The Australian Centre for Electromagnetic Bioeffects Research (ACEBR) is a National Health and Medical Research Council (NHMRC) Centre of Research Excellence. With over 6 billion mobile phone subscriptions world-wide, the electromagnetic energy (EME) that powers this technology is now ubiquitous, as is community concern about the possibility of associated health effects. Responding to this concern, ACEBR has embarked on a 5-year research program to promote Australia's EME health both in the immediate future, and through the development of human research capacity in this field into the future.
Science & Wireless 2017
ACEBR together with PRESEE will be hosting the annual Science & Wireless event this October 19 at RMIT (4:30pm - 8pm), Melbourne, Australia (Building 12, level 7, 402 Swanston St, Melbourne VIC 3000).
The focus of this year's event will be a summary of some of the latest research coming out of the ACEBR and PRESEE research groups, as well as a keynote presentation given by Joe Wiart: 'Looking to the future in terms of 5G'.
The results of the US National Toxicology Program’s cell phone studies in rats are inconsistent and unconvincing - an ACEBR Position Statement
There has been considerable interest in a recent report from the US National Toxicology Program (NTP), concerning a study designed to determine whether mobile phone-like radiation (radiofrequencies, RF) can cause cancer in rats and mice. The results have been eagerly awaited by researchers in the field because the study was backed by substantial investment and was thus able to address a range of potential limitations with previous studies. The resultant report was strongly worded (e.g. concluding that “the observed hyperplastic legions and glial cell neoplasms of the heart and brain observed in male rats are considered likely the result of…RFR”; p. 15), and thus suggests an important contribution to the RF health debate. The report is described as including only ‘partial findings’. The study was designed to include rats and mice. No increases in cancers or hyperplastic lesions were seen in female rats. No results on mice are given; they will be released at a later time. The report includes the comments of several reviewers, which include several major issues relevant to the interpretation. The issues raised are not dealt with in the report.
An overall consideration is that the NTP release does not contain sufficient information to enable adequate review. For example, without detail of other endpoints tested it is hard to correctly interpret the statistical analyses. This is pointed out by Reviewer Dr Lauer, from the National Institutes of Health, who writes
“Why aren’t we being told, at least at a high level, of the results of other experiments (i.e., male and female mice, tissues other than heart and brain, tumors other than glioma and schwannoma)? Given the multiple comparisons inherent in this kind of work… there is a high risk of false positive discoveries. In the absence of knowing other findings, we must worry about selective reporting bias.” Dr Lauer states “I am unable to accept the authors’ conclusions”. Indeed as pointed out by the NTP Associate Director Dr Bucher at the NTP press conference, about 20 to 30% of the scientists within NTP who examined the report did not agree with the conclusions; an indication that more thorough scrutiny was required. Without such thorough scrutiny the conclusions cannot be taken as more than the provisional positions of the authors, rather than a scientific contribution.
However, even given the limited information provided, there are a number of issues that stand out and question the relevance of the NTP report to public health. These include both methodological and interpretational issues.
A few methodological issues are of particular importance:
1/ There were strong and unexplained anomalies in the control (unexposed) rats. Survival rates of the exposed rats were 50, 56 and 60% for the 1.5, 3 and 6 W/kg GSM RF groups respectively, and 48, 61 and 48% for the corresponding CDMA RF exposure groups (which are within the normal survival range in NTP control rats, given as 24 – 72% in the report). Conversely, rates were only 28% for the control rats in this study. This large difference between the control group and all of the RF exposed groups is unexplained. This means that the control rats had a shorter lifespan, so for this reason alone will develop fewer tumours, and indeed no tumours did develop (whereas close to the expected rate of tumours, from previous experience, was found in the exposed rats). It is difficult to know how best to deal with such methodological issues, particularly as the rats’ age, death and time of tumour development will all be heavily correlated and thus cannot be adequately accounted for statistically. In fact as noted by Reviewer Dr Lee, it would only take 1 tumour in the control group for the ‘statistical significance’ to disappear.
2/ The number of statistical comparisons performed is also crucial for determining the relative importance of the results. For example, if 10 comparisons were performed, finding 1 of these ‘statistically significant’ would be equivalent to tossing a coin and getting a ‘head’. Thus given that there were numerous analyses performed, and only some of these reported, it is difficult to determine whether the results were due merely to chance.
A few interpretational issues are of particular importance:
1/ The NTP report claimed to have found significant dose-dependent effects of the exposure. However the data provided does not show a regular graded disease response as a function of exposure for glioma. For example, the malignant glioma rates for CDMA exposure in male rats that were claimed to be dose-dependent were 0, 0, 0 and 3.3%, for the four groups of control, 1.5, 3 and 6 W/kg respectively (where the historic range in controls is given as 2%, range 0-8%). So the cancers only occur in the 6 W/kg group, which is not what is normally taken to mean a dose-dependent or graded relationship. There is no clear increase with increasing exposure. For heart schwannoma in male rats, the results show occurrences of 0, 2.2, 1.1 and 5.5% for GSM, and 0, 2.2, 3.3 and 6.6% for CDMA exposure, for control, 1.5, 3 and 6 W/kg groups respectively (where the historic range in controls is 1.3%, range 0-6%). So although there is some evidence of higher rates at the highest dosage, the dose-response is still far from clear, and hard to interpret because of the zero events in the control group, where the lifespan was shorter. To conclude there is a dose-dependent effect requires further confirmatory research.
2/ Although the authors claim that the exposures are relevant to human telecommunication-related exposures, this is only in the sense of potentially showing a relation between RF and cancer, but not in terms of the magnitude of exposure that people are likely to be exposed to. The exposures used in the study range from 1.5-6 W/kg whole body exposure, which are all many times larger than the International Commission on Non-Ionising Radiation Protection (ICNIRP) limits for both the general public (0.08 W/kg) and occupational exposed workers (0.4 W/kg). The lowest dose used, 1.5 W/kg, is similar to the public exposure limits for localised exposure of the head from a mobile phone, but the other dosages are much greater. Further, the exposures are for 9 hours in each 24 hours, from before birth and throughout life. It may be argued that although whole body exposure is very high, that the magnitude of local exposure is none-the-less relevant to mobile telecommunication. This is indeed true, however if that was the purpose of the study it would be seriously confounding the results due to the effect of increasing body core temperature. For example, 6 W/kg whole body average SAR in some studies has been shown to increase body core temperature in rats by approximately 1oC (and modelling suggests similar increases in humans), which the ICNIRP (1998) Guidelines treat as potentially harmful and set restrictions to protect against. Scientifically this does not make the results meaningless, as even at 6 W/kg this would be a novel finding of a relation with cancer that has not been demonstrated to date, and one that could in principle be predictive of harm at much lower levels. It should also be recognised that these SAR values were chosen so as not to compromise the thermoregulatory system in rats, rather than to be representative of normal human exposures.
3/ It is also important to note that although the schwannoma was interpreted in the study as an analogue of acoustic neuroma (vestibular schwannoma) in humans, there was no evidence of increased schwannomas in other regions of the rats (which were as common as those in the heart). This also raises the question of why RF would affect schwannomas in one but not other regions of the body.
4/ Internal consistency in the study makes interpretation of the results difficult. That is, effects were reported for males but not for females, and pathology from other regions of the body were not reported as being associated with exposure. This suggests that if the results are correct, then there must be important gender- and region-specific processes, as well as species differences, that mediate the effect of RF and that are not currently known. This does not in itself invalidate the results, but particularly given the other methodological issues described above, this makes it more difficult to accept these results without further clarification. The results in mice, not yet reported, need to be assessed.
The NTP report is thus not able to comment on whether the magnitude of RF exposure that people may be exposed to from telecommunications devices is capable of affecting their health. Therefore, although the NTP study has many strengths, at present it must be taken as preliminary and as such does not contribute to the mobile telecommunications health debate; we are left with the current consensus that there is no evidence that mobile telecommunications-related RF causes cancer.