retired, homemakers or unemployed, 53 cases, 79 controls), the ORs of leukemia plus non-Hodgkin lymphoma death (38 cases, 56 controls) by increasing tertiles of proximity were 1, 3.44 (95 % CI 1.04-11.37, significant) and 3.25 (95 % CI 0.82-12.87) (p trend 0.083).
Conclusions: No increased risk of HM death for males and females combined living close to the refinery was found. However, the findings for females and a subgroup plausibly spending most of their time at home suggest a relation between increased risk of HM death and residential proximity to the refinery.