Integration of external exposome data

Source	"Combination of geographical information system (GIS) data on built and natural environments with new data on micro-scale population statistics." Boshuizen, Peuters et al. paper in preparation
Level of evidence	Suggestive evidence
Approach	We examined correlations between 57 exposome indicators across all Equal Life cohorts. We aimed to gain insights for the reduction of complex datasets, as well as to provide context for guiding design and interpreting results.
Findings	Population density underpins many correlations between exposome indicators. This suggests that the health effects of exposures may vary significantly between densely populated urban, and less-populated rural areas. Examples of correlations between population density and other exposome indicators included: ·       Built environment indicators (urbanity, traffic-related air pollution and noise, transport-related infrastructure, and single-person and single-parent households), some of which can be explained by causal relationships (e.g. the presence of major roads implies higher traffic NO2 and traffic noise), and socio-demographic trends (higher percentage of single person families in densely populated areas). ·       Social indicators (old-age poverty, child poverty, foreign citizenship, migration background, general unemployment rate and old age ratio). Migration background was highly interrelated with certain indicators, which could be an area for further research. ·       Local neighbourhood indicators (local streets and crossings, sleep disturbance index, close by urban green and water). ·       Natural and rural environment (NDVI, percentage of tree cover, proportion of youth living in the neighbourhood, distance to local streets and major roads, and ozone concentrations).  ·       Other correlations may not hold for every region (each of the 9 cohorts).
Recommendations	For researchers: Urban/rural divide: This finding holds important implications for future exposome research, indicating that studies may need to stratify analyses by population density or may need to form distinct hypotheses or interpretations depending on whether the study population resides in a rural or urban setting. Social neighbourhood exposures: Social neighbourhood indicators are primarily inter-connected with each other, but also clearly relate to the broader network of physical neighbourhood indicators. This finding highlights the importance of incorporating social neighbourhood exposures in exposome research, as they have often been overlooked or treated merely as confounders in traditional exposure studies.
Target audience	Researchers, Policy makers, urban planners