E-POSTER GALLERY (ID 409)

P-0647 - Application of time-activity estimates in a representative population to a cohort study: the ELSA cohort

Abstract Control Number
2605
Abstract Body
Background/Aim
Application of modelled air pollution exposure estimates to cohort study respondents most often rely on the assumption that the individual remains outside (at their residential address) and static throughout the exposure period. This method likely leads to overestimation of exposure to outdoor air pollution. A lack of information on time-activity patterns and movement of a population throughout a study area therefore has the potential to limit heterogeneity in exposure estimates applied to cohort respondents in epidemiological study. In the present study, the London Travel Demand Survey (LTDS) and London Hybrid Exposure Model (LHEM) incorporate time-activity information of London residents alongside air pollution model estimates applied to respondents of the English Longitudinal Study of Ageing (ELSA).
Methods
Annual average CMAQ-urban (Community Multiscale Air Quality) dispersion model estimates were applied to the residential address of ELSA respondents. LTDS data for ~75,000 London residents were incorporated into the LHEM in order to calculate average time-activity patterns stratified by age and location. Estimates of exposure to NO2, PM10 and PM2.5 for ELSA respondents were then calculated as a reflection of both CMAQ-urban and LTDS/LHEM adjustments.
Results
Time-activity patterns were shown to be significantly different between age groups, as well as by location across London. At recruitment to the ELSA study, London-based respondents (n = 1,037) showed a 62.9% reduction in modelled NO2 exposure when representative time-activity estimates were incorporated. A reduction of 67.7% and 38.3% were observed for PM10 and PM2.5, respectively.
Conclusions
A framework through which estimates of time-activity patterns in representative populations are calculated will likely improve the accuracy of modelled outdoor air pollution estimates when both are applied to respondents of cohorts in epidemiological study. The framework described here will be utilised in a study assessing the true association of air pollution on cognition in the elderly respondents of ELSA.