Interrelations between Environmental Exposures: Air Pollution, Noise and RF-EMF – Potential for Confounding?

Abstract

Background If environmental radiofrequency electromagnetic field (RF-EMF) exposures are correlated with other environmental exposures, spurious results could be obtained if studied outcomes are associated with the respective exposures, but not with RF-EMF. The same accounts for socioeconomic position (SEP) which could provide the potential for confounding. Methods We used the Amsterdam Born Children and Development (ABCD) study to asses correlation between different environmental exposures, at the home address when children were 7yrs of age. Correlation between RF-EMF exposure from base stations was assessed with a geospatial model (NISMap) and from other indoor sources (DECT/WiFi) using parental self-reports. Traffic related air pollution (NO2) exposure was derived from land-use regression. Road/tram traffic noise and aircraft noise were extracted from corresponding noise maps. SEP indicators included maternal educational level, parental occupational attainment, and parental financial situation. Results We included 3,249 children in the analysis. Highest correlation between exposures was observed for RF-EMF from base stations with road/tram noise (Spearman correlation coefficient of 0.18, p<0.0001). Correlation of RF-EMF with air pollution and aircraft noise was low (all r <0.05, all p>0.7). There was a tendency of higher base-station exposure in children of mothers with lower educational level and lower occupational attainment (both r = 0.05, both p<0.01). For indoor RF-EMF sources, the associations were rather the opposite, with higher exposures in children of families in a better financial situation and of parents with higher educational level or occupational attainment (all r > -0.09, p<0.01). Conclusions All assessed correlations were low and potential for confounding therefore is expected to be weak. However, for outcomes where SEP and noise are important predictors, it is advisable to include these factors in future studies to assess the potential for confounding.