A Critical Evaluation of Exposure Surrogates in RF-EMF Research: Recommendations for Improving Exposure Assessment

Abstract

ISEE-0401 Background: So far, radiofrequency electromagnetic field (RF-EMF) health research has focused on the head's exposure originating mainly from the use of mobile phones. For some health outcomes, however, whole body radiation absorption may be more crucial. Recently developed personal RF-EMF meters (exposimeters) are suitable to estimate RF-EMF exposure of the whole body. However their application is limited in large study collectives and for long term exposure measurements, thus reliable exposure surrogates are needed. The aim of this analysis is to compare different exposure surrogates with personal exposure measurements. Methods: Personal RF-EMF exposure measurements of 12 frequency bands ranging from FM radio (88 MHz) to W-LAN (2.5 GHz) were collected from 166 individuals who carried an exposimeter (EME SPY 120) during one week. In addition, five exposure surrogates were derived: i) distance between place of residency and closest mobile phone base station, ii) self estimated exposure, iii) spot measurement in the bedroom, iv) geospatial modelling of RF-EMF from fixed site transmitters at place of residency, and v) exposure predictions combining geospatial modelling with data about exposure relevant behaviour. Spearman's rank correlations between these surrogates and personal exposure measurements were calculated. Results: In the study collective, mean personal RF-EMF exposure during one week was 0.22 V/m (range: 0.08–0.58 V/m). Compared to personal exposure Spearman's rank correlation for distance to the closest base station was −0.03 (P = 0.75); for self-estimated exposure 0.07 (P = 0.31), for spot measurement 0.44 (P < 0.01); for geospatial modelling 0.28 (P < 0.01), and for predictions combining geospatial modelling with exposure relevant behaviour 0.49 (P < 0.01). Conclusions: Simple exposure proxies such as distance to base station are inappropriate for epidemiological research. More sophisticated approaches considerably improve exposure assessment. Nevertheless, still little data about RF-EMF exposure of the population are available and substantial uncertainty about the relevant predictors for exposure in different countries exists.