Without ever fully understanding the reasons for the global increase in the prevalence of asthma during the later part of the 20th century, we are now faced with the challenge of explaining an apparent decrease in prevalence. A challenge indeed! During the 1980s and early 1990s several repeat cross sectional studies, conducted in widely varying regions of the world, reported an increase in the prevalence of symptoms of asthma, diagnosed asthma, and medication use for asthma.^1,2 Although most studies relied on self-reported symptoms and diagnoses—and hence could be subject to reporting bias—at least one³ observed substantial changes in airway hyperresponsiveness, lending credence to the view that this was not simply related to changes in disease classification and labelling. Parallel increases in hospitalisation rates⁴ and mortality rates⁵ also were consistent with a true rising tide of asthma. This trend led to major national and international initiatives to control the disease.^6–9

There are now several reports to suggest that the tide has turned. During 2004 there have been a number of repeat cross sectional studies from Hong Kong,¹⁰ Korea,¹¹ Switzerland,¹² the UK,¹³ Scotland,¹⁴ Belmont, Australia¹⁵ and Melbourne, Australia¹⁶ that have reported a levelling off in asthma prevalence. In some studies there has been a statistically significant decrease in the prevalence of asthma symptoms. Although numerous environmental factors—including indoor and outdoor air quality, infections, diet, sibship, breastfeeding, and pet ownership—have been linked to the aetiology of asthma in cross sectional studies, none of these provides a robust explanation for the previous increasing trend or the apparent reversal of that trend. This is particularly the case as the trends have been observed in a wide range of countries, including those with both a high and low baseline prevalence of asthma.

In this issue of Thorax Mommers et al report the results of a series of four repeat cross sectional studies in Dutch school children.¹⁷ The studies conducted in 1989, 1993, 1997, and 2001 measured respiratory symptoms and treatment in 8–9 year olds. The strengths of this series of consecutive studies are that all the studies included well over 1000 children and had participation rates of over 95%. Identical sampling methods and questionnaires were used, giving us confidence that the results are representative of Dutch school children in this region and that the differences are not due to changes in measurement instruments.

Between 1989 and 2001 there was a downward trend in the prevalence of recent wheeze which was statistically significant for boys but not for girls. The largest reduction was over the most recent period (1997–2001). Over this 12 year period there was no statistically significant trend in the prevalence of shortness of breath, coughing with phlegm, or chronic cough, with the prevalence remaining fairly stable in both boys and girls. However, there was a significant trend towards an increased use of asthma medication by boys who reported recent wheeze over this period. Girls who reported recent wheeze also used more medication but this trend was not significant. So, in Dutch school children between 1989 and 2001 it appears as though the prevalence of asthma symptoms has remained stable or, in the case of recent wheeze, the prevalence has decreased while the use of asthma medication in those children who have symptoms has increased. This apparent paradox may contain a clue to the explanation for the recent trends.

The authors suggest that increasing use of treatment may be responsible for this trend by suppressing asthma symptoms in a larger proportion of the population. An alternative explanation is that the diagnosis of asthma is being used more sparingly in 2001 than it was in previous years. In other words, children with episodic or intermittent wheeze, who do not require asthma medication, may have been less likely to be labelled as having asthma symptoms in 2001. This would explain the observed increase in the proportion of symptomatic boys who used asthma medication during this period. The finding of no change in the prevalence of airway hyperresponsiveness between 1992 and 2002 in an Australian series¹⁵ tends to support the view that there has been no real change in disease prevalence over that period and, rather, the observed decreases in reports of asthma reflect more specific application of the diagnostic label in recent times.

Repeat cross sectional studies provide a useful insight into changes over time. Comparison of studies from different countries and regions are informative, but these comparisons have been limited by the lack of standardised methods between studies. The most accurate insights into international time trends will be gleaned from repeat cross sectional studies conducted as part of the major multicentre studies of the International Study of Asthma and Allergies in Childhood (ISAAC) and the European Community Respiratory Health Survey (ECRHS). Three of the studies from 2004 were repeat cross sectional studies from ISAAC.^10,13,16 When we have more studies reporting international time trends, we will be in a better position to explain the ebb and flow of asthma.