Objective: To investigate cause-specific mortality in a population exposed to high levels of 50 Hz magnetic fields, in a district of Rome.
Design: A cohort of 357 subjects that are resident in an area of about 100 meters on the two sides of a 60 kV power line was studied. Exposure to 50 Hz magnetic fields was evaluated by spot and long-term measurements and by adopting a model based on current load, characteristics of the line and distance of dwellings from conductors. The mortality experience of the cohort was observed from January 1st, 1980 through December 31st, 2003. Analysis took into account duration of residence and latency. Separate analyses were performed for three subcohorts characterized by different distances from the line and different estimated and measured field levels.
Main outcomes: Cause-specific standardized mortality ratios (SMRs) were computed comparing the observed number of deaths with the corresponding expected figures derived from the Latium Region mortality rates. This was done both in the overall cohort and in the subcohorts.
Results: All-causes mortality of the cohort did not differ from expected values (SMR 0.99; 95% CI 0.73-1.35; 40 observed). All cancers mortality showed a non-significant increase (SMR 1.34; 95% CI 0.82-2.18; 16 observed), that reached statistical significance in the analyses limited to subjects with more than 30 years of residence (SMR 2.09; 95% CI 1.05-4.19; 8 observed). Mortality from lymphohaematopoietic malignancies was increased, based on two cases of leukemia. All cancers mortality was highest in the subcohort closest to the power line and exposed to highest levels of magnetic fields. An increased risk for digestive tract neoplasms (SMR 3.57; 95% CI 1.49-8.58; 5 observed) was observed, which was mainly caused by pancreatic cancer.
Discussion and conclusion: Some increase in mortality of the overall cohort and especially of the highly exposed sub cohort was detected, notwithstanding the low power of the study. Other studies are currently on-going in the same area: a cross-sectional investigation on health status, an analysis of hospital discharge cards, a project for estimating individual exposure levels and a research in veterinary epidemiology. It is recommended to replicate the present study in areas presenting similar exposure patterns, in order to be able to pool data with the aim of estimating risk associated with the exposure levels of interest.