The process of skepticism and scientific inquiry, it has become very apparent to me, requires constant learning and re-evaluation of one’s unfounded beliefs. My own personal journey has been one that came from a position of simplistic certainly about the world and has resulted in a necessary embrace of uncertainty and a probabilistic understanding of the universe. How we know A caused B instead of B occurring coincidentally with A is the point, really, of the increasing sophistication of scientific inquiry; with the past 60 years seeing a overhaul of medical investigation.
Enter cell phones. We are constantly reminded of the failure of society to recognize the dangers of tobacco, let alone do something about it, and the industry led effort to suppress information and increase uncertainty in the pubic is held up as proof that all industries will stop at no lengths to protect their investment, despite dangers to the public. In the face of this, we need a scientific outlook to unblinker us from determining an unbiased truth. A new systematic review published in October’s Bioelectromagnetics is an excellent illustration of how we determine causality.
Rapacholi et al set out to systematically review the current knowledge of the effects of cell phones on brain cancer from the two major avenues used to determine causality: epidemiology and in vivo experiments (done in living animals). Through a very comprehensive systematic review format, where all of the studies matching a pre-determined set of criteria are evaluated for their quality and conclusions are drawn given all of the evidence, the researchers sought to lay out the state of the evidence for any connection between the use of cell phones and the incidence of several types of cancerous tumors in the head. They used all studies published in peer-reviewed journals up until January 2010, so they did not include the newest Danish epidemiological study, which showed no relationship between cell phones and cancer.
Modern efforts to determine causality in medicine have culminated in the Hill Criteria, first described in 1965 by Dr. Austin Bradford Hill, and the conclusions drawn from the current evidence by Rapacholi et al use these criteria, along with a comprehensive qualitative weighting system that takes into account the design and reporting of the individuals studies. The study design and paper are only available behind a paywall, so the average reader does not have access to the search and analysis criteria, unfortunately.
The results of the study should not surprise any reader of Skeptic North over the past year. The investigators concluded that the epidemiological studies found no relationship between cell phone use, both short term or long term, and meningiomas, gliomas, acoustic neuromas, or parotid gland tumors. This anaylsis also included a meta-analysis of pooled data results, so it was also evaluated statistically, not just by qualitative measurement: both facets of the investigation found no relationship.
The authors found a consistent disagreement between all of the epidemiological studies and the results of Hardell et al, with the theme of this disagreement being poor study design and control of recall bias – the bias that creeps in when you ask a group of disease suffers to write down their exposure to a possible causative agent over many years of use – I do not know about you, but there is no way that I could accurately recall the amount of cell phone use I had 10 years ago without consulting a bill. Hardell is indicted consistently throughout the paper for having poor study design and was accorded less weight than the other more rigorous studies that were based on objective usage data or took extensive measures to investigate bias in their study, like the Interphone study (a new iteration of which has found no relation between cell phone use and acoustic neuromas).
Ecological studies look at population level statistics, and Rapacholi et al looked at these studies as well. The risk of tabacco smoke was noticed by public health officials first when they say an increase in lung cancer deaths at the beginning of the 20th century and wondered what the cause was. If similar studies show an increase in brain tumors through the wide-spread use of cell phones starting in the late 1990′s then public health officials would be raising similar alarm bells. Rapacholi et al found no such increase at the population level in any of the countries studied from the early 1970′s to the early 200′s. Overall, tumor rates have remained flat in the northern European countries, the UK and Ireland as well as in the US.
Cordless phones were not left out of this studies (I hope it is clear that this was a very comprehensive review!) and no association can be drawn from those studies that were given the greatest weight between cordless phone use and brain tumors. Once again, Hardell et al is the outlier in these studies and the authors could not account, overall, for the reasons Hardell is finding associations when others are not. In one study in 2010, Hardell relied on reporting data from next-of-kin so it appears as if his group is grasping at straws here.
The in vivo studes were no different in their conclusions. These studies were attempting to determine a possible mechanism for any tumor formation from cell phone use as well as control for other variables that are difficult in epidemiological studies. Overall, it was clear that those studies that had better dosimetry and were properly blinded, and were therfore accorded more weight, showed no relationship between RF exposure and genotoxicity, or DNA damage. Concordently, no relationship was found between RF exposure and tumor production, in both the qualitative analysis and the pooled data analysis. This is not looking good for those still purporting a link.
This study represents a very comprehensive evaluation of the evidence so far for a causal relationship between cancer and cell phone use. In the conclusion, the authors apply their data to the Hill Criteria and find little strength of evidence, no dose-response relationship in a very heteraganeous data set, with no temporal relationship in the in vivo studies between exposure and tumorgenisis. As well, they are quick to point out that so far:
“Despite extensive research over many years, no interaction mechanism has been established whereby exposure to low level RF fields (below the level where heating is the dominant mechanism) from wireless phones could cause or contribute to disease in living organisms. Overall, the lack of an appropriate mechanism and the results of the in vitro and in vivo studies do not provide any support for causality.”
It is important to note that there is little if any data for more than 10 years use and we should continue to use ecological and epidemiological studies to see if an increase in brain tumors occurs. As well, there is very little if any study in children so we should continue to be vigilant, though given the above statement, it is doubtful that we will see any difference between childhood and adult exposure.
Two of the principle investigators did report funding from agencies tied to the cell phone industry, but original studies that they did were excluded from the analysis so they did not affect the outcome. As well, there were 13 other authors on this paper, and all 15 authors approved the inclusion and analysis criteria as well as participating in the analysis. Given this, it would be very difficult for any industry bias to confound the outcomes of the study.
Once again, science progresses, and given this new study, along with the Danish study, it is unfortunate that the WHO did not wait until the beginning of 2012 to make their decision about including cell phones on the 2b list of possible carcinogens. It is nigh impossible to get anything removed from this list, because you have to now show that there is no possibility of carcinogenicity by the agent. However, given this most recent systematic review, people can once again rest assured that no link has been shown to exist between cell phones and cancer.
Repacholi MH, Lerchl A, Röösli M, Sienkiewicz Z, Auvinen A, Breckenkamp J, d’Inzeo G, Elliott P, Frei P, Heinrich S, Lagroye I, Lahkola A, McCormick DL, Thomas S, & Vecchia P (2011). Systematic review of wireless phone use and brain cancer and other head tumors. Bioelectromagnetics PMID: 22021071