PM2.5 and NO2 assessment in 21 European study centres of ECRHS II: annual means and seasonal differences

The follow-up of cohorts of adults from more than 20 European centres of the former ECRHS I (1989–1992) investigates long-term effects of exposure to ambient air pollution on respiratory health, in particular asthma and change of pulmonary function. Since PM2.5 is not routinely monitored in Europe, we measured PM2.5 concentrations in 21 participating centres to estimate ‘background’ exposure in these cities. Winter (November–February), summer (May–August) and annual mean (all months) values of PM2.5 were determined from measuring periods between June 2000 and November 2001. Sampling was conducted for 7 days per month for a year. Annual and winter mean concentrations of PM2.5 vary substantially being lowest in Iceland and highest in centres in Northern Italy. Annual mean concentrations ranged from 3.7 to 44.9 μg m−3, winter mean concentrations from 4.8 to 69.2 μg m−3, and summer mean concentrations from 3.3 to 23.1 μg m−3. Seasonal variability occurred but did not follow the same pattern across all centres. Therefore, ranking of centres varied from summer to winter. Simultaneously, NO2 concentrations were measured using passive sampling tubes. Annual mean NO2 concentrations range from 4.9 to 72.1 μg m−3 with similar seasonal variations across centres and constant ranking of centres between seasons. The correlation between annual NO2 and PM2.5 concentrations is fair (Spearman correlation coefficient rs=0.75), but when considered as monthly means the correlation is far less consistent and varies substantially between centres. The range of PM2.5 mass concentrations obtained in ECRHS II is larger than in other current cohort studies on long-term effects of air pollution. This substantial variation in PM2.5 exposure will improve statistical power in future multi-level health analyses and to some degree may compensate for the lack of information on within-city variability. Seasonal means may be used to indicate potential differences in the toxicity across the year. Across ECRHS cities annual NO2 might serve as a surrogate for PM2.5, especially for past exposure assessment, when PM2.5 is not available.