Last month, a jury rejected arguments by two National Security Agency workers who claimed that their brain diseases were caused by a magnetic tape-erasing system they used in the course of their work. The judge dismissed some of the plaintiffs' arguments as junk science and the jury rejected the rest, to the delight of the defendant, Electro-Matic Products Co., manufacturer of the tape-erasing system. This was a victory for science and the legal system, which have both been abused in recent decades by people stoking fear of electric and magnetic fields.
There is no convincing evidence that low-level electric and magnetic fields (EMF), such as those surrounding power lines, have any effect on human health, despite what you may hear from activists, a few researchers, and more than a few journalists.
One of the basic principles of science is that for test results to be meaningful, they should be repeatable. Otherwise, people end up pointing to one or two studies here and there in which the results, possibly just by random chance, appear to support their hypothesis.
Selective reporting of results has clearly occurred in the anti-EMF movement, and that selective reporting reveals a fundamental misunderstanding of the workings of science. Anti-EMF activists will say, for instance, that a Swedish epidemiological study is "the" proof of an EMF-cancer connection, though in no other public health debate is it argued that one study constitutes final proof of a general connection. Epidemiology doesn't work that way. Relying on one unusual study that supports your case is a bit like hypothesizing that flipping a coin tends to result in heads coming up two thirds of the time and then doing many trials until you get a streak in which heads does come up two thirds of the time and pointing to that streak as "proof" of the hypothesis.
Analogous problems seem to exist with each type of test pointed to as supposed evidence for an EMF health threat. Occasionally, isolated animal studies or in vitro experiments will be pointed to as proof, but these experiments are never successfully replicated. A lot can go wrong with delicate measurements of heat and cell size being made in a petri dish. That's why scientists don't normally leap to conclusions based on results from a single in vitro experiment. Their results must be replicable. EMF experimental results that at first sounded frightening have not been successfully repeated.
Some of the studies have been individually flawed as well. The Wertheimer-Leeper study that started the EMF scare in 1979, for instance, was conducted on Denver residents in an area where it had already been noticed that a few clusters of cancer victims existed near power lines. In other words, the study was not a random sampling, only an official head count of a previously-noted cluster (a bit like noticing six out of eight people with baseball caps at a bus stop and then using that abnormal cohort for a survey of how many people, on average, wear baseball caps). Is it a coincidence that the Wertheimer-Leeper study is one of the very few to show a "big" (though still epidemiologically tiny, given the rareness of the illnesses being studied) apparent risk increase of 39%? (See ACSH's Cancer Clusters booklet for more examples of this type of flawed statistical analysis.)
Many reports about EMF danger have made the more complicated mistake of simply counting any ailment in EMF-exposed people as a possible sign of an EMF effect. If the group has slightly more lymph node cancer than average, call lymph node cancer a potential symptom of excess EMF exposure. If the group instead has slightly more miscarriages, call miscarriages a symptom. You're likely to find a higher than average rate of at least one disease in any group of people.
Compounding problems in the field is the fact that new reports showing that there are no health effects from EMF are less likely to be published, especially in journals devoted to the topic, than reports that appear to show effects because studies showing no EMF health effects are already so common. An apparent health effect, perversely, becomes news precisely because such a conclusion is atypical of the research that has been done. So even someone looking conscientiously at the entire body of published evidence may get a skewed picture. Further, people very sympathetic to EMF fears, such as H. Keith Florig of the group Resources for the Future, and Paul Brodeur, whose New Yorker articles did so much to fan EMF fears, tend to discuss the studies that suggest increases in health problems instead of the many that don't.
In epidemiology, a strong correlation would be something like the link between smoking and lung cancer, where the likelihood of getting the disease appears to increase something like twelve or twenty times for those engaged in the risk-increasing activity. By contrast, even the "positive" EMF studies show increases of only a few percent, meaning the ratio of disease cases in the general population to disease cases in the "over-exposed" population is still very close to 1:1 (or, as it is often stated, those studies have a "relative risk ratio" close to 1, while a twenty-fold increase would be called a relative risk ratio of 20).
Even the results from the oft-cited Swedish studies involved only about 13 "extra" cases of childhood leukemia in the entire country, which is a small variation from which to draw any sweeping conclusions. Again and again, the relative risk ratios in the studies pointed to by groups like the Bioelectromagnetic Society hover around 1, just as we would expect if there were no health effects at all. Some 16 childhood leukemia studies have a median relative risk around 1.5 (again, that's tiny compared to the many-times-larger risk increases normally required for epidemiologists to say they see real evidence of a health problem), and childhood leukemia is usually considered by activists to be the clearest example of an EMF health effect.
And, of course, even if there is a real, minute disease increase being detected in these populations, there could be other causes that are merely found in proximity to power lines or in populations who live near power lines. People living near power lines tend to be poorer, to move more frequently (which may affect not only their lifestyles but their likelihood of answering residential surveys about their health), and to breathe in more benzene, a better-substantiated carcinogen from car exhaust, which is found in greater abundance on the main roads usually located near power lines. There are enough confounding statistical factors that a responsible epidemiologist wouldn't point to low-level EMFs as the "culprit" even if the numbers did reflect a minute increased risk.
Occupational studies, such as a Canadian-French study purported to show electrical workers at increased brain cancer risk, have also been based on tiny handfuls of extra cancer cases in small study populations, and the increases observed have not been confirmed in other studies. With some 25 occuptional studies of brain cancer having been done as of the mid-90s, when the case against Electro-Matic was being assembled, the median relative risk was reported to be about 1.2, but it is the occasional "high" studies to which activists point.
In addition, the anti-EMF activists present a great deal of anecdotal evidence, advocacy journalism, and perfectly understandable indignation at power companies for using eminent domain to put frightening power lines on people's property (sometimes reducing their property values), but none of this counts as scientific evidence though the public fear generated by these factors has been great enough to affect public policy debates and result in EMF reduction ordinances in many localities.
As the American Physical Society and others have noted, many of the physical theories put forward in an effort to explain how EMF could cause cancer are implausible, considered fringe science by most people outside the insular community of EMF-health-effects researchers and activists. Elaborate physical mechanisms have to be hypothesized in order to "explain" epidemiological data that is weak to begin with. EMF researcher Dr. Ross Adey and Louis Slesin, editor of the anti-EMF publication Microwave News, have gone so far as to suggest that EMF dangers may be evidence that we stand on the verge of a whole new scientific paradigm, a complete transformation of knowledge in which biological research will be reoriented around the belief that very low-level energies interact with complex, atomic-level webs of signal-receivers running throughout the human body (possibly via subtle motions of calcium atoms as inter-cell signal mechanisms). All this, say the believers, may usher in a new age of sub-molecular biology.
Then again, it may mean that when mainstream physics is against your theory, you resort to rejecting mainstream physics. According to former chairman of the Harvard Physics Department Dr. Richard Wilson, the mainstream science careers of the EMF-danger researchers are essentially over, but they continue trying to reinforce each others' work and promote each others' results.
According to scientists such as radiation and cancer expert John Moulder of University of Wisconsin Medical School, even if you use the activists' figures, the level of EMF in 99% of American homes is hundreds or thousands of times lower than the amount of EMF needed to create even such tiny and benign effects as the movement of especially-reactive, already-magnetized particles found in the human eye.
In the end, as Nobel laureates and respected experts have said repeatedly during the EMF debate, the important thing is to remember that no one study by itself "proves" much of anything. For every favored study the anti-EMF activists point to, there are others they downplay because their results were negative or ambiguous. And, as electromagnetic effects researcher Eleanor Adair of Yale University says, the larger and better controlled the EMF health experiments get, the closer the risk ratios get to 1, which means that more research seems likely to yield less and less reason to fear.
Though there is rarely a "final word" in science, it appears that EMF may be a scapegoat for some of life's randomly-distributed tragedies. Still, one of science's basic strengths is the never-ending search for new evidence. There will always remain the hypothetical possibility that some undetected EMF problem exists, and for a long time to come there may well be people trying to prove it. Sometimes the oddballs and mavericks turn out to be right, and that thought is enough to keep the anti-EMF activists and researchers struggling against the establishment and the accumulating evidence.
May 20, 2002
I suspect the average person who uses a hair dryer has more EMF exposure than bald people, but you don't see anyone calling for a ban on hair dryers.
(I do call for a movement to shave everyone's head, but that is only to get my kids ready for school faster.)
May 23, 2002
Thanks for this fair coverage. I would like to add that we recently collected experimental evidence that an individual's claims about sensitivity to "electrosmog" correlates to a bunch of other irrational claims, ranging from belief in extrasensory communication to more schizophrenia-like ideas of reference. This does not "prove," of course, that electromagnetic fields do not affect us, but it is evidence for a weak causal connection between electromagnetic fields and electromagnetic fears! (See StÅ¡ckenius S. & Brugger P. (2000): "Perceived Electrosenitivity and Magical Ideation." Percept. Motor Skills 90, 899-900.)
May 26, 2002
I followed the EMF controversy since its start with the New Yorker series. It was obvious from the physics that there was minimal or no danger and that the Denver study, though it gave no obvious clue that it was wrong, was not convincing. It was observational, not prospective, made use of calculated rather than measured field strengths, and lacked basic plausibility.
We explored the issue in a AAAS (Pacific Division) annual conference at which physicists and engineers scoffed at the health claims. I had a representative from the Electric Power Research Institute (EPRI) in Palo Alto talk at my course at Stanford Medical School, hoping an authority would demonstrate the implausibility. Surprisingly, he stated that there could be something to the relationship of EMF to cancer. The same speaker stated the same in a JAMA article. Only later did I realize that EPRI subsisted partly on research grants to study such phenomena. Of course, they would not dismiss the theory. Welcome to the political science of science politics.
I was then associate editor of a monthly health newsmagazine and we tried to put together a booklet or special issue on EMF, in which we would present the scientific implausibility of the dangers, debunk the concept with data, physics, and reasoning. I studied all of the studies up to then (1991-2) and was convinced there was no EMF cancer relationship. I even found some interesting inconsistencies in the Swedish study. We needed funding to produce the issue. Where to get it? PG&E, Southern California Edison, and the Association of Independent Local Companies, of course!
All told us that they wanted to minimize the publicity, not make people more aware of the situation. They all had meetings, some national, on the issue and decided on their political positions. PG&E specifically stated that they had to remain neutral on the issue or else the environmental and anti-EMF people would accuse them of sponsoring propaganda to silence the concerned citizenry. Or that they were covering up the true information.
Not only would they be neutral, they would acquiesce to, even encourage "more studies of the issue" so as to appear cooperative and to minimize lawsuits and threats of terror bombings and all. These things were told to us "in confidence" and none of them was willing to put it on paper.
An electrical engineer friend and I testified at a hearing on whether to move half of a school's classrooms that were close to power lines the "Santa Clara Schools case" that made national news. The PG&E representative there advised the "prudent avoidance" rule: move the classrooms or the students. They moved the students out.
An estimated cost of $14 billion or so over ten years was incurred as U.S. power companies were forced to move offending power lines (sometimes placing them underground) and as fear lowered property values across the U.S.
We never could get financing for the special issue.
Wallace Sampson, M.D.
October 10, 2002
Enjoyed your "Editor's Rant" on EMF and health.
I'm mayor of a small town in which the utility successfully sued after being denied a permit to reconstruct a transmission line, based for the most part on EMF hysteria. The court blasted the city council's denial of a permit as "unjustified, arbitrary and capricious." I think this is the first case in Minnesota to actually go to court in which one side alleged that just the possibility of EMF being harmful is sufficient reason to deny a permit.
What amazes me is that well-educated people forget what they learned in Logic 101: the fact that A is associated with B doesn't necessarily mean that A causes B. Yet we hear "my wife has breast cancer and we live near the powerline, so obviously EMF caused her cancer, since I've heard that EMF causes cancer."