Objective: To examine the relationship between air pollution and coronary events in Rome in the period 1998-2000, considering both out-of-hospital deaths and hospitalisations.
Design: Time-series of daily counts of out-of-hospital deaths and hospitalised events, implementation of Generalised Additive Models.
Setting: The air pollutants taken into account were PNC (Particle Number Concentration--a measure of ultrafine particles), PM10, CO, NO2, SO2 and O3. The association was studied with respect to either single days or the cumulative effect on more consecutive days; furthermore, effect modification by age was tested (for the age groups 0-64, 65-74 and 75+).
Participants: People resident of Rome and died/hospitalised for coronary causes into the city in the period 1998-2000.
Main outcome measure: Association between pollutants and out-of-hospital deaths/hospitalised events. Distinction between fatal events (out-of-hospital deaths + hospitalisations with death within 28 days of admission) and non fatal events (hospitalisations with survival longer than 28 days).
Results: Significant association between PNC, PM10, and CO with out-of-hospital deaths, smaller effect on hospitalised events. For a variation of PNC of 28000 particles per cm3 (interquartile range of the distribution) the increase in the risk of out-of-hospital coronary death at lag 0 was 8.1%; for hospital admissions, the risk increased by 4%. The association was stronger in subjects older than 65, and was more evident for total fatal events than for non-fatal hospitalisations.
Conclusion: The study showed increased risks of coronary events associated with air pollution in Rome: ultrafine particles, directly generated by vehicular traffic, are the environmental indicator which best characterizes the health risk.