|
Order back issues of BENER | Feedback |
The authors conducted an international case-control study of the cancer risk associated with occupational exposure to magnetic fields in male electric utility workers. The study was nested within 3 cohorts: 170,000 men employed by Electricite de France-Gaz de France (EDF), 31,543 men employed by Ontario Hydro, Ontario, Canada (OH), and 21,749 men employed by Hydro-Quebec, Quebec, Canada (HQ) from 1970 through 1989. There were 4,151 cases of cancer newly diagnosed in the HQ, OH, and EDF cohorts between 1 January 1970 and 31 December 1988, 1 January 1973 and 31 December 1988, and 1978 and 31 December 1989, respectively. The median year of birth of the cases diagnosed at HQ, OH, and EDF was 1923, 1918, and 1935, respectively. The cases were identified through searches of company files (EDF) or Quebec and Ontario tumor registries (HQ and OH). Controls were chosen at random from lists of other employees from the same utilities without cancer, matched to each case by year of birth. There were 4 controls selected per case for the cancers of a priori interest (those previously associated with magnetic fields; leukemia, hematopoietic cancer, brain cancer, and melanoma); 1 control for each case for other cancers (for a total of 6,106). Because control status was determined for the time of diagnosis, a man selected as a control could become a case later in the study. For each case and control, a detailed occupational history was obtained from company records. Each subject's cumulative exposure to magnetic fields was estimated using a 2-step process. Measurements were made of current workday exposures of 2,066 workers (829 at EDF, 771 at OH, and 466 at HQ) who performed tasks similar to those in the cohorts. Workers wore Positron dosimeters (Positron Industries, Montreal, Canada) attached to their belt or shirt pocket for a 5-day work week. They also recorded the type of work, location, line voltage and current, and other details for the measurement period. Information was obtained on the work histories of cases and controls, past current loads, work practices, and usage from interviews with company personnel and reviews of system data. This information was then used to construct a job exposure matrix for each utility using slightly different methods in each case. Estimates were also made of possible confounders (exposures to ionizing radiation, chemical agents listed as carcinogens or possible carcinogens by the International Agency for Research on Cancer and the American Conference of Governmental Industrial Hygienists, smoking, and socioeconomic status), again with differences in the information available among the utilities. Possible associations between cumulative magnetic field exposures of the workers with 31 types (sites) of cancer were examined by computing odds ratios (ORs) and their 95% confidence intervals (CIs). The median magnetic field exposures experienced by the HQ, OH, and EDF workers were 0.19, 0.22, and 0.10 uT (1.9, 2.2, and 1.0 mG), respectively. The median cumulative magnetic field exposure at these facilities (the product of the time-weighted arithmetic mean exposure in each job and the years of work at that job) was 6.34, 4.91, and 2.21 uT-yr, respectively. The mean cumulative lower quartile exposures were 3.63, 2.62, and 1.70 uT-yr and the mean cumulative upper quartile exposures were 11.83, 9.10, and 3.07 uT-yr, respectively. Overall, the median cumulative magnetic field exposure for all 3 facilities was 3.15 uT-yr. Exposures for selected jobs were reported and demonstrated considerable variation between job title and utility: among the highest were OH hydroelectric operator (mean 5.39 uT), OH general trades (2.06 uT), HQ substation equipment electrician (2.36 uT); among the lowest were HQ nuclear operator (0.13 uT), and EDF hotstick distribution lineman (0.09 uT). The risk for all cancers combined was not significantly elevated for workers experiencing cumulative magnetic field exposures above the 3.15 uT-yr median, OR of 1.01 (CI 0.91-1.13). Among cancer sites of a priori interest, the only risks that were significantly elevated for workers whose cumulative exposures exceeded the overall median were for acute myeloid leukemia (AML), OR of 3.15 (CI 1.20-8.27), and for acute nonlymphoid leukemia (ANLL), OR of 2.41 (CI 1.07-5.44). When examined within each utility separately, the ANLL and AML ORs varied widely and showed no dose-effect relationship. For example, the ORs for ANLL for workers at HQ, OH, and EDF who had cumulative above median magnetic field exposures were 0.75 (CI 0.00-100+), 6.24 (CI 0.95-77.49), and 1.78 (CI 0.57-5.39), respectively. The overall risk for brain cancer was not significantly elevated, OR of 1.54 (CI 0.85-2.81). When subdivided into astrocytomas and glioblastomas, the astrocytoma risk among workers having cumulative magnetic field exposures above the 90th percentile value, 15.7 uT-yr, was significantly elevated, OR of 12.29 (CI 1.05-143.5). This OR, however, was based on only 5 cases and was highly dependent on adjustment for socioeconomic status; when adjusted by a single white collar/blue collar variable, the OR was reduced to 1.53 (CI 0.38-6.25). No associations with magnetic fields were observed for any of the other specific cancer sites including chronic lymphoid leukemia (CLL), a leukemia subtype previously associated with EMF exposure. The absence of an association of magnetic field exposure with CLL may have been the result of limited follow-up beyond the age of 60. Adjusting for confounding by other physical and chemical agents did not alter the results. The authors concluded that, in spite of attempts to conduct a study with adequate statistical power, definitive evidence of an association between exposure to magnetic fields and leukemia and brain cancer was not obtained. The data do support the hypothesis that magnetic field exposure is associated with an increased risk for AML and ANLL.