On February 2nd, 2018, the National Toxicology Program (or NTP), a US government toxicology research programme, produced two technical reports [1, 2] and a press release on the results of experimental studies on the effects of radiofrequency electromagnetic waves in rats and mice. Begun in 2003 and carried out over a ten-year period with a budget of US $25 million, the study is touted as the most comprehensive ever to evaluate the effects of exposure to mobile phone emissions on rodent health. These reports are currently being reviewed by a panel of scientists whose findings are expected by the end of March, and so cannot yet be considered final.
The objective was to identify potentially carcinogenic and toxic effects in two laboratory animal species (rats and mice) chronically exposed to radiofrequency fields at two frequencies (900 MHz and 1900 MHz) and signal modulations of the type used in GSM telephony (European and American system) and CDMA (US system) corresponding to 2nd and 3rd generation (2G, 3G) phones. Due to the start date of the research program, newer or emerging systems were not considered (4G, 4G-LTE, 5G).
The development of the exposure device had required considerable work prior to the experiment, in order to better control exposure conditions and limit experimental biases. The exposure chambers (reverberation chambers) allowed for exposure over the animals’ whole bodies comparable to that of distant transmitters such as relay antennas (known as “whole body” exposure).
In the experiments, chronic exposure (7 days/7, alternating 10 minutes on/off for a total of 18 hours and 20 minutes per day, that is, 9 hours of actual exposure per day) lasted from 28 days up to 2 years. The animals were free to move around their cages.
The mice were about ten days old at the beginning of the exposure. For rats, exposure started at the fetal stage (in utero) from day 5 or 6 of gestation (maternal exposure), then continued for 28 days after birth (lactation and after weaning), or over a period of 2 years. A sample of the animals was terminated in each group at 14 weeks for an interim study. Each group was composed of males and females in equal numbers.
The rats were exposed to the 900 MHz frequency (GSM and CDMA) with SARs of 1.5 to 6 W/kg (see table), while the mice were exposed to the 1900 MHz frequency (GSM and CDMA) with SARs of 5 to 15 W/kg for 28-day experiments, and 2.5 to 10 W/kg for two-year experiments. Finally, a group of control animals was similarly housed but without radiofrequency exposure.
Several biological and physiological parameters were studied during the experiment (body weight, blood parameters, general condition …). Studies investigating genetic damage, known as genotoxicity studies , were carried out in circulating blood, brain and liver cells after 14 weeks of exposure. When the animals were euthanised (at 28 days, 14 weeks or 2 years) or died during the experiment, the organs were examined and weighed, and the tissues were systematically analysed using various methods to look for lesions (histopathology), changes in the genital functions (sperm quality, oestrus, genitals …), existence of cancers and tumour analysis when relevant.
Overall, exposure to radiofrequency fields had no effect on reproductive functions or survival in the rodent groups. The results on genotoxicity did not show a trend towards proof of any specific result. Finally, the histopathological study showed some effects, but the data is of limited use, because of a lack of coherence in the results. In particular, abnormalities were observed in the control populations in male rats and in female mice.
So male control rats (not exposed to radiofrequency fields) did not live as long as exposed male rats. According to the authors, this excess premature mortality (from 75 weeks) was due to chronic renal pathology. In absolute terms, this could lead to an artificially increased incidence of tumours (number of new cases per year) in the exposed groups, since cancerous pathologies develop more with age. In female control mice, the incidence of lymphoma was lower than that usually observed (i.e. in historical controls), by contrast to exposed mice. So, the increase in the incidence of lymphomas in exposed mice compared with control animals seems to be, in this study, the consequence of a decrease in incidence in the control population. It is therefore not a priori an effect of the radiofrequency fields.
However, an increase in a heart tumour, the malignant schwannoma, was observed in male rats at the highest exposure level. This could be related to a more sustained requirement of the heart muscle to regulate body temperature due to induced heating under these conditions (less than 1°C but continued for 2 years). It is surprising, however, that only the exposed male rats exhibited this condition, which was not found in female rats or in exposed male or female mice. In humans, cardiac schwannomas are generally benign tumours and are a very rare pathology, where an increase in frequency in the general population could not go un-noticed.
Finally, the glioma data were particularly keenly anticipated, as it is a brain tumour for which some epidemiological data has suggested an association with an intensive use of mobile phones (in the category of “heavy users”). However, the incidence observed in male and female rats and mice exposed for life in the NTP study remains within the range observed in nature for these rodent lines. So, for gliomas the study is very reassuring.
Can the results be extrapolated for the current use of mobile telephony?
A previous report covering the preliminary results from the rat study was published by the NTP in 2016 , accompanied by a press release . The authors concluded that exposure to mobile phone radiofrequency fields slightly increased the frequency of two rare types of brain tumours (gliomas) and heart tumours (schwannomas). Several inconsistencies were then picked up in the results (see  for examples).
Similarly, in the titles of the reports, as well as in the text of the 2018 press release, mobile phone radiofrequency fields are explicitly mentioned, suggesting that the research programme covers exposures comparable to those of mobile phones. This is however not the case. The difference is a major one, because when the entire body of an animal is exposed at the same level (an exposure referred to as “whole body”), the level of exposure at which the first effects due to heating occurs is 4 W/kg. This finding was used to define the average exposure threshold for humans at 0.08 W/kg maximum for the public (that is, 50 times less). In comparison, exposure to mobile phones mainly involves the head (known as “local” exposure) and the SAR from which effects are observed is then 100 W/kg for localised exposure, leading to a regulatory threshold of 2 W/kg in humans. Exposure to radiofrequency fields induces tissue heating, which increases with the level of exposure; the greater the volume of the body exposed, the harder it is for the body to regulate the heat generated. As an example to illustrate the importance of this distinction: it is less dangerous to dip the tip of your finger in boiling water than to fall into it.
For example, considering rat experiments, the SAR values for whole body exposure were 1.5 W/kg, 3 and 6 W/kg depending on the group of rats; during pregnancy and through the whole lifespan. As regulations limit whole body exposure in humans to 0.08 W/kg, the exposure given to the rats was therefore respectively 18, 37 and 75 times greater than the exposure authorised for humans .
Could this study challenge the IARC classification?
In 2011, the International Agency for Research on Cancer (IARC) concluded that radiofrequency fields from mobile phones were “possible carcinogens” (listed as 2B classification) for gliomas or acoustic neurinomas .
This classification means that, in the absence of sufficient evidence, radiofrequency fields are not “certainly” or “probably” but “possibly” carcinogenic, as are pickled gherkins and many other risk factors they assessed. Since 2011, several epidemiological studies conducted in different countries, most recently in Australia  and in Japan , do not indicate an increase in the incidence of brain tumours in the population despite the massive rise in mobile phone usage in recent decades. Obviously, the NTP study does not provide sufficient scientific evidence to change this classification on objective grounds.
At the end of this study, the authors are not in a position to say that the exposure to radiofrequency fields studied has an effect on health, and that is at levels well above those encountered in a normal setting. NTP’s contribution is just one in a long line of research conducted in the area of radiofrequency fields. It consolidates current knowledge and reinforces the fact that when effects of mobile radiofrequency fields can be observed, it is at exposure levels that far exceeds the maximum permissible exposure values. In practice, these limits cannot be reached with commonly used wireless communication technologies (relay antennas, mobile phones, WiFi …).
By releasing, as early as 2016, preliminary results of a study conducted under exposure conditions that have nothing to do with those generated by mobile telephony or other wireless communications systems in actual use, the NTP has needlessly encouraged a climate of rumour and worry for months, without helping those who are not specialists in the subject to understand the findings. Ultimately, the final and in-depth analysis of the results of this study is, despite its weaknesses, reassuring: it does not reveal any tangible proof of the existence of non-thermal effects of radiofrequency fields.
 NTP report on Toxicology and carcinogenesis studies in HSD: Sprague Dawley SD rats exposed to whole-body radio frequency radiation at a frequency (900 MHz) and modulation (GSM and CDMA) used by cell phones.
 NTP report on Toxicology and carcinogenesis studies in B6C3F1 / N MICE exposed to whole body radio frequency radiation at a frequency (1,900 MHz) and modulation (GSM and CDMA) used by cell phones. https://ntp.niehs.nih.gov/ntp/about_ntp/trpanel/2018/march/tr596peerdraft.pdf
 Report of Partial Findings from the National Toxicology Program “Carcinogenesis Studies of Cell Phone Radiofrequency Radiation in Hsd: Sprague Dawley® SD rats (Whole Body Exposures)”, Draft 5-19-2016. https://ntp.niehs.nih.gov/go/834451
 Perrin A., Lagroye I., Yardin C. Lien entre cancers et téléphones portables : la communication tendancieuse qui s’est organisée autour du rapport américain sur les ondes électromagnétiques [The link between cancers and mobile phones: the biased communication that has been organised around the US report on electromagnetic waves] 2016. Atlantico.
 International Commission on Non-Ionizing Radiation Protection (ICNIRP), ICNIRP statement on the “Guidelines for limiting electromagnetic fields, up to 300 GHz”, ICNIRP Statement (2009).
 IARC Monogr Eval Carcinog Risks Hum. 2013; 102 (Pt 2): 1-460. Non-ionizing radiation, Part 2: Radiofrequency electromagnetic fields.
 Chapman S et al. Has the incidence of brain cancer risen in Australia since the introduction of mobile phones 29 years ago? Cancer Epidemiol 2016; 42: 199-205. doi: 10.1016 / j.canep.2016.04.010
 Sato Y. et al, 2016, Time trend in incidence of malignant neoplasms of the central nervous system in relation to mobile phone use among young people in Japan, Bioelectromagnetics, Volume 37, Issue 5, July 2016, 282–289.
What is SAR?
In the frequency range 10 MHz to 10 GHz, the Specific Absorption Rate (SAR) is the magnitude that quantifies exposure to radio waves. It is the power absorbed per unit of mass of material, at the whole body level or locally (organs, media or biological tissue), expressed in watts per kilogram (W / kg).
 GSM: Global System for Mobile Communications
 CDMA: Code Division Multiple Access
 SAR: Specific Absorption Rate
 Study of the integrity of the DNA molecules that make up the genetic material or genome.
 Data on historical controls are data from the numerous carcinogenesis studies conducted by NTP and which provides a reference to overcome experimental problems such as those encountered in the control populations of this NTP study.
This post is also available in: FR (FR)