Bioelectromagnetics 35:210^221 (2014)
Assessment of Extremely Low Frequency Magnetic Field Exposure From GSM Mobile Phones n,1* Darren Addison,1 Terry Mee,1 Richard Findlay,1 Carolina Caldero 1 Myron Maslanyj, Emmanuelle Conil,2 Hans Kromhout,3 Ae-kyoung Lee,4 geVarsier,2 Malcolm R. Sim,5 MasaoTaki,6 Nade 2 JoeWiart, and Elisabeth Cardis7,8 1
Centre for Radiation, Chemical and Environmental Hazards, Public Health England, Chilton, UK 2 Whist Lab, Paris, France 3 Institute for Risk Assessment Science, Utrecht University, Utrecht,The Netherlands 4 RadioTechnology Research Department, Electronics and Telecommunications Research Institute (ETRI), Daejeon, Korea 5 Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Victoria, Australia 6 Department of Electrical Engineering,Tokyo Metropolitan University,Tokyo, Japan 7 Centre for Research in Environmental Epidemiology (CREAL), Barcelona, Spain 8 Biomedical Research Centre Network for Epidemiology and Public Health (CIBERESP), Barcelona, Spain Although radio frequency (RF) electromagnetic fields emitted by mobile phones have received much attention, relatively little is known about the extremely low frequency (ELF) magnetic fields emitted by phones. This paper summarises ELF magnetic flux density measurements on global system for mobile communications (GSM) mobile phones, conducted as part of the MOBI-KIDS epidemiological study. The main challenge is to identify a small number of generic phone models that can be used to classify the ELF exposure for the different phones reported in the study. Two-dimensional magnetic flux density measurements were performed on 47 GSM mobile phones at a distance of 25 mm. Maximum resultant magnetic flux density values at 217 Hz had a geometric mean of 221 (þ198/104) nT. Taking into account harmonic data, measurements suggest that mobile phones could make a substantial contribution to ELF exposure in the general population. The maximum values and easily available variables were poorly correlated. However, three groups could be defined on the basis of field pattern indicating that manufacturers and shapes of mobile phones may be the important parameters linked to the spatial characteristics of the magnetic field, and the categorization of ELF magnetic field exposure for GSM phones in the MOBI-KIDS study may be achievable on the basis of a small number of representative phones. Such categorization would result in a twofold exposure gradient between high and low exposure based on type of phone used, although there was overlap in the grouping. Bioelectromagnetics 35:210–221, 2014. © 2013 Wiley Periodicals, Inc. Key words: EMF; ELF; magnetic flux density; mobile phones; exposure assessment; MOBIKIDS
Grant sponsor: European Community’s Seventh Framework Programme (FP7/2007–2013); grant number: 22687 3 – the MOBIKIDS project, a study on communication technology, environment, and brain tumours in young people.
*Correspondence to: Carolina Calderon, Centre for Radiation, Chemical and Environmental Hazards, Public Health England, Chilton, Didcot, Oxon OX11 0RQ, UK. E-mail: [email protected]
Conflicts of interest: Prof. Masao Taki’s department received a grant to support numerical modeling work under a universityindustry partnership. Whist Lab is funded by France Telecom. None of this funding was used to support the research described in this paper.
Received for review 3 September 2012; Accepted 20 September 2013
2013 Wiley Periodicals, Inc.
DOI: 10.1002/bem.21827 Published online 6 November 2013 in Wiley Online Library (wileyonlinelibrary.com).
Assessment of ELF Magnetic Fields From GSM Mobile Phones
INTRODUCTION Mobile phones have become increasingly popular, particularly among adolescents over the last decade, and there has been much interest in the possible health effects associated with the radio frequency (RF) electromagnetic fields emitted from them. In 2011, the World Health Organisation International (WHO) Agency for Research on Cancer (IARC) classified radiofrequency electromagnetic fields as possibly carcinogenic to humans (Group 2B), based on an increased risk for glioma, a malignant type of brain cancer, associated with wireless phone use [Baan et al., 2011]. Mobile phones also produce extremely low frequency (ELF) magnetic fields, mainly due to the current being drawn from the battery during transmission. Global system for mobile communications (GSM) phones transmit pulsed RF signals with a frame duration of 4.615 ms and as a result the current drawn from the phone battery is also pulsed, and has a frequency spectrum with a fundamental peak at 217 Hz (corresponding to 1/4.615 ms) with associated harmonics. An additional 8 Hz signal is also produced due to the multiframe structure of GSM; this signal is around 25 times smaller than the 217 Hz component [Andersen and Pedersen, 1997]. Although several studies have measured the ELF magnetic flux density emitted by phones [Jokela et al., 2004; Tuor et al., 2005; Straume et al., 2007], the number of phone models assessed so far has been quite limited, making it difficult to assess whether ELF exposure from GSM mobile phones is dependent on particular characteristics of the phones. The current in the phone may depend on the circuitry elements of the phone (e.g., type of RF amplifier), its power efficiency, battery position and voltage. Therefore possible factors that could give rise to different levels and patterns of the magnetic field include manufacturer, shape of phone (e.g., bar, flip, or slide), and age of phone. This paper describes the two-dimensional magnetic flux density measurements taken from a sample of phones as part of the exposure assessment for the MOBI-KIDS international case–control study [Goedhart et al., 2011]. MOBI-KIDS aims to assess associations between the use of communication devices and other environmental risk factors and brain tumors in young people. The main objective of the electromagnetic field exposure assessment protocol is to develop exposure gradients for the RF and ELF electromagnetic fields arising from the use of mobile phones, as have been developed for the Interphone Study for RF exposure [Cardis et al., 2011]. The laboratory
measurements reported in this paper form part of a much larger exposure assessment exercise and are used to characterise the magnetic flux densities for GSM phones, and to identify representative phone models. These measurements will be used to fit an equivalent loop model to the representative phones data, and predict the distribution of magnetic field in the volume occupied by the brain and numerical modeling techniques will allow the induced current density to be estimated at the tumor location. Details of the full algorithm are outside the scope of this article and will be published elsewhere. The main challenge at this stage is to identify a small number of generic phone models that can be used for numerical modeling and to classify the ELF exposure for the many different phones reported in the study. MATERIALS AND METHODS Selection of Mobile Phones Mobile phones were collected from the various countries participating in the MOBI-KIDS study, through voluntary donations. Forty-seven phone models, released between 1997 and 2008 (Fig. 1), were used for the study. Twenty-two phones (47%) were from the most popular manufacturer in the period of interest. In terms of phone shape, 32 phone models (68%) were bar shaped, while 11 (23%) were slide phones and 4 (9%) were flip phones. In terms of battery type, 35 (74%) were lithium-ion (Li-ion), 8 (17%) lithium-polymer (Li-Po), and 4 (9%) were nickel-metal hydride (NiMH). All batteries except two had a voltage of 3.6 V (the other battery voltages were 2.4 and 4.8 V). Where original batteries were no longer functional, genuine replacements were sought but, if not available, generic substitutes were used. Initial Measurements of Magnetic Fields The variation in magnetic field 25 mm from the front of the phones was initially mapped using a proprietary two-dimensional scanner. The measurements over a two-dimensional grid were performed with a Bartington Mag03-MS1000 fluxgate magnetometer (Bartington Instruments, Witney, UK) (DC to 3 kHz,