From the authors:
The two major issues that are addressed in the letter from B. Brunekreef are: 1) the effect of particles with a 50% cut-off aerodynamic diameter of 10 µm (PM10) on lung function growth; and 2) the possible explanation for this observation being regression to the mean because initial lung function was adjusted for in the generalised estimating equations models. Different models for each season were calculated because conceptually we thought that the effects of air pollution might depend heavily on season. Such an effect was observed in our ozone studies where lung function growth tended to recover when ozone levels diminished in the cold season 1. Hence, our results can only be interpreted in respect to season. For PM10 we found a significant effect on lung function growth in the summer, but not in the winter season. Hence, lung function growth from our paper is always “seasonal growth”. However, we agree that the estimates are still rather large and we do not have a good explanation for this observation. We have adjusted for initial lung function in order to observe effects strictly limited to the study period. However, the issue of regression to the mean is indeed important and we have re-run our models without this variable and observed no meaningful changes in parameter estimates or p-values. Hence, regression to the mean did not explain our results.
- © ERS Journals Ltd