ORIGINAL ARTICLES: Author Commentaries
Two studies, conducted by our group 1 and by Feychting and colleagues, 2 address the possible association of extremely low-frequency magnetic fields (ELFMF) with neurodegenerative diseases. The studies overlap in their topic, case ascertainment, exposure assessment, and (partly) in their study populations. Our study included Swedish engineering industry workers with a high probability of ELFMF exposure, while the Feychting study included employed Swedes in general. The most important differences are probably in the study design.
The outcomes of the 2 studies partly corroborate each other. Neither study shows associations with Parkinson’s disease or multiple sclerosis. For Alzheimer’s disease (AD) the results were partly consistent, while for amyotrophic lateral sclerosis (ALS) there was a clear difference; our study indicated an association, while this was not the result in the Feychting study.
An essential question for a comparison of the results is to what extent comparable data were used in the exposure assessment. We believe the 2 studies have used exposure classifications that are largely identical, although we are not certain. We used the exposure data developed in our case-control study of 1993, 3 covering 170 jobs. For the 100 most common job titles a job exposure matrix was later published, 4 and for another 70 job titles the data have not been published. Feychting et al refer to a job exposure matrix (JEM) developed in a case control study published in 1998. 5 We infer from the exposure distribution and number of unclassified subjects that their exposure assessment is based on the same or nearly the same 170-category matrix.
For the present analysis, Feychting et al 2 used a closed cohort, assessing exposures from job titles in 1970/1980 and linking to mortality from 1981/1987 through 1995. We 1 used a dynamic cohort, yielding a more even age distribution over time and a reduced potential for survival effects. We entered subjects each year from 1985 to 1994, assessing exposures from job titles in 1985/1990 and following mortality from 1985 through 1996. Our cohort thus includes individuals with a closer time lag between definition of exposure and the outcome.
The median age at death among our AD cases was 69 years (7–8 years younger than in the other study). Our follow-up thus captured early-onset AD with a comparatively short survival. The exposure assessment preceded the outcome by 11 years or less, thus referring to a late stage in the disease process. Our study suggested an exposure-response pattern with a 4-fold increase in risk in the highest exposure group (>0.53 μT). The Feychting et al analysis most similar to ours corroborates this finding, with a risk estimate of 3.4 for men exposed >0.5 μT, based on occupation in 1980, with follow-up through 1990 of the survivors of 1987, ie, within 10 years of the exposure assessment. In their analysis where the survivors of 1987 were followed through 1995 according to the exposure 17 years earlier (1970) no association was found. Note that 80% of the cases in this latter analysis occurred 20–25 years after the exposure assessment. Both studies thus support the view that a follow-up limited to 10 years after the last known occupation is of importance, although some results of the Feychting study are not consistent with this pattern.
Our finding of an association for ALS may be due to our cases being younger (median age 63 versus 69-70 in the other study). Most of our cases occurred before the usual age of retirement. If high ELFMF exposure is a late-acting factor, with no or short lag time, this could contribute to the lack of association in the Feychting study. Another explanation could be that certain exposure characteristics are of importance. Extremely high peak values should be more prevalent among the exposed engineering industry workers compared with exposed workers in general. Feychting et al reported an increased risk for welders but not, for example, for train engineers, which supports the possibility that high peak values may be important relative to ALS.
1. Håkansson N, Gustavsson P, Johansen C, Floderus B. Neurodegenerative diseases in welders and other workers exposed to high levels of magnetic fields. Epidemiology. 2003; 14: 420–426.
2. Feychting M, Jonsson F, Pedersen NL, Ahlbom A. Occupational magnetic field exposure and neurodegenerative disease. Epidemiology. 2003; 14: 413–419.
3. Floderus B, Persson T, Stenlund C, Wennberg A, Ost A, Knave B. Occupational exposure to electromagnetic fields in relation to leukaemia and brain tumours: a case-control study in Sweden. Cancer Cause Control. 1993; 4: 465–476.
4. Floderus B, Persson T, Stenlund C. Magnetic field exposures in the workplace: Reference distribution and exposure in occupational groups. Int J Occup Environ Health. 1996; 2: 226–238.
5. Feychting M, Pedersen NL, Svedberg P, Floderus B, Gatz M. Dementia and occupational exposure to magnetic fields. Scand J Work Environ Health. 1998; 24: 46–53.