Development of historic monthly land use regression models of SO2, NOx and suspended particulate matter for birth cohort ELSPAC

Vulnerable windows in child development utero and after birth are critical time points for uncovering the links between environment health. Particular attention is paid to first 1000 days of life from conception second year life. The ELSPAC (European Longitudinal Study Pregnancy Childhood) cohort, launched early 1990s, a rich source longitudinal data about health events, based mainly Brno, Czechia. There currently no air quality concentration maps that can be used assess exposure pollutants this period 1990s Central Europe. Simply transferring current models 1990's burdened with error introduced by temporal change emission sources land use area. Therefore, Czech monitoring were develop monthly regression (LUR) models, which combine collected spatial variables predict variation exposures pollutants. Monthly pollutant concentrations regressed against GIS-based potential predictor LUR following supervised forward linear regression, several predefined constraints. We constructed 180 sulphur dioxide (SO2), nitrogen oxides (NOx) suspended particulate matter (SPM) 1990–1994, completely cover all study participants. final showed, on average reasonably good performance (adjusted R2 = 0.59 hold-out validation (HOV) 0.40 SO2; adjusted 0.75 HOV 0.35 NOx; 0.61 0.31 SPM; mean number stations 74, 38 41, respectively). For these roads greenness predominantly selected as best predictors. modelled will serve many subsequent epidemiological studies, but our may also other possibly European cities period.