DATE: Thursday, July 3, 1997                TAG: 9707030664

SECTION: FRONT                   PAGE: A1   EDITION: FINAL 

SOURCE: BY CURT SUPLEE, THE WASHINGTON POST 

DATELINE: WASHINGTON                        LENGTH:   99 lines

POWER LINES DON'T INCREASE CANCER RISK, STUDY SAYS MAJOR STUDY ANSWERS QUESTIONS RAISED IN '79 ABOUT CHILDREN WHO LIVE NEAR THE LINES.

Children who live near high-voltage power lines do not have a greater risk of developing cancer than other youth, a long-awaited major study has concluded.

The eight-year, $5 million project, coordinated by the National Cancer Institute and involving 1,250 subjects in nine states, is the largest U.S. attempt so far to investigate one of the most emotionally incendiary public-health issues in American life: Whether exposure to electrical and magnetic fields from power lines and other everyday sources can cause or promote childhood leukemia.

The research ``tips the scale'' in that debate, said Lawrence J. Fischer of Michigan State University, who heads NCI's outside advisory committee on the study, toward the conclusion that power-line EMFs ``are not a major, and probably not even a minor, component to the cause of cancer,'' he told a news conference Wednesday.

The new report - published in today's issue of The New England Journal of Medicine - already has met with mixed reactions among EMF health experts.

Jerry R. Williams, director of the radiobiology lab at The Johns Hopkins Oncology Center, called it ``the best, strongest study so far,'' and one that may hasten the day when an incontestable accumulation of negative results means that ``at some point, we're going to have to say there's no risk.''

But epidemiologist David A. Savitz of the University of North Carolina at Chapel Hill, an internationally recognized leader in the field, said the new finding, though ``an important addition,'' nonetheless ``shares the uncertainties'' of many prior attempts to estimate a person's past exposure to varying EMFs by calculations made years after diagnosis of cancer.

``I wish I could say they'd succeeded completely,'' Savitz said. ``This may be as close as we come for a while to a major new effort.''

Such efforts have proliferated since a 1979 study in Denver first suggested that children living close to high-energy power lines were two to three times more likely to develop cancer. Dozens of research groups worldwide pounced on the issue, and two decades later there is some evidence suggesting that workplace or household exposure to high EMFs may heighten the risk of brain cancer, leukemia and other maladies.

But results have been notoriously ``conflicting and at times confusing,'' Robert N. Hoover, director of NCI's epidemiology and biostatistics program, said Wednesday. They are widely regarded as inconclusive.

For example, about half the residential studies found no correlation between proximity to power lines and cancer rates; about half found an increased risk of between 50 percent and 200 percent. But there are two general ways to estimate the EMF exposure a person may have received at home: rating each home by its distance from a power line and the kind of wiring configuration on the power line (called ``wire codes''); and actual measurements of EMFs in the homes.

The most worrisome positive studies showed cancer incidence correlated with wire codes. But none has ever showed a relationship between leukemia and direct field measurements, ostensibly a more accurate indicator of past EMF exposure.

With puzzlement abounding, the National Academy of Sciences assembled a panel to evaluate existing research. Its report, released in October, concluded that ``the current body of evidence does not show that exposure to these fields presents a human-health hazard.'' But the panel noted that so many studies had indicated a statistical association between estimated residential EMFs and childhood leukemia that the relationship could not be ruled out.

That kind of uncertainty was the motivation for the NCI study. Like nearly all previous efforts, it employed the ``case-control'' method - a system that has been used to link cigarettes with lung cancer and toxic shock syndrome with tampons.

In such studies, researchers identify a large number of people who have the disease of interest (in this case, childhood leukemia). They are the ``cases.'' Investigators then locate an equally large number of people of the same age, sex, race and general living conditions who do not have the disease. They are the ``controls.'' Researchers then try to see if some factor (in this case, estimated exposure to EMFs) is more prevalent in the diseased case population than it is among the controls.

In the new study, researchers from NCI teamed with the Children's Cancer Group, a nationwide network of dozens of centers that collectively treat about half of all U.S. children with cancer. They identified 638 children aged 15 or younger with the most common childhood cancer: acute lymphoblastic leukemia, which strikes about 1,600 children a year in the United States, usually between the ages of 2 and 5.

The team then found 620 controls. For each subject, technicians measured the magnetic fields in one or more homes in which the child had spent most of his or her life. Measurements were taken for hours in each child's sleeping area, and for shorter intervals at other locations around the residence, including the room in which the mother slept while pregnant.

In addition, technicians evaluated the subjects' homes' proximity to power lines of different wiring configurations, and rated each dwelling on the ``wire code'' scales used in prior studies.

After analyzing the data, the researchers found that wire codes were not correlated with the incidence of acute lymphoblastic leukemia. ``Unequivocally, there's nothing there,'' NCI researcher Martha S. Linet said Wednesday.

Nor was there a statistically significant link between leukemia and direct field measurements. Children who lived in the fewer than 10 percent of homes that showed extremely high measured fields (4 to 4.9 milligauss, or mG, a unit of magnetic strength) appeared to have a risk three times normal.

Send Suggestions or Comments to webmaster@scholar.lib.vt.edu