Study type: Epidemiological study (observational study)

A pooled analysis of magnetic fields, wire codes, and childhood leukemia. Childhood Leukemia-EMF Study Group. epidem.

Published in: Epidemiology 2000; 11 (6): 624-634

Aim of study (acc. to author)

The association between 50/60 Hz magnetic fields and childhood leukemia was investigated. For this purpose a pooled analysis was conducted on the primary data from fifteen studies carried out in England (Coghill et al.), New Zealand (Dockerty et al.), Mexico (Fajardo-Gutierrez et al., report published in Spanish), Sweden (Feychting et al.), Rhode Island/USA (Fulton et al.), Ontario/Canada (Green et al.), Eastern USA (Linet et al.), Los Angeles/USA (London et al.), Canada (McBride), Germany (Michaelis et al.), Denmark (Olsen et al.), Denver/USA (Savitz et al.), Sweden (Tomenius), Norway (Tynes et al.), Finland (Verkasalo et al.), and Denver/USA (Wertheimer et al.).

Further details

The authors of the 15 studies supplied anonymous records on individual subjects of their studies. The target measure for magnetic field exposure was defined as a child's exposure up to 3 months before diagnosis. Therefore the magnetic field exposure estimate was recalculated for each child. Wire codes used in all North American studies were separately analyzed. Four studies recorded magnetic fields and wire codes.

Endpoint/type of risk estimation

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



Exposure groups

Group Description
Reference group 1 magnetic field category: ≤ 0.1 µT
Group 2 magnetic field category: > 0.1 - ≤ 0.2 µT
Group 3 magnetic field category: > 0.2 - ≤ 0.3 µT
Group 4 magnetic field category: > 0.3 µT
Reference group 5 low current code LCC (includes ordinary low current, very low current and underground)
Group 6 ordinary high current code (OHCC)
Group 7 very high current code (VHCC)


Study size

Type Value
Participants 13,159

magnetic field measurements: 2656 cases and 7084 controls; wire code information: 1457 cases and 1962 controls; magnetic field measurement and wire code information: 850 cases and 1004 controls

Statistical analysis method: ( adjustment: )

Conclusion (acc. to author)

The magnetic field estimates tend to show little or no association of fields below 0.3 µT with childhood leukemia, but all studies with cases and controls in the categorie more than 0.3 µT showed positive associations. The association of wire codes with childhood leukemia varied considerably across the studies (odds ratios ranging vom 0.7 to 3.0), therefore summary estimates were omitted due to the extensive unexplained heterogeneity. The results of the pooled analysis suggested that appreciable magnetic field effects, if any, may be concentrated among relatively high and uncommon exposures.

Study funded by

Comments on this article

Related articles