Epidemiology, the study of negative health effects, is a thorny subject. Because it’s very scary to think of things secretly influencing our health, the public often responds to epidemiological studies with fear, without any attempt to place that study in its proper context. One epidemiological study showing a correlation proves very little. The so-called “Bradford Criteria” lists what you generally need in order to prove something is causing death or disease in epidemiology. These include strength, consistency, specificity, temporality, biological gradient, plausibility, coherence, experiment, and analogy. Without these standards, you can make anything appear harmful.
Let’s test these criteria on an example. What about cell phones and cancer? To be sure, you can find papers out there claiming to link cell phones to brain cancer. Some of them seem to have a statistically strong correlation, which satisfies the first requirement. (1) Some people see this, and immediately go out to buy one of those nifty cell phone pads to protect them from the deadly radiation. You can hardly blame them. Fear is the mind killer, and when a scientific paper tells you to be afraid, it can be very hard to think or investigate further.
So let’s take a look at a few of the other Bradford criteria. What about consistency? Well, that’s where it all starts to fall apart. There has been study after study finding no correlation between cancer and RF-EMF frequency radiation. The same is true for the electromagnetic fields generated by other electronic devices.
At worst, an increased risk of childhood leukemia at exposures above a time-weighted average of 0.4 microteslas is supported by some evidence. However, this is a fairly high level of EMF exposure. It is possible to remain well below this level of exposure while still using plenty of electronics, cell phones included. It is also the strongest correlation that can be found at this time, with other correlations appearing much weaker. Other types of cancer or health problems do not seem to have a consistent correlation with EMF exposure. (2) Another one of the Bradford criteria, biological gradient, says that the incidence of disease should be more-or-less proportionally higher with higher levels of exposure. As the review cited explains, some of the evidence for EMF hazard fails this standard as well.
As for plausibility, that’s even worse. It’s true that EMF can produce electrical currents in the body, which is one supposed mechanism for them to cause damage. However, the natural electric currents produced in your body by nerve signals can be much stronger than anything produced in the body by electronics. So are the electric fields we are exposed to from the Earth itself. Because of this, the International Agency for Research on Cancer (IARC) goes so far as to say that there is “no scientific explanation” established for the association between >0.4 microtesla exposure and childhood leukemia, and suggests that it may be due to selection bias.(4)(5) This is not an unreasonable conclusion, since the mechanism for EMF-caused cancer is practically nonexistent. Furthermore the IARC points out that even >0.4 microtesla EMF seems to have no effect on childhood brain cancer, solid tumors, or adult cancers of any kind. This is a problem, because not only do we need a mechanism for EMF to cause cancer, we also need a mechanism by which it would cause only one type of cancer. So perhaps the selection bias explanation is not so far-fetched.
Hold on though, you might say. Everyone knows radiation causes cancer, and RF-EMF from cell phones and other electric appliances is radiation, isn’t it? Well, sort of. RF-EMF radiation is non-ionizing, which by definition means it lacks the energy necessary to remove an electron from its atom. (6) It cannot therefore directly damage DNA, the way ionizing radiation (e.g. a gamma ray) does. Yet as I have pointed out in my post on energy generation, even a fair amount of ionizing radiation is completely natural. The potassium-40 in the ocean, in bananas, and in your own body, emits gamma rays. In fact, some regions have natural background radiation high enough that they would be evacuated if the cause was anthropogenic, but since it’s natural, people live there. In general, people living in places with elevated natural background radiation like Ramsar, Iran show no increase in cancer rates.(7)
If the body can handle such high levels of natural ionizing radiation, then what chance do the weaker non-ionizing frequencies have of doing any real damage? You would increase your risk of brain cancer more by holding a banana to your head, since that actually contains potassium-40, which gives off gamma rays. Whatever mechanism you think a cell phone has for causing cancer, it should apply even more to a banana. For a comparison, gamma rays have orders of magnitude more energy (Joules per mole) than visible light, whereas radiowaves and microwaves have orders of magnitude less. (6)
Thus far, laboratory testing shows no evidence for EMF caused DNA damage, even at exposure levels well above 0.4 microtesla. In fact, many organisms live perfectly well in the presence of fields measured in thousands of millitesla. There are also particle accelerator workers who are exposed to 300 milliT fields at times, with no increase in cancer. There has been a failure to find any association between static magnetic fields and DNA strand breaks, chromosome aberrations ,sister chromatid exchanges, cell transformation, mutations, or micronucleus formation . (8)
The WHO has classified RF-EMF as a possible carcinogen. Of course, people are fearful when they hear that something is a “possible carcinogen”, but after so much exhaustive research, it actually reflects a failure to find all that much in terms of evidence.
So when it comes to consistency and plausibility, there are some very large marks against EMF as an etiological agent. That doesn’t necessarily mean it’s not harmful at all, but the list of things that *might* be harmful is infinite. Before we start darting at shadows of small risks that might be there, we need to worry about the ones that are large and obvious enough to measure. In my previous posts, I have expressed similar views about vaccines and nuclear power, while acknowledging the measurable dangers of infectious disease and air pollution. In my next post, I will apply the same standards to GM crops. I have my preconceptions about this subject, but as always, I do not yet know for certain what my position will be. Wish me luck!