Personal exposure to radio frequency electromagnetic fields and implications for health

Exposure to radio frequency electromagnetic fields (RF-EMFs), as produced by mobile phone base stations, broadcast transmitters and cordless phones, has considerably increased over the past 20 years, especially due to the rapid expansion of the mobile phone communication network. Little is known about typical RF-EMF exposure levels and the spatial and temporal variability of RF-EMFs in our environment. Moreover, the contribution of the various exposure sources to total exposure has not been quantified. In general, two types of exposure sources can be distinguished: sources operating close to the body such as personal mobile devices, and environmental far-field sources such as e.g. mobile phone base stations resulting in homogenous whole-body exposure. Only recently have portable exposure meters (exposimeters) become available. These devices are promising for quantifying individual exposure to the most relevant environmental far-field RF-EMF sources during their typical daily life activities, but are not expected to realistically represent exposure from sources operating close to the body because the measurements are heavily influenced by the distance between the emitting device and the exposimeter. In addition, exposimeters are not suitable for use in large-scale epidemiological studies, particularly due to the high costs and the tremendous effort for study participants involved.
Parallel to the increase in RF-EMF exposure, public concern has grown regarding possible adverse health effects of RF-EMFs, in particular concerning non-specific symptoms such as headache. However, to date, only a few epidemiological studies have addressed the possible health effects of environmental RF-EMF exposure. The main reason for that is that assessment of RF-EMF exposure in everyday life is highly challenging. Most epidemiological studies conducted so far were of cross-sectional design, where data on exposure and health are collected at the same point in time. These studies have several drawbacks; in particular they are limited for drawing conclusions about a causal relationship between exposure and health outcomes.
The aim of this thesis is to determine the distribution of individual RF-EMF exposure levels in daily life and to identify the factors relevant for the exposure in order to develop an RF-EMF exposure model. In addition, possible non-specific health effects resulting from everyday RF-EMF exposure are examined.
This thesis was conducted within the framework of the QUALIFEX project (health-related quality of life and radio frequency electromagnetic field exposure: prospective cohort study). QUALIFEX consists of two parts: the exposimeter and main study.
In the exposimeter study, 166 volunteers from the region of Basel carried an exposimeter for one week in order to measure their individual RF-EMF exposure. The participants completed an activity diary and a questionnaire on exposure relevant behaviours. In a validation study, we repeated the exposure measurements of 32 study participants on average 21 weeks after the first measurement. Moreover, spot measurements in the bedroom of the participants and data on exposure levels as perceived by the participants were collected and the geo-coded distance to the closest fixed site transmitter (mobile phone base stations or broadcast transmitter) was computed. The mean residential RF-EMF from fixed site transmitters was computed using a geospatial propagation model. We developed a nonlinear full exposure prediction model by combining the exposimeter measurements, the questionnaire data and the modelled residential RF-EMF.
In the main study, a questionnaire survey investigating potential health effects caused by RF-EMF exposure was conducted in a randomly selected sample of 1375 participants. The questionnaire contained standardised questions on non-specific symptoms (somatic complaints, headache and sleep impairment) and tinnitus. Environmental far-field RF-EMF exposure was assessed using the full exposure prediction model. In order to estimate exposure to close to body sources, objective operator data on mobile phone use as well as self-reported data on mobile and cordless phone use were collected. A follow-up survey was conducted one year after the baseline survey.
In the exposimeter study, the mean RF-EMF exposure to environmental far-field sources for one week was 0.22 V/m. The individual mean values ranged from 0.07 to 0.58 V/m. Mobile phone base stations, mobile phones and cordless phones represent the main contributions to exposure. Radio and television broadcast transmitters, wireless LAN and Tetrapol were shown to be minor exposure sources. Mean values were highest in public transportation vehicles. We identified the following relevant factors for RF-EMF exposure: The modelled RF-EMF at the participants' homes from the geospatial propagation model modified by housing characteristics, ownership of wireless communication devices, and behavioural aspects such as the amount of time spent in public transport. The variance explained (R2) by the full exposure prediction model was 0.52, and the sensitivity and specificity were 0.56 and 0.95, respectively (cut-off: 90th percentile). We were able to show that the full exposure prediction model can also be used to quantify mean exposure for a period of several months as the model reliably predicted the data of the validation study (sensitivity: 0.67; specificity: 0.96). Concerning other exposure assessment methods used in previous studies, we found that the mean individual exposure measured using exposimeters correlated best with the values derived from the full exposure prediction model and the spot measurements. Individuals’ perception of their exposure and geo-coded distance to the closest transmitter turned out to poorly represent personal exposure.
Regarding the health outcomes in the main study, our results do not indicate an impact of RF-EMF exposure in everyday life on somatic complaints, headache, sleep impairment or tinnitus. Neither exposure to environmental far-field sources nor to sources operating close to the body was associated with non-specific symptoms. This finding is in line with a systematic review of the scientific literature on potential health effects of exposure to mobile phone base stations which was conducted in the framework of this thesis. A tendency could be observed in our data that individuals suffered more frequently from non-specific symptoms if they believed to be subject to higher exposure as compared to the general Swiss population.
The mean exposure levels measured in our study were well below the current reference values. We were able to demonstrate the feasibility of modelling individual RF-EMF exposure. This makes it possible to assess exposure without expensive and time-consuming individual measurements. The results of our study allow a better interpretation of previous research and a more efficient planning of future epidemiological studies with large populations. We found that crude exposure assessment methods such as calculating the geo-coded distance to the closest fixed site transmitter are not suitable to represent individual exposure levels.
QUALIFEX is the first study to investigate potential unspecific health effects of RF-EMF exposure in daily life using a cohort design. The results allow us to make more robust conclusions in comparison with cross-sectional analyses used in previous research. Moreover, we used objective measures for both environmental far-field and close to body exposure. We did not find indications for a connection between RF-EMF exposure and non-specific symptoms or tinnitus. However, the mean exposure levels were very low and the changes in exposure were small. Our data do not allow us to draw conclusions about possible consequences of higher exposure levels, e.g. values close to the reference values, or effects due to larger exposure changes which may occur in the future. More data on long-term exposure will have to be collected and analysed in order to satisfactorily answer the question whether long-term RF-EMF exposure can cause adverse health effects. This study has successfully evaluated the methods as well as provided a systematic approach which can be used as a guideline for future research on RF-EMF exposure.