Exposure chain of urban air PM2.5—associations between ambient fixed site, residential outdoor, indoor, workplace and personal exposures in four European cities in the EXPOLIS-study

Abstract

In the EXPOLIS study personal exposures and microenvironment levels of air pollutants from 50–201 urban adult (25–55 yr) participants were measured in six European cities during 1 yr from autumn 1996 to winter 1997–98. This paper presents the associations between the personal PM2.5 exposures, microenvironment (residential indoor, residential outdoor and workplace indoor) and ambient fixed site concentrations measured in Helsinki (Finland), Basel (Switzerland), Prague (Czech Republic) and Athens (Greece). Considering the whole chain from ambient fixed site to residential outdoor, residential indoor and personal leisure time (non-working hours) exposure, the correlations were highest between personal leisure time exposures and residential indoor concentrations (non-environmental tobacco smoke (ETS): Pearson r=0.72−0.92, ETS included: r=0.82−0.86) except in Athens, where the correlation between residential indoor and outdoor air was highest (non-ETS: r=0.82, ETS included: r=0.68)). Unfortunately, ambient fixed site PM2.5 concentrations were measured continuously only in Helsinki. Ambient fixed site PM2.5 concentrations correlated quite well with residential outdoor concentrations (r=0.90), and also with residential indoor (non-ETS) concentrations (r=0.80), but concentrations measured at ambient fixed site monitors were poor predictors of personal exposures to PM2.5. They were particularly poor predictors of personal workday exposures (non-ETS: r=0.34, ETS included: r=0.25), but considerably better for personal leisure time exposures (non-ETS: r=0.69, ETS included: r=0.54). According to log-linear regression models combined from all centres of non-ETS-exposed participants, residential indoor concentrations explained 76% of personal leisure time PM2.5 exposure variation and workplace indoor concentrations explained 66% of the workday exposure variation.