Research results on possible effects caused by radiofrequency fields in children are limited because most of the studies published so far have focused on adults, rather than children. Mobile phone use is now widespread, even among younger children. If a biological risk due to mobile phone exposure is found, it might be greater in children because their bodies might be more sensitive to radiofrequency energy. The issue of a possible difference in sensitivity between adults and children begins with whether any difference exists physically in terms of electromagnetic absorption. This paper presents a review of recent publications on dosimetric comparisons between children and adults with respect to radiation from mobile phones. The issue of the health effects of mobile phone use is beyond the scope of the present review. Most of the dosimetry research on possible differences in power absorption between children and adults has been based on numerical modeling and analysis. The understanding of the results so far is presented and needed studies are described.
Public concerns about the possible health effects of mobile phones continue because mobile phone use continues to spead. The International Telecommunication Union has estiated that mobile-cellular subscriptions worldwide reached almost 7 billion by the end of 2014 . South Korea, in particular, has the highest mobile phone penetration rate in the world.
International scientific bodies have stated that no adverse health effects have been established by mobile phone use. The balance of evidence to date suggests that exposure to low level radiofrequency (RF) fields, such as those emitted by mobile phones and their base stations, does not cause detrimental health effects such as increased risk of cancer of the head [2-4].
However, the Ministry of Science, ICT and Future Planning, Korea introduced the system grading mobile phone models based on their specific absorption rate (SAR) values . This is construed as a governmental ‘precautionary approach’ to raised public concern since the World Health Organization/International Agency for Research on Cancer classified RF electromagnetic fields as group “2B” (possibly carcinogenic to humans) in 2011 .
The issue of whether children are more sensitive to electromagnetic fields emitted from mobile phones has been a hot topic among many researchers. However, only very limited research has specifically addressed the issue of a possible difference in sensitivity between adults and children.
One of key questions regarding this issue has been “Do children absorb more radiation power in their heads than adults do from mobile phones?” This is a simple question but very difficult to answer definitively because most dosimetric research compares the absorption of RF power in different individuals. In addition, the structures of the mobile phones examined differ among the various studies.
Nevertheless, some recent publications have examined possible age-related differences in exposure from RF radiation of mobile phones. These are reviewed in this paper to provide a better understanding of the current status of the relevant research.
Mobile phone technology was introduced in the early 1980s, but it was not widely used until the mid-1990s. Some researchers had become interested in possible age-related differences in exposure from RF radiation of mobile phones by the mid-1990s [7, 8]. Dimbylow and Mann  reported that the SAR values, which averaged over 10 g for infants, were comparable or in most cases lower than the values in the adult phantom, while Gandhi et al.  suggested a deeper penetration of electromagnetic energy and a higher 1-g SAR in a smaller head. This controversy continues regarding whether children absorb more RF energy in their heads than adults.
By the mid-1990s, mobile phone use started in earnest and at the same time medical imaging technology was achieving remarkable developments, with the support of high quality computer resources. The first studies used child head models that were linearly scaled down from a realistic adult model. The new advances enabled researchers to develop more realistic human models for numerical simulation of electromagnetic dose in different human tissues and organs [9-11].
Schönborn et al.  analyzed the electromagnetic energy absorption using anatomical head models based on magnetic resonance imaging (MRI) scans of a 7- and a 3-year-old children. No significant differences were found between adults and children in terms of the absorption of electromagnetic radiation in the near field of sources. The same conclusion was obtained when children were approximated as scaled adults.
The use of mobile phones, and especially smart phones, is rapidly increasing, even among preschool children. Electromagnetic absorption testing of mobile phones is typically carried out on a mold in the shape of a large adult head, called the Specific Anthropomorphic Mannequin (SAM). This mold is filled with a liquid that simulates the electrical properties of human tissue. The rationale for choosing that specific head model (i.e., SAM) in standards [13, 14] is based on the following criterion: The peak spatial-average SAR shall be a conservative estimate of the actual value expected to occur in the heads of a significant majority of persons, from adults to young children, during the intended use of wireless handsets.
Children's skulls and scalps are thinner than those of adults. This leads many people to think that the radiation from a mobile phone would penetrate more deeply into the brains of children than of adults. Some publications on the growth and development of the human head were reviewed from a dosimetry perspective in ; the head circumferences of a 1-year-old infant and a 7-year-old child are approximately 84% and 93%–95% that of adults, respectively. This growth mainly takes place in the skull and brain. When considering that children generally do not use mobile phones until they reach school age, the similarity between adults and schoolchildren in terms of internal morphology leads us to expect no significant differences in absorption.
Several studies have measured electromagnetic absorption in realistic child and adult head models that were developed based on medical images and compared this to absorption in the SAM phantom exposed to radiation from a mobile phone antenna [12, 16-20]. Some studies comparing the peak local SAR in the MRI-based head models from adults and children have found either no significant difference or a statistically higher peak SAR in the larger (adult) head [12, 16, 17]. The results of [16, 17] showed that the SAM gives a conservative estimate of the exposure in the anatomical head models that were considered for two standards, IEEE C95. 1-1999 and the International Commission on Non-Ionizing Radiation Protection (ICNIRP) guidelines, if head-only tissue (i.e., exclusion of the pinna during the SAR evaluation) is considered.
Wiart et al.  reported that the peak 1-g SAR in the brain tissues of models of children aged between 8 and 15 years is comparable to that of adult models, while it is about two times higher for models of children aged between 5 and 8 years, due to the smaller thicknesses of the pinna, skin, and skull.
The radiation quality from a mobile phone determines the input antenna power of the phone in a real environment. In most systems, a mobile phone and its corresponding base station check the signal strength between each other and the power level is automatically increased or decreased.
Lee et al.  used modified SAM phantoms with head sizes varying according to age and observed a concerted increase in the peak 1- and 10-g SAR with age for a constant radiated power from a mobile phone. The increase in head volume with age means that a higher input power of a mobile phone is needed for a constant radiation. Nevertheless, the results suggested that the whole-head averaged SAR in children would be higher than in adults for a constant radiation power.
The spatial peak SAR in the head generally shows a strong dependence on the pinna shape and internal morphology of muscle and fat tissue. A difference in internal morphology, such as tissue distribution inside the considered anatomical models, seems to be one of the most important factors influencing spatial peak SAR values .
To date, the reported variability in the peak SAR between adults and children seems to be attributed to individual differences in anatomy rather than age.
Simulation using numerical techniques relies heavily on the dielectric properties of the tissues. In 2001, Peyman et al.  reported the changes in the dielectric properties of various tissues in the rat, from birth up to an age of 70 days, in the frequency range of 130 MHz to 10 GHz. The results showed a general decrease in the dielectric properties with age and provided some insight into possible differences in the exposure assessment for children and adults. However, how and whether these data can be extrapolated to humans is unclear because the development of the head and brain of rats and humans follows a different time pattern.
The effect on SAR of different dielectric properties between adults and children was examined in a few studies [22-24]. The variation in the dielectric properties with age is mainly due to the changes in the water content of tissues. Wang et al. derived the dielectric properties in 7- and 3-year-old child head models as a function of age according to the total body water, but no significant difference was found in peak 1- or 10-g averaged SAR for mobile phone use , because the total body water varies to a great extent in humans under 3 years of age. Peyman et al.  measured the dielectric properties of three age groups of pigs (less than 30 days old, about 100 days old, and fully matured) and applied them to 3- and 7-year-old child head models. They concluded that the effects on an average SAR of 10 g were marginal for walkie-talkie exposure.
Christ et al.  compared the SAR produced by three different phone models in anatomical head models of two adults and four children at different ages, together with agedependent dielectric properties of tissue, based on porcine data reported by Peyman et al. . The used phone models were a generic mobile phone equipped with a monopole antenna, a generic mobile phone with an integrated dual band antenna at the top of the body, and a CAD model of the Motorola Timeport T250 with a helical antenna. Christ et al.  reported that age dependences of dielectric tissue properties did not lead to systematic changes in the spatial peak SAR.
Another difficulty in generalizing age-related differences in electromagnetic absorption, especially the spatial peak SAR between adults and children, arises from the variability in the structure and location of the antenna and the structure of the phone. The public has seen rapid changes in mobile phone technology since the 1st generation (1G) mobile communication system was commercially introduced in the early 1980s; this used analogue technology operating in the 800 MHz range. The 2G systems were launched in most countries in the 1990s and they vary somewhat from each other in the frequency range and access mode. The biggest change between the 2G and 3G systems has been observed in the phone use pattern: the 1G and 2G systems were voiceoriented, whereas the 3G system allows phone users to access data communication applications such as email, internet etc. with higher data rates. Generally, initiating with the 3G phones, the antenna started to be inserted into the phone body.
Fig. 1 shows the changes in outer shapes and antenna types and locations of the most commonly released phone models in South Korea by year. Flip, slider, and bar types make up the largest proportion in the early, mid, and late 2000s, respectively, and the antenna has been primarily located at the bottom of the phone body since the late 2000s. In particular, since the release of smart phones, the most recent Korean models have been bar-type, with the antenna at the bottom of the phone body.
These trends in phone structure might be characteristics limited to the Korean market, but this change needs to be considered in exposure evaluation. However, the research that has investigated possible age-related differences in SAR and that has tested the conservativeness of the SAM phantom has been performed using mobile phone models with the antenna on top of the body, in the form of an external whip, an external helix, or an internal planar inverted F antenna (PIFA) [7, 8, 16, 17, 20, 22, 24].
Long Term Evolution (LTE) is the latest step toward 4G radio technologies designed to increase the capacity and speed of mobile telephone networks: downlink and uplink peak rates reach 100 Mbps and 50 Mbps, respectively, and the advantages expand the functions of a mobile phone. High-speed technology gives mobile phone users instant access to text messaging, email, and other Internet-based information, including entertainment and multimedia content. Another RF radiation which the public almost daily exposed to is coming from the Wireless Local Area Network (WLAN) system, an ad-hoc system operating in the 2.4 GHz or 5 GHz range and set up within limited areas, such as homes, hotels, cafes, or office buildings. It allows the general public to use various devices such as tablet computers, wireless transceiver-equipped notebook computers, and mobile phones (smartphones) with wireless internet access.
Accurate and reliable dosimetry and exposure assessment are key requirements of scientific studies on the health effects of electromagnetic fields. The antenna types and locations on commercial mobile phones have undergone a number of changes from those used in early research on health effects. Therefore, these changes need to be considered in future studies.
Until now, most of the computational simulations have been performed under specific conditions; for example, the standard phone positions against a head model and the maximum output power of a phone model. However, phone usage patterns due to high-speed technology could be quite different among children, adolescents, and adults. Exposure in real environments also needs to be assessed according to different usage patterns between different age groups.