As the grandson and nephew of dentists I was probably the only kid at school who actually enjoyed going to the dentist. In fact, it wasn’t until I joined the Navy that I understood why my classmates never shared my enthusiasm for practitioners of the dental arts – having my wisdom teeth removed in preparation for submarine duty was my first really unpleasant experience in the dentist’s chair (although not my last). Another surprise was finding out that dental x-rays could be taken quickly – my grandfather’s x-ray machine dated back to the 1940s (maybe even earlier) and seemed to take a few seconds for an exposure. In fact, the first time I had an x-ray with a modern machine it was over so quickly that I demanded to see the developed film before I believed they’d actually done something. By today’s standards my grandfather’s x-ray machine was hopelessly antiquated and likely delivered an astronomically high dose of radiation. On the other hand, it’s the same machine that was used on all of Grandpa’s other patients – including my mother and her 5 siblings and countless other family members and not a few friends – all without any apparent ill effects.
Every so often it seems that dental radiation briefly makes an appearance as something that we wonder if we should be concerned about. In 2004, for example, the Journal of the American Medical Association published a paper purporting to show that dental x-rays were linked to low birth-weight children (the authors claimed that dental radiation somehow inhibited thyroid function, causing low birthweight children). The only problem was that the amount of radiation exposure delivered to the thyroid is far lower than anything shown to be able to have any impact whatsoever, and the authors failed to perform a relatively simple test for thyroid hormone levels that could have shown them to be right or wrong in their assertion. This paper was roundly criticized (including this fact sheet by the University of Texas Health Science Center) and to date there have been no other indications that dental x-rays can have any reproductive impact at all. As an aside – during my days as a radiation safety officer I had one pregnant patient whose dentist advised her to terminate her pregnancy after he found she was pregnant after receiving dental x-rays. Thankfully she asked for a second opinion because not even the most high-dose dental x-ray procedure can expose the fetus to enough radiation to cause problems.
Around this same time I was asked to consult on a British lawsuit in which patients were claiming that their dentist was taking too many x-rays, exposing them to unnecessary radiation. The problem was that, adding up the risk factors from the x-rays and comparing them to other risks, it turned out that the riskiest part of this procedure was the drive to the dentist and not the x-rays themselves. And that is if we place any confidence in radiation risk estimates at such low doses (on the order of a few millirem, or a few tens of microSieverts) – the Health Physics Society, the National Council on Radiation Protection and Measurements, and the International Commission on Radiation Protection all take somewhat different stances on this issue but all of these highly skilled and reputable bodies conclude that the risk varies between zero and incredibly low.
The latest round in this saga is a paper published in the journal Cancer just last month, suggesting that dental x-rays increase the risk of meningioma (a usually non-malignant tumor of the membranes surrounding the brain, and the most common form of brain tumor). But a careful reading from the standpoint of a radiation safety scientist leaves me unimpressed. Here’s why.
First is that the authors don’t seem to have calculated the radiation dose to the brain from the dental procedures they are concerned about. As any radiation biologist knows, radiation dose to the organ of interest is everything – every radiation dose-response hypothesis describes a relationship between radiation dose and cancer risk. Without knowing the estimated radiation exposure to the brain (specifically to the meninges) we can’t determine the probability of causation for the specific tumors and can’t see if there is any dose-response relationship. If radiation is causing these tumors then the number of “excess” tumors should increase as radiation dose increases – without knowing what the radiation exposure is then we cannot even guess whether or not radiation exposure is a reasonable explanation for the meningiomas that were seen.
In addition to a lack of dosimetric information the risk ratios reported by the authors are unimpressive. Most of the incidence ratios are very close to 1 (meaning that the cohorts studied have an incidence of meningioma very similar to that of the control group) and many of the groups had a ratio of less than 1. There are some groups with higher incidence ratio but these were only those who had received panoramic x-rays – not the more common bitewing x-rays.
It is also important that the dental x-ray history of those involved in the survey was based in the recollection of those surveyed – not on their dental records. I remember my grandfather (and later my uncle) taking dental x-rays, but I have no idea how many of these I had in my teens, and I have no idea how many of these were panoramic x-rays. I assume that, during my Navy days, I had annual dental exams but even here I honestly have no idea whether or not I had x-rays annually or if the Navy even gave me the panoramic x-rays. I have to admit that I am not a paragon of memory, but I’d wager a guess that my memory is no better and no worse than that of the average subject interviewed for this paper. The bottom line is that the closest thing to radiation dosimetric exactitude is based on the memory of those who were interviewed as a part of this study and memory is fallible and biased. At most, a study based on recollection is suggestive, at worst it is worthless. But it is also plausible to wonder if people with meningioma were more likely to recall a higher number of dental x-rays than those who were healthy.
So let’s take on the topic of dosimetry. Dental x-rays are low-dose procedures – according to the American Dental Association bitewing x-rays give a dose of about 4 mrem and full-mouth x-rays give a dose of about 15 mrem. Seventy sets of full-mouth dental x-rays would give a person a dose of about 10.5 rem – barely above the lowest dose (10 rem or 100 mSv) at which the Health Physics Society suggests that it’s scientifically acceptable to calculate a numerical estimate of risk (below this dose the epidemiology is quite fuzzy). And even at this dose the additional risk of cancer is about 0.5% above the background risk of 25% or so. This isn’t nothing – but it’s pretty low, and certainly lower than the excess risks suggested in this paper.
In fact, an earlier (2009) paper by Joanna Banerjee did show a dose-response relationship between radiation exposure and meningioma. But this paper determined the radiation dose and these doses were all far in excess of 10 rem (100 mSv). The doses reported by Banerjee and her colleagues are certainly high enough to cause problems –giving this paper more credence than the more recent paper that sparked this posting.
But all of this brings up a more fundamental point – that there is an assumption that even the slightest dose of radiation can cause problems and that the role of the author is to reveal these problems rather than to determine if they exist. There are any number of scientific papers that begin with the assumption that even the slightest dose of radiation will cause cancer and the goal of the author(s) is to tease out the relationship instead of trying to determine whether or not a relationship exists. This may seem reasonable, but it’s sort of like a parent starting with the assumption that their kids are getting away with something, rather than a parent giving their kids the benefit of the doubt. When authors assume a priori that every adverse health effect is due to radiation then they are beginning by assuming the conclusion that they are trying to prove. When you start off trying to prove your assumptions then you will frequently be successful – we are good at finding what we are looking for.
The bottom line is that there may (or may not) be risk from dental x-rays, but there is unambiguous benefit that comes from having them. In addition to showing the dentist the location of potential cavities (or abscesses) dental x-rays can also reveal osteoporosis, sinus infections, some tumors, and more. If we are to cast aspersions or raise cautions about dental x-rays then we must also acknowledge the benefits that accrue from them – to fail to do so will give us a skewed view of these procedures. And the paper discussed her not only assumes that all of the “excess” cancers are due to dental radiation, but it also fails to acknowledge these benefits.
The fundamental problem is that, while radiation can cause health problems, it takes far more radiation to do so than most people realize. Radiation can cause cancer, for example, but the cancer risk from 1 rem (10 mSv) of exposure (typical of a CT scan) is far lower than even the risk from driving (and may actually be zero). Similarly, radiation exposure can cause birth defects, but the amount of radiation required for this is far higher than most realize. Thus, radiation finds itself blamed for all sorts of things that it is simply unlikely to have caused.
And here is where I have to reveal my biases as a radiation scientist. I have to admit that I grow irked when papers such as this are published – papers that might make sense mathematically but that make no sense at all when considering the totality of what we know of radiation health effects. Anyone can take a radiation health effects slope factor and calculate a risk for even vanishingly low doses of radiation – but these calculations are meaningless because they are uninformed by reality and they are out of touch with the real world. As an example, I might calculate that an x-ray that gives one a dose of 10 mrem might increase my risk of cancer by 0.005%. Let’s face it – every time I push the “=” key on my calculator I’ll get an answer. But this answer neglects the advice of the Health Physics Society to avoid calculating a numerical risk estimate for dose of less than 10 rem, it ignores the statement by the ICRP that such doses are “trivial,” and it fails to acknowledge not only the risk from driving (1%) but also the background cancer risk of 25% or more. This not only skews the authors’ interpretation of their results but it also gives radiation a weight it does not deserve. I do not object to radiation being blamed for what it has (or is likely to have) caused, but I do object to its being unfairly blamed for effects that are implausible. In this case, unfairly blaming radiation from dental x-rays for causing meningioma is not only inappropriate in that it flies in the face of or experience and the recommendations of respected scientific bodies, but it might (ironically) do a disservice by dissuading people from receiving needed dental x-rays. If this happens then the net result of this paper – and others of its ilk – will be to the detriment of public health.