The Science and Politics of the EMF Puzzle;
The Missing Pieces in the Frontline Story
The irony is astonishing. On the very day that a committee of the National Council on Radiation Protection and Measurements (NCRP) completed its 800-page draft report asking regulatory agencies to pay “serious attention” to EMFs, public television station WGBH aired a one-hour show across the country comparing EMFs to cold fusion. While the NCRP committee called for “a national commitment to further research,” the June 13 Frontline, “Currents of Fear,” asked whether it was time to close down the research effort.
Of course, Frontline dominated the gossip circles at the annual meeting of the Bioelectromagnetics Society (BEMS) held later that same week in Boston, WGBH’s hometown. Only a few insiders knew about the NCRP report.
It's too bad that Jon Palfreman, WGBH’s producer of this show, did not bother to go across town to attend the BEMS meeting. If he had listened to some of the presentations and talked to members of the NCRP committee, he might have realized that his documentary for the Frontline series left out key facts—facts that conflict with his thesis that concern over EMF health effects has no scientific basis.
Then again, it might not have made a difference. There are good reasons to think that Palfreman never approached the EMF puzzle with an open mind. Much like Gary Taubes, whose own attack on EMFs appeared in last November's Atlantic Monthly, Palfreman started with an idée fixe and then went looking for like-minded people.
In a nutshell, Palfreman’s thesis is that the animal studies show no effects, the cellular experiments are irreproducible, the best epidemiological evidence is specious and the postulated biophysical mechanisms of interaction contradict the laws of physics. Unfortunately, it isn’t quite that simple.
* * *
Let's begin with what Palfreman left out:
Palfreman only cited those experiments being carried out at Chicago's IIT Research Institute (IITRI) by Dr. David McCormick, whose first cancer study failed to show any effect. McCormick's results are so new that they have not yet been peer-reviewed or published. Even so, they already have their share of critics, who have raised questions about the study design.
Palfreman built up IITRI’s exposure study as the definitive, best-controlled experiment ever. He let Dr. Gary Boorman of the National Institute of Environmental Health Sciences (NIEHS) make the case:
As you refine your studies, if there really is an effect, the effect should increase, it should become stronger, it should become more focused, and if you cannot, with repeated studies and with better studies, you continue not to find an effect or find only marginal effects, then it becomes obvious that there's really nothing there.
Boorman's scientific logic is of course rational and correct, but, in this particular context, it makes no sense. Very few animal cancer studies have been done to date and McCormick has only completed a single experiment. It's hard to make a case for a trend with IITRI's one data point.
In a press release posted on the Internet, the WGBH public affairs office lost all control of the facts:
Dozens of animal experiments have been carried out in which rats and mice are exposed to very large magnetic fields for long periods —some for their entire lives— but no animal has ever been proven to contract cancer due to this exposure.
If you add up all the animal–cancer studies ever done, you do not reach dozens” and some of those that have been done show some adverse effects. What did Palfreman say when asked about this? Only that he did not write the press release. No apologies, no excuses, no interest in setting the record straight.
At BEMS, both Dr. Craig Byus1 of the University of California, Riverside, and a Finnish team2 reported seeing higher-than-expected rates of cancer in animals exposed to EMFs. For one set of exposures, Byus said that he found a “very, very significant” increase in tumor incidence. Byus’s study is being sponsored by the NIEHS, where Boorman works.
But the most riveting talk on animal experiments was that given by Dr. Wolfgang Löscher of the School of Veterinary Medicine in Hannover, Germany. He and Dr. Meike Mevissen have completed a series of studies at four different exposure levels—much of which has already been published in respected peer-reviewed journals.3 In Boston, Löscher concluded that the magnetic field “promotes the growth and increases the incidence of tumors in a dose-dependent fashion.”
Boorman said that he is impressed by Löscher’s animal studies not only because they show clear and reproducible effects but because Löscher also has confirming experimental evidence on the hormone melatonin and the growth enzyme ODC. It all adds up to a consistent and solid picture of EMF effects on a living organism.
Indeed, Boorman is so impressed with Löscher's work that he recently issued a request for proposals to repeat the animal studies at a cost well in excess of $1 million. Does this sound like a scientist who thinks that there is “nothing there” as Palfreman would have us believe? The Department of Energy (DOE) is so favorably impressed by Löscher that it will sponsor his work directly—one of the very few times the DOE has ever funded EMF research outside the U.S.
“The animal studies were incompletely and too simply presented” on Frontline, Boorman said in an interview. If Palfreman had used Löscher’s studies as his example instead of IITRI’s, he said, the audience would have reached a “very different conclusion.”
Here again, Palfreman based his argument on one study—Battelle’s Dr. Jeffrey Saffer’s unsuccessful attempts to repeat Drs. Reba Goodman and Ann Henderson's experiments showing changes in gene expression in HL-60 cells. At this point, no one knows why the two (three, counting a British team that also found no response) labs got different results. But looking beyond this particular experiment, another picture emerges.
A number of researchers have shown that EMFs can affect gene expression. Among them is Saffer himself. At last November's EMF review in Albuquerque, NM, Saffer reported that he had turned his attention to another cell line, JB6, and found preliminary evidence of a response.4 Saffer later parlayed that finding into a research grant of more than $1 million from the NIEHS.
As it turns out, at BEMS, Saffer said that so far he has been unable to reproduce the effect.5 Nevertheless, another lab, at the Food and Drug Administration, has been seeing a robust response in JB6 cells.6
When asked about the JB6 work, Palfreman said that he knew nothing about it and that Saffer had never mentioned it. Saffer must have gone through a TV interview —never a short affair— without a word about his new million-dollar project on gene expression.
One of the best-known scientists at the BEMS meeting was struck by this omission: "If Saffer does not think there is anything there and doesn't have the scientific imagination to know where to look, why doesn't he give the money back?" he asked, expressing his own frustration on how hard research money is to find.
Most of the data supporting a cancer risk comes from studies of human populations, and many leading epidemiologists —including Drs. Anders Ahlbom, Birgitta Floderus, Sam Milham and Gilles Thériault— have found strong evidence for a link. They are all internationally known and respected and each has led large EMF studies.
Palfreman is unconvinced. He says that “most epidemiologists” regard the EMF field “as something of an embarrassment to their profession,” but does not cite any sources. If he were right, why couldn’t he find a single epidemiologist other than a longtime utility consultant to say this to the camera? And why are so many epidemiologists still working on the EMF puzzle if it so stigmatizes them?
Palfreman's experts on epidemiology were two electric utility consultants: Dr. John Moulder of the Medical College of Wisconsin in Milwaukee, an expert on using radiation to treat cancer, and, to a lesser extent, Dr. Patricia Buffler, an epidemiologist at the University of California, Berkeley. He used them to disparage the Swedish childhood cancer study by Drs. Maria Feychting and Anders Ahlbom. Moulder said that the Swedes made a fundamental error: They made so many comparisons that, by chance alone, they had to come up with some positive associations. Buffler agreed. The Swedish study, they argued, was simply an exercise in data dredging.
Ahlbom dismissed Moulder's and Buffler’s criticisms. “It is absolutely necessary to look at a large number of analyses in any epidemiological study to look for consistencies and inconsistencies in the data,” he said in an interview.
In a review7 published earlier this year, Feychting and Ahlbom concluded, “The evidence on leukemia in children appears rather consistent.” They went on to say that the hypothesis that EMFs lead to the development of cancer cannot yet be considered proven and that we “have to accept the uncertainty."
The Frontline interview with Dr. David Savitz made it clear that Palfreman is not interested in uncertainties. After talking to Savitz for more than an hour on camera, Palfreman only gave him some 20-30 seconds of airtime —and even then Savitz never got to say a word about epidemiology, only exposure assessment. Savitz thinks he knows why he was not allowed to say more: “Palfreman had his point of view and looked for quotes to support it,” he said in an interview.
The data-dredging argument is “completely unsatisfying —it’s no explanation at all,” Savitz said, adding that, “It is a strength of the study to have investigated so many possibilities.” As Ahlbom pointed out, “Suppose that we had not done this, but had been asked to do so after publication, should we have answered: ‘This is an interesting question, but unfortunately we cannot do this because it is not in the study protocol’?”
The contribution of the Swedish study is not that it gave a definitive answer, but that it provided a new and persuasive piece of the EMF puzzle that fits neatly into the existing mosaic. But Palfreman has no patience for anything short of absolute proof.
The earlier Wertheimer–Leeper, Savitz and London–Peters studies all showed a link between childhood cancer and EMF exposures, as defined by the Wertheimer-Leeper wire codes. When measured fields were used, however, the link was weaker. Some observers, especially those from the utility industry, jumped on this apparent discrepancy and dismissed the studies as being internally inconsistent.
Using historical records of the current loads on Sweden’s power lines, Feychting and Ahlbom calculated the magnetic fields when a child got sick-as well as one, five and ten years prior to diagnosis. (This partly explains why they made so many comparisons, which led to the charges of data dredging.) The Swedes found that the calculated historical fields did show a link to cancer. Making their case even stronger, they found a dose-response relationship. Further, like their predecessors, Feychting and Ahlbom did not see an association between present-day magnetic fields and leukemia. This suggested, as many epidemiologists had long speculated, that wire codes are good indicators of long-term EMF exposures.
In short, the Swedes set out to see whether there was a link between power line magnetic fields and leukemia and that is exactly what they found. They improved on past studies and the link grew stronger: precisely the type of evidence that skeptics say they want to see. Feychting and Ahlbom did not answer all the questions, but epidemiology never does.
Biophysical Mechanisms of Interaction
This is the most important part of Palfreman’s argument. Whether an experiment shows an EMF effect in humans, animals or cells becomes moot if it is possible to show that such interactions are theoretically impossible: Yale University physicists Drs. Robert Adair and William Bennett believe this, and, it appears, so does Palfreman. To use the metaphor conjured up by Adair on Frontline, worrying about EMF health effects is akin to being concerned that a cat will damage a tree by breathing on it during a howling wind storm.
Given the recent statement8 by the American Physical Society (APS) that EMFs are of no concern —also cited by Palfreman on the show— one might concludethat all physicists agree with Adair and Bennett. But that would be a mistake.
There are many physicists working in the field of bioelectromagnetics. As Dr. Bill Kaune, a consultant based in Richland, WA, who has a doctorate in physics, put it: “We physicists who do research on EMFs have long been aware of the signal-to-noise problem, but, regardless of our concerns, experiments seem to show that EMFs affect living tissues. I don't see how one can justify flatly discounting the work of a large number of epidemiologists and laboratory biologists solely on the basis of signal-to-noise calculations on highly simplified models of living tissues.”
A couple of years ago, Adair had the opportunity to make his case to the JASONs, a high-level group of physicists, whose advice is routinely sought by the Department of Defense. In his report on behalf of the JASONs, Dr. Steven Koonin of Cal-tech concluded: “The essential point to take away...is that a cellular-level coupling of magnetic fields to biological systems is physically plausible and does not violate any physical principles.”9
Koonin was a member of the APS council that approved the statement, and may well believe that "no plausible biophysical mechanisms" have been identified. But this does not mean, as Adair and Bennett (and Palfreman) contend, that such interactions are impossible.
As for the APS statement itself, it is as much a political as a scientific document. A look at Dr. David Hafemeister's slipshod background paper10 that served as the basis for the statement is convincing evidence of his and the APS' political agenda. Hafemeister is well attuned to the world of politics and the power of a press release given to a responsive reporter, having spent many years in Washington working for the federal government and congressional committees.
Adair was at the BEMS meeting, and when asked how he could explain an organism’s ability to sense magnetic fields as weak as 0.2 mG against the background of the earth's 500 mG field, Adair replied, “Perhaps if you have enough cats and enough trees....”
Or to put it another way, biological systems are complex and are not easily captured by a simple model or a colorful analogy.
Implicit in these discussions is that when we talk about EMFs we are referring to one physical phenomenon. In fact, there are a huge variety of EMFs, each of which may have a different effect. Among the most intriguing are transients-short intense pulses of energy.
At BEMS, Dr. Antonio Sastre, a consultant based in Suffern, NY, and his colleagues showed that when it comes to EMF transients —common occurrences on power lines— the signal can rise above the background noise.11 “The objection that environmental fields are too weak with respect to thermal noise need not apply to transients,” Sastre said in an interview, pointing out that one needs to invoke only “pedestrian physics applied to realistic models of cells" to show this. Sastre's work is sponsored by the Electric Power Research Institute (EPRI).
Sastre’s work on transients follows last fall's announcement by Thériault that he and his colleagues at McGill University in Montréal, Canada, had found a very strong association between exposure to transients and lung cancer among utility workers. Hydro-Québec's first reaction on learning of this result was to take the data away from the McGill research group.12 Seven months later, the conflict remains unresolved and Thériault's team is still barred from probing further into this risk-one of the largest ever observed for any EMF-exposed population.
* * *
Palfreman is a serious journalist. He has won two AAAS science writing prizes, as he is not too bashful to tell his critics. But he clearly came to EMFs with his mind made up. He might have salvaged the show had he taken the trouble to talk to those whose work he is disputing.
Palfreman never interviewed Drs. Anders Ahlbom or Reba Goodman, two of the scientists he skewered on the program. He said that he exchanged faxes with Ahlbom and has the faxes to prove it. This made no impression on Ahlbom, who cannot remember Palfreman among the many reporters who have called him.
The Goodman story is different and more troubling. Palfreman said that Goodman refused to be interviewed, while Goodman maintained that no one from Frontline ever called her. Goodman is right. Palfreman let one of his assistants, Michaela Barnes, contact Goodman, but she conceded that she never did. When asked why not, Barnes cited "political reasons."
One set of Frontline interviews, left on the cutting room floor, involved the controversy over health problems among those living next to an electrical substation in Guilford, CT. Paul Brodeur used this as a case study in a 1990 article published in The New Yorker,13 with Bob Hemstock serving as the protagonist. In the course of his interview with Frontline, Hemstock offered to contact Goodman on the program's behalf, since he was going to be talking to Goodman about a project of his own.
Goodman told Hemstock that it would not be a good idea for him to visit her lab with a television crew. Goodman was swayed, at least in part, by Hemstock's theories on EMFs, which are somewhat idiosyncratic. In any case, by the time Hemstock relayed this message to Barnes, he had himself decided, for his own reasons, that he did not want to take the Frontline crew to Goodman’s lab.
Hemstock never knew and therefore could not tell Goodman about Frontline's interest in Saffer’s experiment. Nor did Goodman know that the public television crew Hemstock mentioned was from Frontline.
No one from Frontline ever called Goodman directly —even though Palfreman planned to tell the world that her results are worthless. Neither Palfreman nor anyone else from Frontline was interested in hearing Goodman's side of the gene expression story— a violation of one of the most basic rules of journalism. The omission is striking given Palfreman’s statement that he is “interested in finding the truth in a world where, increasingly, scientific data is being abused and distorted for political ends.”
* * *
One of the most revealing aspects of the Frontline episode is that so few members of the bioelectromagnetics community have spoken out to correct the obvious errors in the program. Palfreman said that the vast majority —approximately 95%— of the comments have been favorable. The reason for the silence is an important part of the EMF story.
EMF research is an underfunded backwater of the scientific community. Before the congressionally mandated $65 million RAPID program got under way last year, most of the available research funds came from the electric utility industry through EPRI and from the DOE, an agency not known for putting radiation safety ahead of its other program objectives. EPRI and the DOE do not look kindly on those who publicly highlight possible health risks.
This is the grubby side of science, where many researchers are as interested in securing contracts and grants —even if it means making compromises along the way— as they are in doing the actual scientific work.
This also explains why there has been no outcry —indeed we have yet to hear a single word of public protest— at Hydro-Québec’s outrageous behavior in blocking Thériault's access to millions of dollars worth of data that could explain part of the EMF puzzle.
Among the possible casualties of this silence are the emerging biomedical applications of EMFs, for instance the use of pulsed EMFs to heal nonunion fractures. If low-level, non-thermal effects fall into disrepute, then, by definition, such medical devices are ineffective and should not be on the market. But even those who see a bright future for EMF therapies are not speaking up.
Before the New York Power Lines Project began in the early 1980s, neither EPRI nor the DOE moved to confirm or refute the Wertheimer-Leeper cancer study. As a result, it took nine years to repeat it. And as soon as Savitz had done so, EPRI, still unconvinced, decided that it had to be done all over again. This took another five years. Critics of the Feychting–Ahlbom study should not wonder why we don't know more, but rather should marvel that we know as much as we do.
When the Feychting–Ahlbom results were released in 1992, the Swedish government said that it believed that, more likely than not, the cancer link existed.14 Two years later, the Swedes did not reject this conclusion, but decided that they could not rationalize the high economic costs of regulating EMFs given the rarity of childhood leukemia. Swedish government officials have made it clear that if the hypothesized EMF role in breast cancer were to be confirmed they would reconsider the decision.15 Feychting and Ahlbom are now working on an epidemiological study of EMFs and breast cancer.
The significance of the epidemiological studies is not that they point to a cancer epidemic. But they raise the question: If EMFs can cause even a small change in cancer rates, what other biological effects could they have?
In the absence of detailed studies on breast cancer, Alzheimer's disease and depression, among other common health problems, no one knows how great the EMF health risk really is. Those who argue that we now have enough research to conclude that the risk is small, if it exists at all —as Dr. Jack Sahl of Southern California Edison does in a recent report(16)— are engaging in wishful thinking.
The reason the EMF problem has attracted so much attention is not because of pressure from the scientific community. It is the public that has propelled EMFs into the limelight. The Omaha housewives whose children have cancer want answers, as was shown on Frontline. Palfreman portrayed them as naifs who have been brainwashed by Paul Brodeur. This is unfair because they have legitimate concerns and because they are victims of the scientific uncertainty that is a result, in large measure, of years of industry and government foot-dragging.
* * *
So, the animal, cellular and human studies all point to real risks. And physics does not put them out of the realm of possibility. To be sure, these risks have not been conclusively proven —but neither have they been convincingly dismissed.
As the NCRP committee concluded in its draft report: “[F]indings are sufficiently consistent and form a sufficiently coherent picture to suggest plausible connections between ELF EMF exposures and disruption of normal biological processes, in ways meriting detailed examination of potential implications in human health.”
Yet Palfreman is sure he already has the answers. After a quick breeze through the literature and talking to a few like-minded scientists, Palfreman thinks he knows better than the expert NCRP committee that studied the issue for a decade. “The thesis is mine," he said in an interview. “It’s very clear-cut. I don’t feel any doubt.”
Palfreman’s show was simply an exercise in hubris. Only this can explain why Palfreman is willing to follow in Buffler’s footsteps in waging a holy war against prudent avoidance. Neither sees the point, for instance, in telling parents that they may be protecting their kids simply by moving a bed across a room out of a 10 mG field. As Julie Larm, one of the mothers on the show, wrote to Palfreman on behalf of Omaha Parents for the Prevention of Cancer after the June 13 broadcast, “May God help you if you're wrong.”
(1) C.V. Byus, Y. Ma and M.A. Stuchly, “The Ability of Magnetic Fields To Serve as a Co-Promotional Stimulus to the Development of Papillomas on the Skin of the Mouse,” Paper No.18-3, 17th Annual Meeting of the Bioelectromagnetics Society (BEMS), Boston, June 18-22, 1995.
(2) T. Kumlin et al., “A Study of the Possible Cancer-Promoting Effects of 50 Hz Magnetic Fields on UV-Initiated Skin Tumors in ODC-Transgenic Mice,” Paper No. P-126C, BEMS, 1995.
(3) W. Löscher et al., “Tumor Promotion in a Breast Cancer Model by Exposure to a Weak Alternating Magnetic Field,” Cancer Letters, 71, pp. 75-81, 1993; W. Löscher et al., “Effects of Weak Alternating Magnetic Fields on Nocturnal Melatonin Production and Mammary Carcinogenesis in Rats,” Oncology, 51, pp.288-295, 1994; W. Löscher and M. Mevissen, “Animal Studies on the Role of 50/60 Hertz Magnetic Fields in Carcinogenesis,” Life Sciences, 54, pp. 1531-1543, 1994; M. A. Baum et al., “A Histopathological Study on Alterations in DMBA-Induced Mammary Carcinogenesis in Rats with 50 Hz, 100 µT Magnetic Field Exposure,” Carcinogenesis,16, pp.119-125, 1995; M. Mevissen, M. Kietzmann and W. Löscher, “In vivo Exposure of Rats to a Weak Alternating Magnetic Field Increases Ornithine Decarboxylase Activity in the Mammary Gland by a Similar Extent as the Carcinogen DMBA,” Cancer Letters, 90, pp. 207-214, 1995. See also MWN, J/A93, S/O94, J/F95 and M/A95.
(4) J.D. Saffer, S.J. Thurston and N.H. Colburn, “Carcinogenesis in Weak Electromagnetic Fields,” Paper No. A-14, Annual Review of Research on Biological Effects of Electric and Magnetic Fields from the Generation, Delivery and Use of Electricity (DOE), Albuquerque, NM, November 6-10, 1994.
(5) J.D. Saffer, S.J. Thurston and N.H. Colburn, “Tumor Promotion in JB6 Cells by Weak Electromagnetic Fields,” Paper No. 1-5, BEMS, 1995.
(6) R.W. West et al., “Enhancement of Anchorage—Independent Growth in JB6 Cells Exposed to 60 Hz Magnetic Fields,” Bioelectrochemistry and Bioenergetics, 34, pp. 39-43, 1994. See also MWN, J/F95.
(7) Maria Feychting and Anders Ahlbom, “Childhood Leukemia and Residential Exposure to Weak Extremely Low Frequency Magnetic Fields," Environmental Health Perspectives, Supplement 2, pp.59-62, 1995.
(8) See MWN, M/J95.
(9) Emphasis added. See MWN, S/O93.
(10) D. Hafemeister, Background Paper on Power Line Fields and Public Health, Washington: American Physical Society, May 1995.
(11) A. Sastre et al., “Residential Magnetic Field Transients: How Do Their Induced Transmembrane Voltages Compare to Thermal Noise?” Paper No. A-33, DOE, 1994; and G.B. Johnson, R. Kavet and A. Sastre, “Residential Magnetic Field Transients: Effect of Residential Services on Fields Arising from Distribution Line Capacitor Bank Switching,” Paper No. P-130A, BEMS, 1995.
(12) See MWN,N/D94.
(13) P. Brodeur, “Annals of Radiation: Calamity on Meadow Street,” The New Yorker, pp. 38-72, July 9, 1990; reprinted in P. Brodeur, The Great Power Line Cover-Up, Boston: Little Brown and Co., 1993.
(14) See MWN, S/O92.
(15) See MWN, M/J94.
(16) J.D. Sahl and B.S. Murdock, Electric and Magnetic Fields and Human Health: A Review of the Issues and the Science, Azusa, CA: Southern California Edison, April 1995.