 |
 |
|
|
|
|
|
|
||||||||||||||||||
|
|
|||||||||||||||||||
|
|
|||||||||||||||||||
|
|||||
|
|
By Greg Lapin, N9GL, Chairman
ARRL RF Safety Committee
April 18, 2000
In the past two columns, I've presented the theories regarding the formation of cancer and have discussed one way that RF energy cannot fit these theories and another way that it can. I left you hanging last month with a question: "But does it?" This month, we'll look at how science has tried to answer this question. Along the way, the oft-heard phrase "Bad Science" comes up.
It is unfortunate that Bad Science exists. Not just the science, but the phrase itself.
It is all too easy to dismiss useful scientific results as Bad Science and then to ignore them. This happens on both sides of the fence: People who are convinced that RF is totally safe (a group that contains a lot of hams) often label any scientific study that says otherwise as Bad Science. The people who are convinced that RF must be unsafe are quick to label any scientific study that does not agree as Bad Science. I, along with many of my scientific colleagues, sit on the fence. From the science that has been performed to date and analyzed for its inadequacies, it is clear to me that RF can be terribly dangerous, but it can also be used in a manner that is safe.
When I hear the frustration of many of the non-scientists that I speak with regarding the mixed messages that they are hearing about RF safety, the Thomas Dolby song title from the 1980s comes to mind. The difficulty of distinguishing valuable studies from those that are fraught with errors, coupled with what (and how) the press chooses to present to the public, makes it difficult to feel secure with all of the RF that surrounds us.
There is so much detail in the science used to study these effects that it is not possible to cover it all in one sitting. This month I intend to introduce the different types of science that are used to study RF effects and then to follow that up in successive months with some more detail about each of these and what it has shown us to date.
Health effects from environmental factors often are the result of a lifetime of exposure. The things that quickly affect our health already are known, and we tend to avoid them (such as getting hit by a car, drinking poison and getting struck by lightning). What's left are things that take years and decades for their results to be seen. These include smoking, high cholesterol diets, and exposure to X-rays. Exposure to athermal levels of RF also will fall into this category, if any dangers to health ever are proven.
Scientific study has a built-in problem with proving that something is safe; investigation is based on the null hypothesis, which first assumes something will not happen and then tries to disprove that assumption. If a scientific experiment is able to disprove the null hypothesis, it is generally accepted that the effect exists. However, the failure to disprove the null hypothesis does not prove it.
Thus, we can study RF for years without finding an athermal effect, but this does not prove that one does not exist. Someone can always point to this fact and demand more study. This is one of the reasons that the issue has stretched out for decades.
There are two general ways to study long term effects. Both have basic problems and, as a result, are rarely conclusive. Ideally, we could take a laboratory animal, expose it to the environmental factor in question, and see what happens. The problems with this are that laboratory animals have very limited lifetimes and cannot receive an identical form of exposure to what we would like to study.
For instance, a laboratory rat rarely lives longer than a year or two. How can you study a rat for a year to determine an effect that may take more than 20 years to be seen in humans? One technique uses increased dosage in place of increased time of exposure. As discussed in past columns, this often gives incorrect results, such as the saccharin scare of the 1970s. With RF exposure, the problem is even more pronounced. An increased exposure to RF causes heating in tissue that does not exist at lower levels. It is not possible to study athermal effects this way.
We can't study humans in the laboratory, but we can study them in their habitat. This has the advantage of allowing us to look at lifelong exposures. It has the disadvantage of not being controlled, as is a laboratory experiment. The science of epidemiology uses statistics to look for apparent connections between causes and effects. For instance, if you know five hams who got cancer, does it follow that that ham radio causes cancer? No, it doesn't. Since everyone gets sick, it is not enough to observe a group with a high incidence of a given disease and then assume that a common trait of that group causes the disease.
Statistics are extremely powerful and can be dangerous. It is very easy to use numbers to show just about anything. As an extreme, somewhat absurd, example, if I did a study of people who died of cancer and found that 100% of them breathed air, I could correctly state that there is a high correlation between breathing air and dying of cancer.
The first thing that an epidemiologist needs to do is compare the rates of a disease in the specific group to the general population (and since 100% of the general population also breathes air, cause and effect is not supported by my prior correlation). However, even this does not tell the whole story. If a difference is found, it could be due to a common trait of the group, but it also might be another, less evident factor. I'll discuss more about epidemiology in the coming months. For now, I'll state that as powerful as epidemiology can be, it can hardly ever (some say, never) be used as proof that a certain factor causes a disease. Epidemiology is very useful for suggesting possible links between cause and effect. However, Good Science then dictates that the suspected correlation be taken into the laboratory to find the mechanism that positively links cause to effect.
Bad Science is out there, but it is often hard to recognize. It happens for the usual reasons: carelessness, preconceived notions, incompetence, dishonesty and greed. To a non-scientist, the claims made by Bad Science can be convincing, and often scary. It takes considerable expertise and knowledge to recognize Bad Science for what it is.
The scientific community has built up a number of methods and procedures to recognize and discredit Bad Science before allowing it to be published. The foremost method is peer review. Before a research study can be published in a respectable scientific journal, it must undergo review by several experts in the field. The study must successfully stand up to their criticisms related to correct recitation of accepted facts, consideration of prior related studies, experimental technique and analysis methods before it appears in print. (Even these columns are peer reviewed by the ARRL RF Safety Committee). Although the system is not perfect, and some Bad Science does get through, the vast majority of peer-reviewed work that is published is of high quality and can be trusted.
An adjunct to Bad Science is Bad Reporting. Quite a bit of Good Science is recognized as such only by fellow scientists and is often misinterpreted by the general public. This can be the fault of the media, when the results of a scientific study are reported without placing them into the proper perspective.
Consider the different motivations of the scientific press and the popular press. The scientific press strives to add to the base of knowledge about the world around us. The popular press often strives for increased readership, sometimes with stories that are designed to shock people or otherwise grab their attention. Even though the media typically subscribe to certain standards of accuracy, misrepresentation of accurate facts occurs.
My favorite ham radio example of Good Science Gone Bad is an epidemiological study performed on hams in the mid-1980s by Samuel Milham. Milham was an epidemiologist with the State of Washington Department of Public Health. He used QST's "Silent Keys" column to search for call signs from the sixth district. He then looked up those names in the State of California Death Records (California is known to compile among the best health and death records in the nation). He performed the statistics and came up with a number of conclusions, one of which was that his study group had a statistically significant increase in the rate of deaths from leukemia when compared to the expected rate in the general population.
This work was peer-reviewed and published in the American Journal of Epidemiology, a respected publication in the field. An epidemiologist reading this work would recognize it as a preliminary study into the correlation between Amateur Radio and disease. It was preliminary because little effort was made to determine many of the important variables that would affect such correlation, such as what types of exposures the test subjects had (just because someone has a ham license doesn't mean he or she uses it) and what other things the test subjects were exposed to (such as cigarettes, solder fumes or PCB-based transformer oil). Such preliminary studies often are performed to see if the quick and inexpensively obtained results warrant further study. No conclusions should be drawn from this type of study.
One of my favorite statistics from Milham's paper was that the average age of death of the ham population that he studied was 10 years older than the average age of death of the general population, a statistic that has no more meaning than all the others in the study (it could be an indication that the hams were sitting inside using their radios and not outside getting run over by buses). The results regarding leukemia were somehow leaked to the general public, which eventually came to believe that ham radio causes leukemia.
Several years ago at a scientific meeting on RF bioeffects, this statement was made by one of the participants. As the chairman of the session, I took the opportunity to challenge the statement, only to find out that it was something that this person had heard as accepted fact; he didn't even know the source of the original study.
How do you recognize Bad Science? It's not easy. If a scientific study is published in the popular press before it is published in the scientific press, or a scientific study is published in a journal that does not specialize in the field, I question it. When the scientific study fails to reference and discuss previous related studies, it is suspect. Experimental methods must be verified. It is usually not sufficient for someone to develop a new experimental method and use that method to make conclusions about biological effects. This kind of work must be done in two stages; first the new method must be tested and presented to the scientific community for comments and evaluation. Only after it is accepted should it be used to test biology.
One of the most common errors found in RF bioeffects laboratory studies is the use of an incorrect dose of RF energy. It is rarely valid to expose a small animal to the same RF power density as a human. Corrections must be made for the resonances due to the different sizes of the test subjects. Scientific studies that don't take this into account generally cannot be fully trusted. However, even if these results cannot be used directly, they still show something about biological effects to different exposure levels of RF and could be valuable if viewed in the correct light.
In a highly complex field such as RF bioeffects, often there are confounding factors that are unforeseen by the investigators. It is generally accepted that an independent laboratory using the same methods must confirm any study showing an adverse effect. In many cases, the effects can only be seen in the original laboratory, and their methods must be questioned.
I have refrained from citing specific scientific studies that I consider Bad Science. There are several that have been publicized lately. After you read this, I hope that you will become more discerning about which news items you believe. Remember that RF Safety has been studied for more than 40 years, and it is unlikely that an unexpected breakthrough will occur; science tends to crawl by inches rather than jump forward by leaps and bounds. The mountain of research has been examined in light of current experimental knowledge resulting in safety standards (ANSI/IEEE C95.1-1992 and NCRP Report 86 [1986]). The FCC has adopted these standards in its regulations. As long as you continue to follow the FCC regulations, you are operating safely, according to what the majority of the scientific world believes.
Editor's note: First licensed in 1969 at age 13 as WN1NUK in Connecticut, Greg Lapin, N9GL, went on to earn a PhD in electrical engineering from Northwestern University. He started working in the RF Safety world after spending many years first studying cardiac function imaging and then brain tumor kinetics. He is currently chairman of the ARRL RF Safety Committee and a member of the IEEE Committee on Man and Radiation. A former professor of Biomedical Engineering and Neurology at Northwestern, he now works as a consulting professional engineer in the electronics industry. Lapin professes to still be fascinated by virtually all aspects of Amateur Radio. One of his many interests is electronic design, and he is the author of Chapter 8, "Analog Signal Theory and Components" in The ARRL Handbook for Radio Amateurs. His non-ham interests include making things grow in his garden and serving as commissioner of the local children's softball league. At other times when he is not working or doing his kids' homework, you might find him with the local emergency services agency, climbing his tower, building a new QRP rig, playing with his APRS setup, responding to QSL cards, going off on a DXPedition, or trying to get that new one. Readers may contact Greg Lapin at g.lapin@ieee.org.
