BETHESDA, Md--A multistate, case-control study involving more than 1,200
children has failed to document a link between electromagnetic fields in
and around the home and acute lymphoblastic leukemia (ALL), the most common
form of cancer in children.
"Our study, overall, shows no evidence of an increased risk of
childhood ALL at residential magnetic field levels experienced by most
children in this country," said lead investigator Martha Linet, MD,
of the National Cancer Institute. "The results do not support the
hypothesis that children living in homes with high measured levels of magnetic
fields or high wire code levels have an increased risk."
[Wire code levels are based on power-line thickness, configuration,
and distance from the home, and closely correlate with magnetic field levels.]
Dr. Linet and her colleagues at the NCI and the Children's Cancer Group
(CCG) described the findings in a briefing at the National Institutes of
Health. Their full report appeared in the July 3, 1997, issue of the New
England Journal of Medicine.
"Some people claim that epidemiological studies are never definitive,"
Lawrence J. Fischer, PhD, director of the Institute for Environmental Toxicology,
Michigan State University, and chairman of the study's advisory group,
said at the briefing. "But this study certainly, in my view, tips
the scale to indicate that exposure to magnetic fields is not a major,
and probably not even a minor, component of cancer causation."
About 1,600 children will die of ALL in 1997, according to the American
Cancer Society. The issue of whether magnetic fields have contributed to
the rise in ALL that has occurred since 1974 has remained unresolved since
the first report suggesting a cancer link in 1979.
Last year, in an analysis of more than 500 studies, the National Research
Council (NRC) reported a "weak statistical correlation between childhood
cancer and wire configuration," but the new study suggests that even
this weak correlation can be explained by problems with study methodology.
The NCI/CCG study was designed specifically to address factors that
might account for the NRC findings and to overcome a number of shortcomings
noted in previous studies.
For example, the current study is four times larger than the next-largest
US study; in this study, magnetic field levels were measured inside the
home rather than estimated by examining power lines outside the home; measurements
were not limited to a single residence per child and were made within two
years of the child's diagnosis instead of many years later; and, finally,
the data collectors in the NCI study were blinded as to which children
had leukemia and which did not.
The new study enrolled 638 children with ALL who were under 15 years
of age; 58% were younger than age 5. Another 620 healthy children served
as controls, matched for age, race, and telephone exchange. All the children
lived in one of nine states: Illinois, Indiana, Iowa, Michigan, Minnesota,
New Jersey, Ohio, Pennsylvania, and Wisconsin.
The researchers measured magnetic fields in the children's current and
former homes, including those where their mothers lived during pregnancy.
They also assigned wire code levels to the homes.
The exposure measurements consisted of electronic meter readings taken
in a child's bedroom every 30 seconds for 24 hours. In addition, 30-second
"spot" measurements were made in the bedroom, family room, kitchen,
and the room the mother slept in when pregnant with the child.
Dr. Linet acknowledged that even this approach has its limitations.
"It is not known how well a single, 24-hour measurement taken in a
child's bedroom over one day characterizes that child's current magnetic
fields exposure, much less a child's overall exposure during earlier years,"
Four Levels of Exposure
The children were divided into four groups, based on the level of magnetic
fields to which they were exposed: below 0.065 µT (microtesla) (44.6%
of the 1,241 children with magnetic field measurements); 0.065 to 0.099
µT (19.3%); 0.1 to 0.199 µT (23.3%); and 0.2 µT or above
Although the middle two groups each showed a 10% increased risk of ALL,
compared with the children with the lowest exposure, and the highest group
showed a 24% increase, the team said these increases were not statistically
significant. Furthermore, there was no dose-response pattern of increasing
risk with increasing exposure.
Wire code classifications were also divided into four groups: underground
or very low exposure (43.5% of the 804 children classified); ordinary low
(28.6%); ordinary high (21.6%); and very high (6.2%).
The ordinary low exposure group had a 7% higher ALL risk than the very
low group, but the ordinary high group had a 1% lower risk than the very
low group, and the very high group had a 12% lower risk. Again, there was
no statistically significant difference.
"This is a strong study; it eliminates problems present in other
studies," Dr. Fischer said. "The results should be welcomed by
persons who must make decision regarding the mode with which we transmit
electrical power. It should also help scientists turn their attention to
research that may provide more productive results leading to the cause,
treatment, and prevention of this dread disease."
The magnetic fields study is part of a much larger ALL study being conducted
by the CCG involving 1,900 cases and 1,900 controls. Researchers are looking
at both genetic and environmental issues, including infections, parental
occupation and lifestyle, possible medical contributions (x-rays, drugs),
and specific exposures (pesticides, radon, solvents). The first of a series
of reports will be published early next year.