| Author, year | Study population (location) | Exposure assessment | Covariates | Exposure group | RR (95% CI); observed* |
| Deapen and Henderson, 198658 | 518 ALS cases identified from the ALS Society 1977–79, 518 friend controls (USA) | Exposure defined as “electrical occupations” | Electrical occupation | 3.8 (1.4 to 13.0); 19 | |
| Gunnarsson et al, 199159 | 1961 deaths from ALS in the Swedish population 1970–83, 2245 population-based controls (Sweden) | Occupational categories | Age, men only | Electricity workers | 1.5 (0.9 to 2.6); 32 |
| Gunnarsson et al, 199260 | 92 prevalent cases of motor neuron disease at hospitals in central and southern Sweden, 372 population-based controls (Sweden) | Occupational categories | Age, men only | Electricity work | 6.7 (1.0 to 32.1); 4 |
| Davanipour et al, 199757 | 28 clinic-based prevalent cases of ALS, 32 relatives controls (USA) | Occupations classified by industrial hygienist as electrical (high or medium) or non-electrical | Sex | Average lifetime occupational exposure, 75th percentile | 2.3 (0.8 to 6.6) |
| Savitz et al, 19984 | Cohort of 139 905 men working at five US electric utility companies 1950–86. Mortality compared to the general population, and internal comparisons within cohort; 33 ALS cases identified (USA) | Used work-shift measurements of extremely low frequency fields in a sample of workers to categorise occupations | Age, calendar year, race, social class, work status, poly-chlorinated biphenyl and solvent exposure | Employment in exposed occupation, years0–<55–<20⩾20 | 1.0; 221.8 (0.7 to 4.7); 182.4 (0.8 to 6.7); 16 |
| Savitz et al, 199842 | 114 deaths from ALS 1985–91, matched controls selected from among persons who died from other causes not related to electric and magnetic fields (USA) | Occupations classified according to work in “electrical occupations” | Age, social class, men only | Electrical occupation | 1.3 (1.1 to 1.6); 114 |
| Johansen and Olsen, 1998†61 | Cohort of 26 135 men working in an electric utility company 1900–93; 14 deaths from ALS identified. Mortality compared to national mortality rates (Denmark) | Job-exposure matrix; categorised workers into high, medium, low or background magnetic field exposure | Age, duration of employment, men only | Medium to high exposure (⩾0.3 μT) | 2.5 (1.1 to 4.8); 9 |
| Noonan et al, 200244 | 312 deaths from ALS 1987–96 in Colorado, 1248 matched controls among other causes of death (USA) | Three methods: 1) primary occupation classified as electrical or non-electrical, 2) combination of occupation and industry codes classified into four exposure categories, 3) population-based job-exposure matrix constructed from 2400 magnetic field measurements | Age, race, occupational grouping, men only | Electrical occupationNo exposurePossible exposureDefinite or probable exposureAverage exposure<0.1 μT0.1–0.19 μT0.2–0.29 μT⩾0.3 μT | 2.3 (1.29 to 4.09); 191.0; 2421.18 (0.83 to 1.67); 2851.75 (1.00 to 3.06); 191.0; 590.79 (0.54 to 1.15); 1341.21 (0.75 to 1.93); 510.77 (0.37 to 1.59); 12 |
| Håkansson et al, 200345 | Cohort of Swedish engineering industry workers with a large proportion of resistance welders, 537 692 men and 180 529 women followed in Causes of Death Registry 1985–96; 97 deaths from ALS identified (Sweden) | Information on occupation from censuses linked to a job-exposure matrix constructed from over 1000 measurements | Age, sex, socioeconomic status | Average exposure<0.164 μT0.164–0.250 μT0.250–0.530 μT>0.530 μT | 1.0; 151.58 (0.88 to 2.81); 521.95 (0.97 to 3.92); 172.16 (1.01 to 4.66); 13 |
| Feychting et al, 200346 | Economically active Swedish population 1980, 4 812 646 persons, followed in the Causes of Death Registry 1981–95; 1965 deaths from ALS identified (Sweden) | Information on occupation from 1970 and 1980 censuses linked to a job-exposure matrix constructed from over 1000 measurements | Age, sex, socioeconomic status | Average exposure ⩾3 μT 1970 and 19801970 or 1980⩾5 μT1970 and 19801970 or 1980Electrical occupations | 0.8 (0.6 to 1.1); 480.8 (0.7 to 1.0); 1970.6 (0.4 to 1.0); 160.7 (0.6 to 1.0); 831.4 (1.0 to 1.8); 47 |
| Park et al, 200547 | 6347 deaths from motor neuron disease in 22 states during 1992–98; controls were deaths from other causes not related to electric and magnetic fields or solvents (USA) | Information about usual occupation from death certificate linked to job-exposure matrix constructed from 2400 magnetic field measurements | Age, sex, race, region, socio- economic status | Analysed as a continuous variable with 10 equal width exposure strata | 0.94 (0.73 to 1.20); 5965 |
| Sorahan and Kheifets, 200748 | Mortality 1973–2004 among 79 972 employees of the former Central Electricity Generating Board of England and Wales with computerised work histories 1971–93; 68 deaths from motor neuron disease identified. Compared to mortality rates for the general population of England and Wales and internal comparisons (Great Britain) | Information on job and facility (location) used to estimate exposures to magnetic fields. Power station workers considered to have highest exposure | Age, sex, calendar year, socioeconomic status (the latter only in internal comparisons) | Electricity generation and transmission workers† RR per μT year cumulative exposure | 0.87 (0.67 to 1.10); 681.04 (0.84 to 1.30); 78 |
| Roosli et al, 200749 | Mortality 1972–2002 among 20 141 Swiss railway employees. Internal comparisons within cohort, train drivers had highest exposure and station masters the lowest. | Exposure was assessed through on-site measurements and modelling of past exposure | Age, time period.Study was restricted to men | Railway train driversCumulative lifetime exposure >median | 1.31 (0.31 to 5.59) 2.32 (0.70 to 7.73) |
*Observed number of exposed ALS cases; †standardised mortality ratios (SMRs).