Study type: Epidemiological study (observational study)

Residential Magnetic Field Exposure and Childhood Brain Cancer: A Meta-Analysis epidem.

Published in: Epidemiology 2008: 424-430

Aim of study (acc. to author)

A meta-analysis of studies on magnetic field exposure and childhood brain cancer was conducted to evaluate homogeneity in the results and to derive a summary effect estimate. Furthermore, the potential for selection bias in childhood leukemia studies should be assessed by comparing results from studies of childhood leukemia and childhood brain cancer.
Thirteen studies were included in the meta-analysis: 7 studies were conducted in Europe (Feychting et al (1993), Olsen et al. (1993), Verkasalo et al. (1993), Tynes et al. (1997), , UK Childhood Cancer Study Investigators (1999), Schüz et al. (2001), Draper et al. (2005)), 4 in the USA ( Wertheimer et al. (1979), Savitz et al. (1988), Gurney et al. (1996), Preston-Martin et al. (1996)), and 2 in Asia (Lin, abstract in English and dissertation in Chinese (1989) and Saito, paper presented on international symposium (2003).

Endpoint/type of risk estimation

Type of risk estimation: (odds ratio (OR))



Exposure groups

Group Description
Reference group 1 wire code current configuration: low
Group 2 wire code current configuration: high
Group 3 wire code current configuration: very high
Reference group 4 magnetic field exposure: < 0.2 µT
Group 5 magnetic field exposure: ≥ 0.2 µT
Group 6 magnetic field exposure: > 0.3 µT
Group 7 magnetic field exposure: > 0.4 µT


Results (acc. to author)

None of the analyses showed statistically significant increases in childhood brain cancer risk in relation to various exposure assessment of the residential magnetic field. The summary effect estimate for magnetic fields above 0.3 or 0.4 µT was not significantly elevated, but the analysis was based on low numbers. No single study had a substantial effect on the summary effect estimates. There was no indication of publication bias.

Limitations (acc. to author)

The possibility of selection bias or exposure assessment bias in childhood leukemia studies and childhood brain tumor studies cannot be excluded.

Study funded by

Related articles