Health effects of ambient air pollution: Do different methods for estimating exposure lead to different results?

Sex-specific differences in fetal growth in newborns exposed prenatally to traffic-related air pollution in the PELAGIE mother–child cohort (Brittany, France)

“…Although some of these significant associations might be pure chance findings, this is not likely to explain the consistent and significant associations observed across the four different metrics used to assess fetal growth. Our results are in line with some recent studies also reporting a decreased risk of low birth weight at term or higher birth weights associated with prenatal exposure to NO 2 (Laurent et al, 2013;Kashima et al, 2011;Hannam et al, 2014) or other air pollutants (Vinikoor-Imler et al, 2014;Laurent et al, 2013;Sellier et al, 2014;Madsen et al, 2010). In a recent meta-analysis on this topic, Stieb et al observed a high degree of heterogeneity among studies as measured by the I² value (Stieb et al, 2012).…”

Comparing land use regression and dispersion modelling to assess residential exposure to ambient air pollution for epidemiological studies

“…Gulliver et al (2011) compared LUR and DM at 52 routine monitoring stations in London (UK) using a grouped jack-knife approach Results showed that LUR (R 2 = 0.47) outperformed DM (R 2 = 0.28). Most recently Sellier et al (2014) compared LUR and DM estimates for NO 2 at cohort addresses in Nancy and Poitiers (France) finding a good correlation between the two methods (R = 0.87). …”

“…It remains important, however, to test directly in epidemiological studies differences in effect estimates related to exposure models. A recent study from Sellier et al (2014) which applied four different exposure methods, including LUR and DM, to a cohort in Nancy and Poitiers (France), showed some differences in estimated health effect despite moderate to high correlations between NO 2 exposure estimates at the cohort level…”

“…Chemical Transport Models have also been used to model short-and long-term exposure periods (Hennig et al, 2014). Few studies to date have conducted comparisons between LUR and DMs for their performance in estimating exposures (Beelen et al, 2010;Cyrys et al, 2005;Dijkema et al, 2011;Gulliver et al, 2011;Marshall et al, 2008;Sellier et al, 2014). These studies included different models, spatial resolution, pollutants and study areas, factors likely to have contributed to inconsistent findings within individual studies.…”

Residential NOx exposure in a 35-year cohort study. Changes of exposure, and comparison with back extrapolation for historical exposure assessment

“…LUR model estimates are often comparable with dispersion models for the year the LUR was created. A high correlation (r p ¼ 0.77) between LUR and dispersion modeling was found in two French metropolitan areas (Sellier et al, 2014), while a moderate agreement (r ¼ 0.55) was found in the Netherlands (Beelen et al, 2010).…”