| Reference, type of study, country |
Study population |
Exposure assessment method |
exposures category/ magnitude/exposure metric category |
Risk estimates | Comments | |||||||||||||||||||||||||||||||||||||
| No. of cases |
OR (95% CI) |
|||||||||||||||||||||||||||||||||||||||||
| (Floderus et al., 1993) case-control, Sweden |
All male workers in a large region of mid-Sweden who were 20-64 year old in 1980. 250 leukemia cases from cancer registry, 1121 controls matched on age from census records | Exposures from JEMs for TWA and 3 other metrics in job held the longest in the 10 years before diagnosis. JEMs for 169 job categories based on 1015 magnetic field measurements taken at subject's workplace and job (or surrogate) |
|
OR with highest EMF exposure as high as 4.7 (2.2-9.7) after adjusting for ionizing radiation, benzene, and solvents. No change after adjusting for smoking, or pesticides | ||||||||||||||||||||||||||||||||||||||
| (Matanoski et al., 1993); case-control, USA |
Cohort of white males who were employed by AT&T or pensioned during 1975-80. Cases were deaths from leukemia (except CLL) whose job histories were obtained. For each case, 3 controls without leukemia were matched on gender, date of birth, date of hire, work status, and local telephone company | Cumulative exposures from job histories and JEMs for TWA and peaks from magnetic field measurements on 204 workers. Different exposures noted for central office technicians from old relay switches and new solid-state switches |
|
3 tests for trend (one-tailed) done over exposure quartiles measured by quartile ordinals, means, and log of means | ||||||||||||||||||||||||||||||||||||||
| (Sahl et al., 1993); case-control, California (USA) |
36221 southern California Edison Company utility workers. Cancer mortality determined for 1960-88. 10 controls per case matched on date of birth, gender, and race | Exposure based on job history and JEMs for TWA and 4 other metrics from 776 person-days of magnetic field measurements |
|
OR increase per defined exposure magnitude calculated by conditional logistic regression | ||||||||||||||||||||||||||||||||||||||
| (London et al., 1994); case-control, Los Angeles County (USA) |
Men age 20-64 years with cancer diagnosis and occupation in LA County tumor registry. 2355 leukemia cases (121 among electrical workers) compared with all other cancers with the exception of CNS tumors | Exposure based on job title reported on hospital records and on JEM for the TWA and 2 other metrics
derived from full-shift magnetic field measurements on 278 electrical workers and 105 non-electrical workers. Current measurements adjusted based on estimates of time spent on tasks 15-20 years previously |
|
No confounding detected with chlorinated hydrocarbon solvents, ionizing radiation, benzene or gasoline exhaust | ||||||||||||||||||||||||||||||||||||||
| (Thériault et al., 1994); case-control, Canada and France |
Cohorts of male utility workers in 3 large companies: base population of 223292; 4151 cancer cases determined from cancer registries, company records, etc. 6106 controls matched to cases by utility and date of birth | Cumulative exposure based on job history plus JEMs from 2066 workweek EMF measurements (50/60 Hz magnetic
fields, electric fields and pulsed EMF). JEMs for TWAs adjusted for past practices constructed for 260 job titles |
|
Trend calculated over 4 exposure groups. No significant confounding detected for benzene, gasoline, paint, solvents, or ionizing radiation | ||||||||||||||||||||||||||||||||||||||
| (Savitz & Loomis, 1995); case-control, USA |
138,905 men employed for > 6 months in 5 electric utilities followed for mortality in 1950-86 | Cumulative magnetic field exposure estimated from job history plus JEM based on 2842 magnetic field measurements |
|
Risks adjusted for PCB and solvent exposures. Ionizing radiation exposure insignificant | ||||||||||||||||||||||||||||||||||||||
| (Feychting et al., 1997); case-control, Sweden |
Combined occupational and residential exposure assessment: approximately 400000 subjects living within 300 m of transmission lines; 325 leukemia cases | Residential magnetic field exposure from a physical model of distance from transmission line and historical reconstruction of power loads. Occupational exposure for job on census before diagnosis was estimated from JEM developed from measurements of Floderus (Floderus et al., 1993) |
|
CML risks also significant with occupational + residential exposures >0.2 mT [OR 6.3, (1.5-27)]. Crude estimates of confounding from smoking and chemical exposures did not change results | ||||||||||||||||||||||||||||||||||||||
| (Miller et al., 1996); case-control, Ontario (Canada) |
Ontario Hydro cohort of (Thériault et al., 1994), 1484 cancer cases, 50 leukemia | Cumulative exposures estimated from job history in company records combined with JEMs for EMF derived from Ontario Hydro measurements taken by (Thériault et al., 1994). |
|
ORs adjusted for EMF interactions, benzene, herbicides, and ionizing radiation. Without EMF interaction, OR for highest exposure is 3.5 (0.56-22) | ||||||||||||||||||||||||||||||||||||||
| (Johansen & Olsen, 1998); cohort, Denmark |
Employees of 99 Danish utility companies, 32,006 total |
Magnetic field exposures for subject's first job taken from a JEM based on expert judgment and 24-h measurements on 129 workers at 6 companies. JEM has 5 exposure categories (including background and unknown) for 25 job titles and 19 sites |
|
Risks calculated as an SIR adjusted for age, gender, and date of diagnosis relative to background | ||||||||||||||||||||||||||||||||||||||
| (Kheifets et al., 1997a); Meta-analysis |
38 studies of leukemia in EMF-exposed worker populations | Indicators of EMF exposure varied from a single job (e.g. welding) to magnetic field measurements |
|
Significant heterogeneity (p < 0.05) among AML and CLL studies with broad definitions of EMF exposure | ||||||||||||||||||||||||||||||||||||||