Asthmatics have been shown to be greatly affected by environmental exposure. Both experimental [1] and epidemiological studies [2] have shown air pollution to aggravate asthma problems. Concern about the possible adverse respiratory effects of major ambient air pollutants has stimulated investigators to attempt to relate effects to lower and lower concentrations in the search for minimum concentrations at which adverse effects on health are detectable. As a result of such studies, even in relatively weakly polluted areas, the guidelines and limits for several pollutants are now under revision and acceptable limits from a health point of view are expected to be set at increasingly lower levels.Daily diaries of respiratory health are very useful in studies of the effects of air pollution on susceptible persons, particularly in smaller cities and towns where studies of hospitalization or mortality would provide very small numbers. Diary studies, which are often referred to as panel studies, have become common in the study of air pollution and respiratory effects [2][3][4]. Diary information includes health-related endpoints such as exacerbations of asthma, other symptoms, peak flow and airway medication in different combinations [3][4][5][6][7][8][9][10]. Environmental data must include not only air pollution levels but also factors that could operate as confounders. One advantage of the diary studies is that potential confounders are limited to those that can co-vary with ambient pollution concentrations over time, e.g. humidity, temperature and pollen concentrations. The typical confounding factors from crosssectional studies, such as socioeconomic status and smoking habits, do not pose any problems in these studies.Analysis of frequencies of symptoms and medication use can be based on time series of diary data from individuals in a panel or averaged daily prevalence or incidence rates for the panel. Regardless of the method used, it is necessary to address the problem of state dependence in the health variables to avoid underestimation of the standard error for the parameter estimates [5]. If a person has a symptom one day, they will probably present with the same symptom the next day, which accounts for the relatively frequent occurrence of autocorrelation of daily prevalences. A problem with averaged symptom data is the occurrence of missing responses or exclusions due to time spent outside the study area, which leads to daily prevalences based on a varying panel composition, without any relevant adjustments.This report presents data from a study of acute respiratory problems among people with asthma or asthma-like problems and their relation to variations in low air pollution concentrations in the southern Swedish city of Landskrona, which has a relatively mild winter climate.
Methods
Study populationLandskrona is a city on the coast in southern Sweden with approximately 30,000 residents. It is situated close to several larger cities. The air in Landskrona is affected by emissions from traffic on a major high...