Sound Exposure and Risk Assessment of Wireless Network Devices

Final Report Summary - SEAWIND (Sound Exposure and Risk Assessment of Wireless Network Devices)

Executive summary:

Wireless network devices have become an integral part of modern society as they facilitate our lives by offering a broad range of capabilities oriented toward greater user mobility, flexibility, and productivity. Along with this development, our daily exposure to radio frequency (RF) electromagnetic fields (EMF) where we live, work, and play is rapidly increasing. During the course of SEAWIND project, the technology has become even more pervasive due to the growing popularity of smart phones and tablet computers, which had already reached the milestone of 1 billion users in October of 2012 and is expected to double over the next 3 years.

Project Context and Objectives:
Wireless devices are an integral part of modern society that have become ubiquitous in our homes, schools, hospitals, and workplaces in the last decade, as the populace becomes increasingly mobile and technologically sophisticated. Although wireless communications facilitate our lives by offering a broad range of capabilities oriented toward greater user mobility, flexibility, and productivity, our daily exposure to electromagnetic fields (EMF) is rapidly increasing where we live, work, and play. Among the main sources of exposure are wireless local area networks (WLAN), wireless metropolitan area networks (WMAN or WiMAX), and body-mounted and body-worn wireless personal area networks (WPAN). Another rapidly evolving technology is radio frequency identification (RFID).

The exponential growth of wireless network device usage necessitates that the scientific basis for assessment of potential health risks due to EMF exposure in everyday life be broadened, especially since the International Agency for Research on Cancer (IARC) has classified radiofrequency (RF) EMF as possibly carcinogenic to humans (Group 2B) based on studies that show increased incidence of glioma, a malignant type of brain cancer, associated with wireless phone use (see http://www.iarc.fr/en/media-centre/pr/2011/pdfs/pr208_E.pdf online).

In ten scientific work packages under the umbrella of an eleventh work package devoted to management, SEAWIND addressed the entire scientific spectrum from dosimetry to biology with the aim to satisfy the following objectives:

1. In a comprehensive literature review on in situ and laboratory incident exposure evaluations as well as experimental dosimetric evaluations of wireless networks (Work Package 1: 'Review of exposure assessment and dosimetry of wireless networks'), shortcomings and knowledge gaps should be identified at the outset of the project. The literature is monitored throughout the lifetime of the project and subsequently updated at the end of the project.

2. Investigation of current and future wireless network systems from device and system points of view (Work Package 2: 'Review of communication systems, signals, and power modulations'): The most important aspects to consider were power classes and power envelopes of systems as a function of traffic. The former is important for evaluation of compliance and worst-case exposure, whereas the latter is relevant for assessment of average exposure strength and exposure signal characterization.

3. Development of instrumentation and procedures for the accurate exposure assessment of these devices (Work Package 3: 'Development of instruments and calibration techniques'). The knowledge gained is disseminated via Work Package 9 ('Dissemination to standards') to the appropriate standardization groups.

4. Assessment of worst-case exposures required to demonstrate compliance with safety limits as well as for risk assessment (Work Package 4: 'Incident field evaluations for whole-body exposure'). This needs to be performed for the far-field exposure scenarios and for the near-field exposure scenarios (Work Package 5: 'Dosimetry for worst-case partial body and local exposure') are evaluated.

5. Assessment of typical daily-life exposure scenarios (Work Package 6: 'Organ specific dosimetry') by selection of source models and postures to represent exposure of the general population and the generation of anatomical models in appropriate postures and dosimetric evaluations for adults, children, and pregnant women.

6. Development of theoretical propagation models to characterize the exposure, which are thoroughly validated with respect to epidemiological considerations.

7. Testing of the models with various technologies to check that they accurately predict exposure to wireless devices.

8. Derivation of simplified exposure prediction models that can be used by non-engineers and that nevertheless permit reliable determination of exposure for any multi-system/exposure situation.

9. Derivation of guidelines for optimal installation and usage, to maximize connectivity and minimize exposure

10. Translation of the above-mentioned exposure scenarios to organ-specific exposures in the time domain for the entire user population (Work Package 6), which is necessary for accurate risk evaluations (the use of the incident field strength only is too inaccurate for risk evaluations).

11. Application of standardized procedures and development of advanced biological models and experimental procedures to sensitively and quantitatively measure transient and persistent molecular and cellular responses due to EMF exposure; the concerted approach focuses on in vitro and ex vivo analyses of cells with genotoxicity and genomic instability as endpoints (Work Package 7: 'Genotoxicity screening in vivo and in vitro').

12. Evaluation of the four most-dominant signals that address the biological relevance of real-life exposure scenarios (Work Package 7), which is finalized at the end of the project.

13. Development of novel exposure systems for the biological screening of wireless EMF exposures. Three exposure systems in total, two for in vitro and one for rodent exposure, should be produced (Work Package 8: 'Exposure systems and quality control').

14. Dissemination of the instrumentation, calibration techniques, and assessment procedures developed to the relevant international standardization groups (CENELEC, IEC, ICES, ARIB, etc.) (Work Package 9: 'Dissemination to standardization agencies').

15. Performance of a comprehensive risk-governance protocol to integrate the scientific risk assessment results, the risk perceptions by key stakeholders and users, and the wider social and ethical concerns. The goal is to provide society with all the relevant data and background information to make prudent choices and to have a firm basis for making tradeoffs between benefits and risks (Work Package 10: 'Risk governance, integrating assessment, perception, and communication').

16. Development of communication and dissemination tools for improving risk communication practice, which is finalized at the end of the project.

17. The communication systems considered should focus mainly on systems for data communication with special attention to lower-range systems (e.g. WiFi, WPAN, and BAN), which are foreseen as being employed everywhere indoors, e.g. the IEEE 802.11 IEEE 802.15 and IEEE 802.16 families. The exposure characteristics are compared to mobile systems for well-known telephony as well as for data, including GSM, UMTS and its extensions, 4G LTE, DECT systems, Bluetooth, etc. The systems will be grouped according to exposure characteristics, e.g. power levels, bandwidth, duty cycle, modulation, spectrum of the power, frequency, etc. (e.g. Ultra Wide band (UWB) systems, RFID systems, etc.).

The SEAWIND effort consisted of eight expert groups (including SMEs) from five European countries, all with internationally recognized competence in their respective fields. The contributing scientists offered a wide range of expertise in exposure assessment, dosimetry, biology, and risk assessment. The consortium was supported by two external advisors with excellent reputations for their achievements in risk-assessment processes within public institutions.

Project Results:
1. Description of Wireless Network Devices and Communication Systems

Technological progress has led to rapid changes in the situation with regard to wireless network exposure due especially to the transition from simple mobile phones, used for speech and short messages, to complex 'smart phones', which offer high rate data communications. Progress has continued apace even during the lifetime of the SEAWIND project. The main exposure factor of concern remains the use of devices in speech mode with the device (phone) close to the head. This is by far the largest exposure in terms of specific absorption rate (SAR) values in the head. Body-supported or hand-held usage during speech mode, e.g. by means of a hands-free kit, causes similar high exposures of body regions close to the phone. High exposures also occur when the phone is used as data network terminal over mobile networks (tethering). There is also growing public concern related to remote access points in offices and homes. However, not only is the average power of an induced field important for full characterization of exposure, but also the envelope of the frequency- and time-domain power variations play a role. The following is a brief overview of the systems described in lay terms to the best of our ability.

1.1 Global System for Mobile Communications (GSM)
GSM is also known as second generation (2G) mobile communication, a digital system, whereas the first generation (1G) is the now obsolete analog system. The system is imbedded as a choice in all modern phones, and it is a matter of coverage and network provider settings whether or not 2G is used (the user may also choose the system via the mobile phone settings). GSM, counting billions of users, is the most widely used system. In GSM, power is not constantly delivered but is transmitted in short bursts on the order of milliseconds (0.57 ms), allowing multiple users (up to 8) to share frequency band channels between the bursts. The frequencies, which are different for uplinking (from device to base station) and downlinking (from base station to device), lie in Europe in the 900 MHz and 1800 MHz bands. The data rate is small, on the order of tens of kilobits per second (kb/s). Mobile stations have peak powers of 1 W at 1800 MHz and 2 W at 900 MHz. In speed mode, only 1 of 8 time-slots is used during uplink, thus, the frame-average maximum output power is approximately 1/8 of the aforementioned peak power levels. Base stations transmit depending on the cell size up to several hundreds of watts. The dominant modulations of the output power are at frequencies of 2, 8, and 217 Hz and their harmonics. The antenna input power of the mobile unit, which is controlled by the base station over a range of 30 dB, start at maximum power upon handovers between base stations. In the real world, the average output power is dominated by the handovers, resulting in an average of 30 50% of the frame-averaged maximum output power (Kuhn, S.; Kuster, N., IEEE Transactions on Electromagnetic Compatibility 99: 1 - 13 (2012)).

1.2 Universal Mobile Telecommunications System (UMTS)
UMTS, also known as wideband code division multiple access (WCDMA), is the 3G system that is gradually replacing GSM. It is technically very different, as it uses a code technique that allows multiple users to simultaneously access the same frequency bands and time slots. The advantage is obtained through use of a wider bandwidth than in GSM, i.e. 5 MHz, with the additional advantage of a processing gain (due to intelligent encoding of the data and speed communication) of up to a factor of ten thousand in power. This has the important consequence that, although the maximal values of output power are similar to those of GSM, the mean value in practice may be in the micro to low milliwatt range, typically between a factor of 50 100 below the maximum output power (Kuhn, S.; Kuster, N., IEEE Transactions on Electromagnetic Compatibility 99: 1 - 13 (2012)). Of course the mean value of the specific absorption rate (SAR) is reduced accordingly. Power variations in the low frequency range, below 1000 Hz, are more noise-like and do not conform to specific frequencies like GSM signals.

1.3 Long Term Evolution (LTE)
The quest for higher wireless data rates has pushed the development of 4G, LTE, with promised downlink rates of up to 100 Mbps under good propagation conditions. Clearly, the emphasis is not on speech, but rather data communication to laptops and tablets, where the distance to the head is large with ensuing low SAR values. Voice communication in LTE is foreseen via Voice-over-IP (VoIP) technologies, similar to the well-known Skype service. The nature of the signal and access techniques are complicated, adaptive, involving both the frequency-domain (narrow bandwidth slots) and time-domain slots. One novel feature is the adoption of multiple antennas at both transmitters and receivers. LTE has not been studied further in the SEAWIND project.

1.4 WiFi local area networks (LANs)
Wireless LANs are highly pervasive in the field of short-range wireless data communications and are partially replacing not only cellular solutions like 2G and 3G but also wired LANs. They exist now in many private homes, schools, hotels, trains, buses, airports, etc. The maximum power is 100 mW in the 2.45 GHz band and up to 1 W in the 5 GHz band. Personal exposure levels are, in general, very low, except when very close to an access point or a mobile device using WiFi. In the latter case, the maximum SAR levels can be as high as levels from mobile phones. The maximum data rate is 54 Mbit/s for commonly used devices and up to 600 Mbit/s for modern devices. Exposure is highly dependent on the usage scenario with respect to the body, e.g. head-mounted, front-to-face, hand-held, body-worn etc. Average exposure is also heavily dependent on the actual data throughput. Although exposure is often low, the pervasiveness warrants inclusion of the system in the SEAWIND studies.

1.5 Digital Enhanced Cordless Telecommunications (DECT)
DECT is a cable-replacing radio technology suited for voice, data, and networking applications with range requirements of up to a few 100 m. Its use in cordless phones is widespread in many homes. ETSI specified frequency bands between 1800 2500 MHz. The peak output power of the portable part, the phone handset, which is typically used close to the head, is 250 mW; however, the mean value when only one timeslot is used is 10 mW. In standby mode, the fixed part transmits a short packet-bearer beacon for an average power of 2 mW. Like for GSM, there are power bursts, in this case at 100 Hz and harmonics.

1.6 Bluetooth
Bluetooth is a high-speed, low-power wireless link technology operating in the 2400 MHz band typically used to connect phones, laptops, personal organizers, printers, and other portable devices. There are three possible power levels, 100, 2.5 and 1 mW with bursts frequencies of 1600 Hz and harmonics. Devices with a peak output power of 100 mW have to implement a power control to limit the output power to typically less than 2.5 mW. Like WiFi, the actual average output power depends heavily on data throughput. In most cases, the average output power is at least a factor of 10 smaller than the reported maximum output power levels. The low power levels and similarities to DECT and GSM justify that no further studies are performed.

1.7 Radio Frequency Identification (RFID)
RFID is a fairly recent technology, which is assumed to soon migrate to an even wider range of applications, where objects (e.g. pets, luggage, people) are equipped with an integrated unique RF tag that communicates with a remote reader. Many different frequency ranges are used, the common one being the 866 MHz band. In many cases, the tag has no internal energy source, and all the necessary energy for communication, typically 2 W, comes from the reader. A short burst of energy is often sufficient for establishing the identification.

Choice of signals for biological experiments

Based on a thorough analysis, it was decided that the GSM, UMTS, WiFi, and RFID communication systems would be investigated. Representative signals were created for both in vivo and in vitro exposure systems, the details of which are described below.

2. Assessment of Exposure

2.1 Measurement Techniques, Instrumentation and Procedures

With the introduction of modern communication technologies in our everyday life, the mobile communication industries and government agencies have been confronted with the task to accurately assess human exposure to EM fields from wireless devices. At the beginning of the SEAWIND project, existing measurement equipment had been optimized for only 1G and 2G wireless technologies, which limited accuracy for exposure assessment of modern communication technologies. In the course of the project, we have conducted research and implemented methods to overcome these insufficiencies and to re-establish measurement uncertainties for internationally unified standards.

2.2 Fields in the Environment or Incident Fields

At the beginning of the SEAWIND project, there existed several reports in the literature in which exposure levels had been studied in different countries and for different scenarios by means of personal exposure meters (PEMs). The disparate methodologies used in these investigations and the considerable variation in some of the results render the findings to be of limited value. It is clear that, to be able to reliably compare exposure levels across Europe, the same measurement methodologies (protocols) must be used. Campaigns in Belgium and Greece were initiated calling for the same procedures and, in some cases, even the same equipment models, to be employed.

2.3 Fields in the Body

Although incident fields can give an indication of exposure level for comparison to reference levels defined in exposure guidelines, the actual physical quantities related directly to biological effects are the electric and magnetic fields inside human tissues that cannot be readily measured. There are two approaches to the assessment of these fields, namely, via experiment and numerical dosimetry. The former entails the use of phantoms that represent worst-case conditions of maximal exposure, and the latter can be used to simulate realistic exposure situations from which results can be deduced and statistically analyzed. Before the SEAWIND project, the majority of research efforts focused on dosimetric characterization of mobile phones for voice communication, i.e. next to the user's head. However, mobile phones are being replaced by smart phones and laptops by tablets that are used mainly for data communication (including VoIP for conversation) and tethering (i.e. as an access point to the internet). In SEAWIND, the use of these devices as terminals for wireless data transfer with currently available technologies, as well as with emerging telecommunication protocols (including those at higher frequencies), has been studied, whereby it was assumed that the devices are located next to the body rather than the user's head.

A total of eight human models of the Virtual Population (ViP) were used to represent both sexes and various age groups, i.e. Thelonious (male, 6 years), Billie (female, 11 years), Louis (male, 14 years), Duke (male, 34 years), Fats (obese male, 37 years), and, Ella (pregnant female, 26 years, in 3 gestational stages: 3, 7, and 9 months). The positions of the sources (quarter-wavelength dipoles) were chosen to simulate real life exposure situations:

- The wireless device placed inside the trousers pocket (back and front sides); in these cases, the generic source was 5 mm away from the body
- The wireless device placed inside the jacket pocket; in this case, the source was placed 20 mm away from the body
- The wireless device placed in a bag (backpack or 'side-carry' bag); in these cases, the source was placed 50 mm away from the trunk of the body.

2.4 Guidelines for Exposure Minimization

The research results also allowed derivation of guidelines on how exposure to wireless networks can be minimized. In the close vicinity to a wireless network device, exposure is dominated by the device itself and is only minimally influenced by the environment. Farther away, the environment is an important exposure factor, i.e. diffuse incident EM fields from an access point (base station) in the room determine the exposure.

2.4.1 Body-Mounted Devices

Both measurement and simulation results show that there are several critical factors that affect personal exposure, i.e. the fields induced in the body, which are, for the sake of discussion, expressed in terms of energy absorbed in tissues by body-mounted devices.

2.4.2 Access Points

In the SEAWIND project, we investigated the exposure to access points (base stations) of wireless networks by measuring the fields emanating from them. It is clear from the measurement results that, for distances of up to 1 m, the EM source determines the field strength and exposure level. It is known from EM theory that the decay of an electric field close to a RF source is inversely proportional to the distance from the source raised to the power of 2 3. This behavior of the electric fields was confirmed by measurements during SEAWIND. The impact of the orientation of the antenna is large, as well. Close to the access point, there are hardly any depolarization effects, meaning that, when the antenna of an access point is vertically oriented, the electric field values in the vertical direction are much higher than those in the horizontal direction. Therefore, fields couple better to standing persons. Beyond a distance of 1 m, the reflections in the room produce a depolarization effect, i.e. the energy carried from the source to a person in the room is the same for both polarizations.

3. Assessment of Genotoxicity and the potential contribution of Oxidative Stress

It is well established that higher-energetic EMFs such as the ultraviolet (UV)-component of sun light or ionizing radiation have a potential to directly damage the chemical structure of the DNA. The spectrum of damage induced is wide and includes a variety of DNA base modifications, crosslinks and strand breaks.

3.1 Exposure Systems

The history of bioelectromagnetic research, i.e. research on the interactions of EMFs with biological systems, is scattered with reports of effects that were notoriously refractory to independent reproduction by other independent researchers, thus casting doubt on whether any effect was there at all. Unfortunately, experiments and exposures were not sufficiently well controlled or described, such that the precise exposure conditions were poorly defined; furthermore, artifacts and stray exposures were not properly accounted for or characterized, or were minimized, adding to the uncertainty surrounding the observed effects. With this in mind and with reference to the literature, we set out to design experimental approaches and equipment to perform in vivo and in vitro genotoxicity screening according to the best practices. The systems we used provided the tools to allow well-characterized and controlled exposure of cells and rodents. Of particular note is the live cell imaging system, which allowed for the first time a direct, real time insight into cellular response to wEMFs of different modulation characteristics. This application of this system is likely to generate a significant impact in wEMF research. Whereas the types of signals present from GSM and 3G mobile phones are easily characterized and show specific characteristics in terms of their temporal signal variations, signals from WiFi, RFID, and other signals display more random variation as a function of the data traffic being sent over the link.

3.1.1 In Vitro Exposure Systems

In vitro exposure systems are used for controlled experiments with cultured cells. Existing systems were equipped with the new signals for emerging wireless technologies, namely for WiFi and RFID and delivered to Fraunhofer ITEM in Hannover and to the University of Basel. Two exposure units were installed in a cell culture incubator to control accurately the ambient environment. It is well known that changes in culture conditions (e.g. temperature) can have a profound effect on human cells and hence must be minimized such that the EM interaction can be evaluated. The systems were designed such that the maximal difference between sham and exposure is always less than 0.1K. An additional consideration of importance is that the experimenter should not know which of the two chambers is exposed and which remains unexposed (also referred to as sham).

3.1.2 In Vitro Exposure System for Live Cell Imaging

The live cell exposure system is a miniature RF exposure system normally operating at 2450 MHz, which is integrated with a live cell imaging microscope. State-of-the-art microscopic techniques, including confocal scanning and fluorescence microscopy can be performed, allowing live cell observation during exposure. In contrast to the classic approach of post exposure analysis of cells, the instrument enables direct insight into the cellular response to EMF in real time. This computer-controlled setup allows different signals to be selected, as well as the monitoring of exposure and environmental conditions. The system also provides the ability to blind the experiments such that the experimenter does not know if a given exposure in a sequence is sham or exposed.

3.1.3 In Vivo Exposure Systems

An in vivo exposure unit was designed to provide an efficient and flexible setup with maximum exposure homogeneity for mice housed singularly within standard cages. Homogeneity of the exposure is required for an accurate prediction of what each individual mouse has been exposed to and, hence, to relate any possible effects to a given dose. Exposure to SARs well beyond those seen in real life scenarios was made possible to maximize the chance of inducing some effect, however always taking care not to expose at too high a level such that thermal effects become dominant. Exposure was at 2450 MHz with the various modulation schemes defined and agreed on by the consortium. As well as monitoring the EM exposure, environmental parameters were also recorded throughout to have a full record. The system consisted of two reverberation chambers that could be allocated by the user to sham (no exposure) and expose or both to different exposure levels.

3.1.4 Dosimetry

Central to the analysis of experimental results is knowledge of the dose not just the overall dose but also the dose to individual tissues or organs of the mouse. Dosimetry was performed to determine how the dose varies according to the different tissues and anatomy inside the mouse. To achieve this, a male mouse model with a weight of 28 g was used for the numerical dosimetry by computer modeling, and a simple homogeneous phantom was used to verify the results experimentally. Thermal modeling was also performed to assess temperature rise in different tissues.

3.1.5 Exposure Signals (Rationale and Description)

Exposure was performed with various modulation schemes, including CW, GSM, UMTS, WiFi, and RFID. These modulations were chosen from the range of possible signals as having the characteristics that best represent today's variety of exposures, whereas the average power and the ELF power envelopes were synthesized to maximize the likelihood of engendering biological effects, based on current best hypotheses regarding interactions of weak RF exposures with organisms.

CW, continuous wave, simple sinusoidal signals of constant amplitude, phase, and frequency that do not, as such, convey any information, was included as a form of active sham exposure, i.e. without ELF spectral components of the power envelope but generating the same temperature load.

3.2 Genotoxicity Testing in cultured human cells (in vitro)

Evidence for a potential genotoxic effect of wEMF at 0.1 to 2 W/kg was obtained mostly from the so-called "Alkaline Comet assay", a classical experimental test to evaluate and quantify the level of DNA damage, in particular DNA strand-breaks and alkaline labile sites, in cell populations. In the SEAWIND project, we made use of this assay to re-visit the potential of wEMFs to induce DNA damage in in vitro cultured human cells. In addition to the standard test, we applied a modified version of the assay that allows a specific and more sensitive detection of oxidative DNA base lesions. Moreover, to address potential wEMF effects more comprehensively, the SEAWIND project not only tested previously reported observations but extended the investigation to a wider range of wEMF signals and modulations that have become important for wireless data transfer in recent years and to the exploration of potential co-genotoxic effect.

The first series of experiments focused on the replication of previously reported observations with the alkaline Comet assay that indicated an induction of DNA damage by mobile phone-specific UMTS and GSM signal modulations in primary human fibroblast and immortalized human trophoblast cells (HTR-8/SVneo cells). Despite considerable efforts to optimize the sensitivity of our assays, however, these effects could not be confirmed in the two SEAWIND partner laboratories. Likewise, the subsequent systematic testing of potential effects of the WiFi, RFID and CW exposure signals in the Comet assays failed to produce evidence for DNA damage induction.

To validate these results and to address the hypothesis that wEMFs might alter intracellular levels of reactive oxygen species (ROS), we systematically assessed the wEMF signals for induction of oxidative DNA damage. For this purpose, we applied the enzyme-modified version of the Comet assay, designed to specifically detect a common oxidative base modification with high sensitivity. As the standard alkaline Comet analyses, however, this assay produced no evidence for increased oxidative DNA base lesions in the two cell types tested. Consistent with this lack of effect in the enzyme-modified Comet assay, we did not observe any EMF-induced change in intracellular ROS using a highly-sensitive, fluorescence microscopy-based method that traps the short-living ROS and converts it into a more persistent fluorescence signal. Notably, this assay was negative, even if it was performed under 50Hz EMF exposure that previously produced a measurable effect in the alkaline Comet assay. Thus, as both the intracellular ROS detection and the oxidative Comet assay are very sensitive methods, we conclude that ROS-mediated oxidative DNA damage is not an important contributor to potential genotoxic effects related to wEMF exposure under the experimental conditions.

The development and application of standardized experimental procedures, including the blinded exposure and evaluation as well as the independent confirmation of key findings in the two partner laboratories, was an important conceptual strength of this part of the SEAWIND project, meant to avoid eventual inconsistencies in reproducibility that often occurred in this field of research. Minor variations of experimental conditions, due to differences in the technical infrastructure between laboratories, however, are difficult to eliminate and may have contributed to the contradictory outcome of similar experiments in the past. We addressed the issue of experimental variation, focusing on the consistency of results generated by different Comet analysis pipelines. This showed, not unexpectedly, that the methods used in the two partner laboratories produced slightly different results from otherwise identical experiments or even data sets. Thus, minor technical issues like this may add up in the multistep protocol of the Comet assay and culminate in an overall experimental variation that, if the effect under investigation is small, may produce small positive results in some analyses and negative outcomes in others. This is a negligible issue if the agent assayed is clearly DNA toxic and affects homogeneously most cells in a population with a consistent dose-dependency. EMF exposure, however, has never produced strong dose-dependent responses and the data underlying positive effects (e.g. ELF-EMF exposed human cells) suggest that only a subpopulation cells in a culture may be affected. All considered, the inconsistent nature of EMF effects in the Comet assay strongly suggests that this type of exposure does not damage the chemical structure of the DNA. It may, however, affect certain aspects of cell physiology in a way that alters the steady-state level of naturally occurring DNA strand breaks in some cells. Cells undergoing DNA synthesis, for instance, have intrinsically increased levels of DNA strand breaks that are picked up by the Comet assay, and there is evidence from ELF-EMF studies that the dynamics of such processes might be influenced under exposure.

In conclusion, considering both the results reported in the literature and the observations of the biological in vitro part of the SEAWIND project, there is no evidence for a direct DNA damaging potential of wEMFs. However, wEMF exposure might impact cellular processes that, in combination with other environmental stressors, could result in molecular readouts resembling those of genotoxins. Such effects could explain the occasionally reported positive result with the Comet assay.

3.3 Genotoxicity Testing exposed mice (in-vivo; ex-vivo)

To address persistent in vivo genotoxicity and co-genotoxic effects of wEMF exposure in living animals, this part of the SEAWIND project examined micronucleus (MN) formation in bone marrow and peripheral blood erythrocytes (PB) and also in keratinocytes of mice. Male B6C3F1 mice were exposed to the 1.95 GHz CW and the three wEMF signals UMTS, WiFi, and RFID. For each signal, exposure levels in terms of average whole body SAR of 0 (i.e. sham), 1.6 4.0 and 10 W/kg were used. Groups of 6 mice were exposed in reverberation chambers for 2 weeks, 20 h/d to the CW and wEMF signals and SAR levels. In addition, the well-characterized cyclophospamide monohydrate (CP) was used for co-exposure experiments with 1.95 GHz CW signal to investigate a potential co-mutagenic effect. Finally, malondialdehyde (MN) levels were evaluated in the erythrocyte fraction as a measure for lipoperoxidation, a signal of oxidative stress.

3.4 Development and Application of Novel Assays

Beyond applying classical cytogenetic tests to assess the genotoxic potential of EMFs, an important objective of the SEAWIND project was the development and application of advanced experimental procedures to elucidate putative interactions between wEMFs and DNA-related cellular processes.

DNA modifications including strand-breaks naturally occur through a variety of DNA transactions associated with cell metabolism and proliferation. In general, these are efficiently repaired and therefore not detectable as DNA damage or genetic alterations by classical tests that require the DNA repair capacity to be saturated by induced DNA damage. Thus, modulation of DNA repair capacities either genetically or by chemical inhibition of repair enzymes may increase the sensitivity of these assays towards minor changes in the steady-state of endogenously occurring DNA lesions, e.g. by accumulation of unrepaired DNA damage. As EMFs were proposed to induce low levels of DNA single strand-breaks and ROS-triggered oxidative DNA damage, we considered the inactivation of respective DNA repair activities a promising approach towards enhancing the sensitivity of the alkaline Comet assay for the detection of potential wEMF damage.

To this end, we evaluated a series of PARP1/2 inhibitors with respect to cytotoxicity and impairment of DNA repair in the two cell models used in the SEAWIND project. The activity of PARP proteins play a key regulatory role in the recognition and processing of DNA base damage and strand-breaks. Alkaline Comet assays performed with wEMF exposed and PARP inactivated cells indicated a slight accumulation of DNA damage in UMTS exposed cells, while the GSM, WiFi and RFID signals had no such effect. Although both, the signal dependency of this effect and the question whether it indeed reflects DNA damage accumulation requires confirmation, these results underline the feasibility and power of the modulation of DNA repair capacity for future research into the potential DNA directed effects of EMFs.

Effects of EMFs on cellular processes appear to be generally subtle and transient, impeding any kind of endpoint analysis. The SEAWIND project intended to develop equipment and experimental procedures to investigate molecular and cellular responses in real-time under wEMF exposure, allowing the detection of small and transient effects. We thus designed, developed and applied a wEMF exposure chamber for live cell imaging microscopy (sXclive-2450) as described under "Exposure Systems" (see above).

One application of such a tool is the tracing of DNA damage by detecting the transient appearance of DNA repair proteins in focal nuclear structures representing sites of ongoing repair, so-called repair foci. To explore the potential of such an approach, we established human cells lines expressing a fluorophore-tagged XRCC1 protein. XRCC1 is a key regulator of the repair pathway that fixes DNA single-strand breaks and oxidative DNA base modifications and is known to be rapidly recruited to sites of damage. Remarkably, we noticed that the monitoring of the XRCC1 repair activity under the fluorescent microscope significantly increased foci formation, suggesting that the excitation of GFP-fluorescence at about 488nm is sufficient to induce simple DNA base damage and strand-breaks. While this is an observation to be taken seriously by the research community, as the technology has been widely used to address the kinetic and dynamic aspects of cellular DNA damage response, the high background of repair activity induced by the observation made an assessment of transient induction of repair foci in wEMF-exposed cells impossible. Nevertheless, the results highlight the sensitivity and the feasibility of live cell imaging of cells under EMF exposure, which can be applied to monitor virtually any cellular process that can be tracked microscopically (e.g. the real-time assessment of ROS formation described before). We also used the technology to investigated a putative impact of wEMF exposure on the dynamics of DNA repair processes, as changes in repair activities might explain transient disturbances of genome integrity as well as co-genotoxic effects described in some studies.

4. Risk Communication

4.1 Insights about Concerns and Risks from Lay People

Eight focus group exercises were conducted 4 in Switzerland and 4 in Greece with lay people to gather insights about concerns, risks, and benefits of wireless network devices. A focus group is a moderated and structured process of debate involving some 6 to 15 participants representing a social group or a relevant selection of stakeholders, designed to explore the social resonance of arguments for and against certain controversies or value conflicts, to better understand patterns of perception and potential concern, and to anticipate demands for communication and information. In total, about 90 persons attended the focus groups in Switzerland and Greece. Please note that these results are not representative!

In their assessment of sources, channels, and information requests, the focus groups in Switzerland and Greece revealed five user types: (1) sceptical users, (2) 'convinced' frequent and pragmatic careless users, (3) interested frequent users, (4) adaptable users, (5) low-tech non frequent and non-users.

The focus groups also produced basic recommendations on an integrated policy of communication. Focus group participants demonstrated a preference for a centralized information source integrating different independent (scientific) sources and channels, and authorized by a trustworthy central institution (such as the European Union). The provider should also be obliged to provide reliable information about potential health risks and recommend precautionary measures. The information supplied by the provider merely constitutes one element of an integrated communication policy.

The participants were of the opinion that political decision-makers should be encouraged to initiate a set of international and long-term studies on the issue.

Independent international studies by scientists were judged by the respondents to be the most reliable source. This was true for all types of users. Studies should be written or translated into everyday language accessible to a lay readership. The information should also be published on the website of every EU member country.

The participants would also appreciate risk comparisons with other familiar risk situations (i.e. smoking, use of mobile phones).

The participants of the focus groups expected any communication to address the different types of users mentioned above, with the message targeted to the needs of each of the user groups (by modifying message framing, channel, frequency, information type, language).

The scope of communication favoured by the participants ranged from a) providing full information on technology, spatial diffusion, exposure-related health effects, exposure, protective action (including precautionary measures), public policies and standards, to b) limited communication about health effects, or only when health risks are clearly indicated (supported by 'pragmatic careless users', who feel unable to change their habits without good reason).

The highest consensus on communication strategy was achieved in the focus groups with relation to the integration of health information into a single format, including physical risks, psychological online addiction, data security, social exposure, pressure, social fragmentation, monitoring juvenile Internet use, etc.

The focus groups also made some demands and offered concrete political recommendations, e.g. providers should not only release information on risks, but should also be encouraged/obliged by law to reduce exposure. Participants also suggested that school curricula should include information (similar to traffic safety information) on the risks associated with wireless network devices (health as well as other).

Other suggestions included that the best knowledge available be summarized in the form of a leaflet and / or a YouTube video sponsored by the EU Commission, providing EMF measurement tools (i.e. exposure calculators) in public places such as schools and online, installation of a common regional network (there should be one single wireless network per region instead of a variety of overlapping signals), and production of urban risk maps featuring telephone antennas, high voltage power lines, and radio amplifiers, not to mention publication of information brochures and online articles.

Several focus groups had the chance to examine a web tool created by iMINDS for the prediction of indoor exposure.

4.2 Communication Strategy and Strategy to Communicate Uncertainty

DIALOGIK conducted a group Delphi to devise the best strategy for communicating the IARC 2b classification of agents (c.f. International Agency of Research on Cancer (IARC) 2012). The group Delphi was also designed to develop approaches on how to communicate uncertainty to a lay audience.

It must also be mentioned here that the results of a group Delphi cannot be representative; the statistical values simply serve as a guide for discussion. In the SEAWIND Delphi, 14 experts participated in the group (in total, over 130 experts were requested to participate in the group Delphi)! These participants were experts in EM compatibility, technology assessment, radiation protection, risk research, EMFs and microwaves, public health, information technology and society, medical radiation biology, future studies and technology assessment, mobile communication, etc.

Between the participating experts there was broad consensus (1) that public authorities, health related organizations and scientists, and consumer associations are the most relevant communicators to supply the public with risk classification information.

There was also broad consensus during the first round of the group Delphi (2) that national agencies and scientists should cooperate. During the second round, the experts argued that the national agencies and scientists should be complemented by a tandem of scientists and journalists to ensure that expert information is presented in an understandable and comprehensible way. The experts also agreed about (3) the need for risk avoidance/benefit recommendations that should form the content of the communication. Regarding the objectives of the communication (4), there was consensus between the participating experts about the need for transparency when evaluating and managing risks. They also agreed that there is no need to present just one position, and instead favored a presentation of the arguments for each of the conflicting positions (with strong emphasis on communication of uncertainties).

For the experts, the most effective communication channels (5) depend on the communication objective and the target group. Effective communication channels could be websites, press conferences, scientific publications, TV, and radio. In terms of credibility (6), there was consensus that the most credible communication channels are scientific publications and information supplied by the relevant authorities.

According to the experts, risk awareness (7) can be enhanced by apps and product labelling. Mobile measurement instruments were rejected as inappropriate, so more research is needed here.

The last question (8) dealt with whether uncertainty should be communicated at all. In this instance, the participating experts were unable to reach agreement, and concluded that uncertainty should not always be communicated. However, there was consensus that uncertainty is a key element of risk communication, and that it is important to achieve a balance between the desire to provide people with the information necessary to enable them to evaluate and assess risks on their own, and the potential for causing undue fear, incurring more damage by cementing risk aversion, which, in turn, can lead to missed opportunities.

The qualitative studies of this part of the project aimed at understanding perceptions about, concerns associated with, and the need for information on wireless network devices and the corresponding technological infrastructure. They focused on two specific characteristics: a) that devices are frequently used by large sections of the population and even have an impact on those not engaged in active use and b) that the uncertainty surrounding long-term health risks is ongoing.

This constitutes an extraordinary challenge to political planning, both in terms of risk governance and risk communication. The main strategic communication recommendation foresees integration of information from a unified, independent, and credible source, and tailoring information and communication programmes to the needs of the different user types. These recommendations exceed the context of SEAWIND technologies, since the same communication problems also apply to various other technologies that remain shrouded in uncertainty, such as nano-particles, genetically modified organisms, dosimetry, and others.

Many experts felt that the general public's sense of risk awareness was underdeveloped. Some people overestimate while others underestimate risk, and there is a lack of judgement when it comes to appropriately balancing risks and benefits.

During the group Delphi, two opposing opinions emerged, concerning how and when uncertainty should be communicated: one group favored full disclosure of comprehensive information, including guidelines for prevention, while the other favored a more limited information approach, whereby all information that could cause unnecessary fear and worry is filtered. Which of these two opposing strategies is more appropriate is likely to depend very much on the individual situation. It is important to achieve a balance between causing undue worry and concern on the one hand and a careless approach to risk on the other. Communication should aim to provide people with the background information, thus allowing them to judge for themselves how much protective action they need/want to take.

Potential Impact:
1. Impact

The potential impact of SEAWIND is expected to be on several levels, the societal impact being the most important. Until recently, research has concentrated on mobile phones while less attention was paid to the pervasive exposure of wireless local or metropolitan area networks, body-mounted and body-worn wireless personal area network devices, and specific wireless applications in industry, e.g. novel RFID logistics applications. However, that people are increasingly exposed to these signals during their daily life causes considerable public concern about the safety of these technologies. The most significant impacts are seen in the following areas:

1.1 Exposure Signals

The exposure signal has been analyzed, in particular with respect to daily-life and maximum exposures. The main characteristics and differences have also been summarized in layman terms. This will allow technical experts, communication experts, politicians, and the public to rationally discuss the different technologies (Work Package 2: Review of communication systems, signals, and power modulations).

Based on this analysis, signals designed and applied (Work Package 7: Genotoxicity screening in vivo and in vitro) well represent the technology but also maximize the likelihood of generating effects. The rationale is also provided. It is suggested that these signals shall be used or at least considered for any future biological experiments.

1.2 Measurement Technology

The methodologies and instrumentations were developed to reduce the uncertainties of exposure assessment (Work Package 3: Development of instruments and calibration techniques). These results were disseminated to the standards agencies (Work Package 9: Dissemination to standards) and have already been adopted. This enables reliable exposure assessments that are of great importance for industry, regulators, and health agencies for industry. It also had a significant economic impact for the participating SME, as it could demonstrate its lead in providing exposure assessment technologies.

1.3 Incident and Induced Field Exposures

For the first time, the exposures due to wireless networks were systematically analyzed. The spatial and temporal RF exposures at typical indoor microenvironments (schools, crèches, offices, and homes) were measured in Belgium and Greece. Furthermore, methods to extrapolate instantaneous exposure to maximal daily exposures were developed. These values are important today to enable quantitative exposure values (Work package 5: Dosimetry for worst case partial body and local exposure and 6: Organ specific dosimetry).

More important is the novel propagation model that enables estimation of the exposures inside closed rooms. This model will be widely used in the future for exposure estimations and wireless network optimizations (Work package 4: Incident field evaluations for whole-body exposure).

Also for the first time, a comprehensive analysis for induced fields was performed. This allows estimation of the maximum tissue and organ-specific exposures based on frequencies and antenna input power and distance or incident field strengths. This not only enables assessment and comparison of the exposures related to today's technologies but of any future technologies (Work package 6: Organ specific dosimetry).

All of the above findings have also been experimentally validated.

1.4 Exposure Systems

Several exposure systems (in vitro system, novel table-top reverberation system for mice, and a live-cell imaging system) were developed or have been adapted for the purpose of this study. In particular, the live imaging system, which allows direct real-time insight into cellular response to EMFs of different modulation characteristics, which could have a significant scientific impact, is the first of its type to be reported. The output from Work Package 8 (Exposure systems and quality control) was instrumentation that was exploited directly by the partners in the project. Furthermore, the new equipment was presented at the foremost conference for bioelectromagnetics, to allow other researchers in the field to share the new developments. It is expected that one of the partners will offer a version of these systems for investigations of impact of RF on other biological endpoints.

1.5 Genotoxicity Screening In Vivo and In Vitro.

Based on epidemiology and supported by some laboratory studies, EMFs were classified as possibly carcinogenic (2B) by IARC. Although many cellular pathways may promote the formation of cancer, agents that directly attack the DNA exhibit in general a carcinogenic potential. Therefore, the SEAWIND project aimed at systematic re-evaluation of the potential impact of wEMF signals on genome integrity, which was previously very controversially discussed in both the scientific and the public community. By carefully controlled experimental in vivo and in vitro systems and approaches, previously reported induction of DNA damage by mobile phone-specific signals could not be reproduced. In addition, there was no indication for a direct DNA-damaging potential of the newly explored signal modulations used in modern data transfer technologies. However, our investigations revealed novel hints about modulation-specific multifactorial effects and how EMF may interfere with cellular homeostasis. These results clarify and advance the scientific understanding about potential health impacts of EMFs and, in particular, will stimulate and guide future investigations into the role of EMFs as a putative co-carcinogen or co-stress factor. Such research will be supported by the novel tools developed during the SEAWIND project. Hence, the outcome of the present project will direct future biological investigations towards a better understanding of the interactions between EMFs and biological systems, which will facilitate the risk assessment concerning health effects of EMF exposure (Work package 7: Genotoxicity screening in vivo and in vitro).

1.6 Risk Governance: Integrating Assessment, Perception and Communication

In communicating uncertainty, the dilemma of finding the appropriate balance between undue precaution (rejecting a beneficial technology) and carelessness about risks (focusing on benefits only) must be addressed. The panelists recommended: Integrate information about health risks in a broader information context about social risks and benefits of WLAN and mobile internet use in general, take advantage of different channels, vary the frequency with which information is given, and experiment with different formats and frames in line with the needs and concerns of the various user-typed identified above. The focus groups also recommended establishing a centralized structure for authorized independent information (e.g. by EU institutions).

Furthermore, on the practical side, the EU was advised to produce YouTube videos on the issue and to consider establishing awareness programs at schools. These measures should be augmented by installing radiation measurement devices in computer rooms, to create exposure and risk maps about radiation exposure, and to invest in a single common wireless network to minimize exposure to radiation from overlapping EMFs caused by a variety of parallel wireless networks.

1.7 Standards

The following standards greatly benefit from the outcomes of this project:

- IEEE 1528: Recommended Practice for Determining the Peak Spatial-Average Specific Absorption Rate (SAR) in the Human Head from Wireless Communications Devices: Measurement Techniques (adopted the methodology developed in Work Package 3)

- IEC 62209-1: Human Exposure to Radio Frequency Fields from Hand-Held and Body-Mounted Wireless Communication Devices: Human Models, Instrumentation and Procedures - Part 1: Procedure to determine the specific absorption rate (SAR) for devices used in close proximity to the ear (frequency range of 300 MHz to 6 GHz). (adopted the methodology developed in Work Package 3)

- IEC 62209-2: Human exposure to radio frequency fields from hand-held and body mounted wireless communication devices Human models, instrumentation, and procedures Part 2: Procedure to determine the specific absorption rate (SAR) for wireless communication devices used in close proximity to the human body (frequency range of 30 MHz to 6 GHz) (adopted the methodology developed in Work Package 3)

- IEEE C95.3: Recommended Practice for Measurements and Computations of Electric, Magnetic and Electromagnetic Fields With Respect to Human Exposure to Such Fields, 0 Hz to 300 GHz (is currently under revision and adaption of the methodology is also tabulated)

2. Dissemination

The SEAWIND dissemination aimed to promote knowledge sharing among the scientific community and standardization bodies and to increase awareness of the project results on the part of the public. Various instruments were and are used to reach that goal.

2.1 Project website

A project website was established at the beginning of the SEAWIND project and was continuously updated during the course of the project. The purpose of the website was to exchange information and share confidential data within the consortium as well as to promote the project and its activities to the wider scientific and user communities. It, therefore, consisted of public and private sections. The public section comprised all material accessible to the general public, whereas the private section is intended for the internal organization of the project and could only be accessed via login with a username and password. The structure of the SEAWIND website at http://www.seawind-fp7.eu was designed in a clear and consistent way so that visitors and users could easily locate all information intended for them. The main sections were listed in the left navigation pane.

In the public domain, the opening page referred to 'News and Events' to clearly highlight the latest developments regarding the project and project-related issues. Events, such as a workshop in which the SEAWIND experts reported project results in the course of a public event, were announced there. Detailed information about the project aims, the deliverables of the project and management structure, the consortium partners, and external advisors were provided on linked pages.

2.2 Project leaflet

A project leaflet written in generally understandable language, containing a summary of the main objectives and methodology and the project partners, was developed. The leaflet aimed not only to promote the project to the public, but was also used to inform people who were affected by the studies, e.g. pupils (and their parents) of the schools and other facilities where measurements were performed. For that purpose, the leaflet was translated also in Greek. The leaflet is available for download from the project website and was distributed in printed form at conferences and other events during the lifetime of the project.

2.3 Dissemination towards the scientific community and the general public

Dissemination activities of SEAWIND towards the scientific community included various channels, such as participation in scientific meetings, conferences, and workshops, publications, and liaisons with other research projects. The SEAWIND project and its results have been presented at around 20 national and international scientific meetings, conferences, and workshops in the form of posters and oral presentations and distribution of the project leaflet.

Dissemination of the project results to the scientific community and industry and general public has also been established via a workshop. The IT'IS Foundation, coordinator of SEAWIND, organized a scientific workshop covering, among other topics, SEAWIND-related research results. The workshop EMF Health Risk Research - Lessons Learned and Recommendations for the Future 7 years later took place in Autumn 2012 (October 21 26, 2012) at the Center Stefano Franscini, in Monte Verità, Ascona, Switzerland. The workshop brought together world-renowned researchers in the field of EMF and health as well as government health protection experts and standardization committees to analyze and synthesize newly available research results. It provided a forum for extensive discussions that has been available neither at scientific meetings nor at official standard/assessment group meetings, namely, to focus on those experiments that are not in line with current understanding of EMF interaction with biological systems. In the frame of the workshop, a public event was held for local people entitled Health Risk from of Exposure to Wireless Network Devices?, which focused on SEAWIND research. For the first time, the principal investigators of the SEAWIND consortium presented the results and conclusions of the project to the general public. The researchers and representatives of health agencies were available to answer questions asked by the public The event has been videotaped for dissemination and is available on the SEAWIND webpage (http://seawind-fp7.eu/deliverables-and-publications/) the Monte Verita webpage (see http://www1.itis.ethz.ch/mv-2/ online) and other related sites.

2.4 Contacts and co-operation with other European projects

The SEAWIND consortium is closely liaised to the European Framework Programme 7 project ARIMMORA ('Advanced Research on Interaction Mechanisms of electroMagnetic exposures with Organisms for Risk Assessment'), which deals with the exploitation of biophysical mechanisms that could explain the effects of weak environmental extremely low frequency (ELF) fields in support of a possible causal relationship between cancer and ELF magnetic field (ELF MF) exposure. Four SEAWIND Partners are also members of the ARIMMORA consortium.

2.5 Dissemination to standards

The objective of Work Package 9 ('Dissemination to Standards') was the direct dissemination of the results of the Work Package 2 ('Review of communication systems, signals, and power modulations'), Work Package 3 ('Development of instruments and calibration techniques'), Work Package 4 ('Incident field evaluations for whole-body exposure'), Work Package 5 ('Dosimetry for worst-case partial-body and local exposure'), Work Package 6 ('Organ specific dosimtery') to the relevant standard committees, with the advantage that the findings could be evaluated at the earliest possible opportunity by academic, industrial, and governmental experts and adopted as soon as possible. The dissemination work began in early 2010, with presentation to the standards committees, first of the open questions and then the results. The work has a direct impact on measurement standards where complex signals must be accurately measured.

List of Websites:
http://seawind-fp7.eu

Final Report Summary - SEAWIND (Sound Exposure and Risk Assessment of Wireless Network Devices)

Related documents

Share this page

Download