Viel JF, Clerc S, Barrera C,Rymzhanova R, Moissonnier M, Hours M, Cardis E. Residential exposure to radiofrequency fields from mobile-phone base stations, and broadcast transmitters: a population-based survey with personal meter Occup Environ Med. March 30, 2009 Ahead of print.
People are exposed to RF electromagnetic fields from many sources including radio and TV transmitters, telecommunications links and satellite communication, mobile phones and their supporting transmitters. Both the public perceptions and most epidemiological studies rely on the assumption that the distance of a residence from the base station mast is an appropriate surrogate for exposure to RF fields. Although power density in the RF beam decreases with increasing distance, the relation between exposure and distance from the antenna is usually very complex, especially in urban areas. Many of the RF measurements reported so far were spot measurements that are unlikely to be a good indicator of exposure over time in view of time-dependent signal variability and of the movements in space that characterize most of human activities. Recently developed personal exposure meters (PEM) offer a way forward for direct exposure assessment of individuals over time and space.
The goals of this study, using such a personal exposure meter, were to characterize the distribution of residential exposure from antennas and transmitters, to assess how exposure to RF fields varies with distance from these point sources, and to test the association of RF exposure with the level of urbanization of the area.
The study was conducted in the Department of Doubs (485,000 inhabitants) in eastern France. Stratified sampling was carried out by dividing the study population in urban, periurban and rural subgroups according to the place of residence. A total of 200 people was enrolled into the study (urban: 50, periurban: 100, rural: 50). Data on individual, house and workplace characteristics and on activities that might influence RF exposure were collected at the beginning of the study using questionnaires. All participants were supplied with a personal exposure meter for 24 hour measurements and kept a time-location-activity diary in which they noted their location and activities every 15 minutes for 24 hours. The PEM records the electric field strength present in 12 different frequency bands with 0.05 V/m lower detection threshold and a 5 V/m upper recording threshold. A database giving details for all antennas located in the Department of Doubs was provided by the French Radiofrequency Agency. The antenna list provided by the Agency was cross-checked with the data on the Cartoradio Web site. Residential address was geocoded, and distance from each antenna was calculated based on the coordinates of the residence location and the antenna location. Two exposure metrics were defined for each participant and each radiofrequency: the proportion of measurements above the detection limit and the maximum value of electric field strength. Statistical analyses were performed in terms of field strength, which is mathematically related to power density. Data were fitted with cubic smoothing P-splines with 10 degrees of freedom using a logarithm scale for distance. The focus was on the shape of the curve over 0 – 1,000 m range for GSM, DCS and UMTS, and the 0-30,000 m range for FM and TV. Non-parametric Kruskal-Wallis tests were performed to compare exposures across urban, periurban and rural areas.
Quality controls indicated personal exposure meter failure for 15 participants. One more subject was excluded because of no stay at home during the 24 hour period. Thus, 184 subjects (45, 92 and 47 in urban, periurban and rural areas, respectively) were left for analyses. Much of the time the recorded field strength was below the detection level (0.05 V/m) except FM band with a (still low) mean proportion above the detection threshold of 12.3%. The maximum electric field strengths, all below 1.5 V/m, were between 20 and 400 times lower than the International Commission on Non-Ionizing Radiation Protection (ICNIRP) reference level. Exposures varied greatly, even at similar distances from GSM and DCS base stations. The greatest differences were observed in the short range, i.e. in urban and periurban areas. Exposure to GSM and DCS waves peaked around 280 m and 1000 m from the antennas. A downward trend was found within a 10 km range for FM. Conversely, TV and UMTS signals did not vary with distance. RF exposures were smaller in rural areas serviced by lower base-station densities.
Interpretation and Conclusion
Despite numerous limiting factors entailing a high variability in RF exposure assessment, but owing to a sound statistical technique, the authors have found what theory implies. Exposures from GSM and DCS base stations increase with distance in the near source zone to a maximum where the main beam intersects the ground. The authors believe that their results will contribute to the information needed by both local authorities and members of the public about the siting of base stations and their associated emissions.