[Open questions in the interpretation of daily time-series studies on air pollution and health. Part I: estimate of exposure]. / Problemi interpretativi delle analisi di serie temporali su inquinamento atmosferico e salute. Parte I: le stime di esposizione.

This paper is a commentary to the MISA study (Meta-analysis of the Italian studies on short-term effects of air pollution). The interpretation of the relationships observed in the time-series analyses is not straightforward. I will review a number of open questions: the exposure estimates; the meaning of the daily mortality and hospitalisation rates, in terms of impact on life expectancy, incidence of specific diseases, exacerbation of pre-existing chronic diseases; the pattern of the estimated dose-response curves; the role of possible effect modifiers; the hypotheses about the underlying mechanisms. These interpretative problems often present overlapping aspects. Some of the pending answers are relevant for policy making. For editorial reasons this review is divided into two parts. This first paper focuses on the exposure estimates. The second part, to be published in the next issue of this journal, will deal with the meaning of the outcome variable, the shape of the dose-response curves, possible effect modifiers and underlying mechanisms.