Impact of residential mobility on exposure assessment in longitudinal air pollution studies: A sensitivity analysis within the ESCAPE project
2012 (English)In: Scientific World Journal, ISSN 1537-744X, E-ISSN 1537-744X, Vol. 2012, 125818- p.Article in journal (Refereed) Published
Exposure misclassification in longitudinal studies of air pollution exposure and health effects can occur due to residential mobility in a study population over follow-up. The aim of this study was to investigate to what extent residential mobility during follow-up can be expected to cause exposure misclassification in such studies, where exposure at the baseline address is used as the main exposure assessment. The addresses for each participant in a large population-based study (N>25,000) were obtained via national registers. We used a Land Use Regression model to estimate the NOx-concentration for each participant’s all addresses during the entire follow up period (in average 14.6 years) and calculated an average concentration during follow up. The Land Use Regression model explained 83% of the variation in measured levels. In summary, the NOx concentration at the inclusion address was similar to the average concentration over follow-up with a correlation coefficient of 0.80, indicating that air pollution concentration at study inclusion address could be used as indicator of average air pollution concentrations over follow-up. The differences between an individual´s inclusion and average follow-up mean concentration was small, and seemed to be non-differential with respects to a large range of factors and disease statuses, implying that bias due to residential mobility was small.
Place, publisher, year, edition, pages
2012. Vol. 2012, 125818- p.
Environmental Health and Occupational Health Human Geography
IdentifiersURN: urn:nbn:se:umu:diva-60677DOI: 10.1100/2012/125818OAI: oai:DiVA.org:umu-60677DiVA: diva2:561975