Discussion
After a continuous rise in the prevalence of wheeze/asthma since 1978 and a phase of stabilization from 2003 to 2008, the sixth (2013) and seventh (2018) follow-up surveys of Patras epidemiological study showed a continuing decline in wheeze/asthma rates among 8- and 9-year-old schoolchildren in Greece.
Interestingly, persistent and non-current wheeze/asthma followed this overall trend, while the prevalence of recent-onset disease remained essentially unchanged. In our surveys, both the persistent and the non-current group included children in whom wheeze/asthma was diagnosed more than two years prior to the survey, whereas the recent-onset group consisted of children with symptoms that were manifested more recently (i.e. solely within the last two years). It seems, therefore, that the prevalence trends of childhood wheeze/asthma in Patras were primarily determined by changes in the dissemination of wheeze/asthma diagnosis at younger ages (i.e. to children younger than 6-7 years).
It is worth noting that the diagnosis of asthma was also particularly applied to persistent and non-current wheezers (Figure 3), i.e., to children in whom the onset of wheeze occurred at preschool age. Conversely, asthma was reported at significantly lower rates in children with recent-onset wheeze. These findings suggest that Greek paediatricians tend to assign the diagnosis of asthma to preschoolers with troublesome respiratory symptoms, thus resulting in transfer of asthma diagnosis towards younger ages. Conversely, asthma may be underdiagnosed and therefore undertreated in older children with recent-onset symptoms. Our spirometric findings of lower FEV1 and FEV1/FVC in recent-onset wheezers/asthmatics (see below) reinforce the above assumption.
Many studies from Europe, North America and other parts of the world have reported an increase in the prevalence of asthma through the late 1990s1-7,13,18. However, studies of the last 20 years have shown that the ‘asthma epidemic’ has reached a peak and may even be declining in high prevalence societies8-18. A series of repeated surveys from 1964 to 2014 in the city of Aberdeen, Scotland, indicate a continuing decline in asthma prevalence after 2004 among the 5-12-years-old schoolchildren14,15; after its peak in 2004 (29.5%), the prevalence of lifetime asthma (i.e. ever-had asthma) declined to 22.7% in 2009 and to 18.6% in 2014, i.e. lower than the 1994 levels (19.5%)14,15. Data from the National Health Interview Survey also indicate a decrease in the prevalence of current childhood asthma in the USA, from 9.6% in 2009 to 8.4% in 201718,23.
The Patras epidemiological study is one of the longest worldwide (40 years; second only to the Aberdeen study) that has captured the prevalence of childhood asthma at several time points by using identical methodology. The two most recent surveys of 2013 and 2018, showed -for the first time- a decline in wheeze/asthma prevalence: from 6.9% in 2008 to 5.2% in 2013 and 4.3% in 2018 for current wheeze/asthma, and from 12.6% in 2008 to 9.6% in 2013 and 6.6% in 2018 for lifetime wheeze/asthma. These rates are two- to almost four-fold lower than those reported in the Aberdeen and other studies14,15,18,24, which may be attributed to the more ‘restrictive’ phrasing of our questionnaire (i.e. physician-diagnosed wheeze/asthma) and the prompting of parents to respond with a negative answer if in doubt. Nevertheless, the relative prevalence changes between Patras surveys (Figure 2) are comparable to those of the Aberdeen19.
In contrast to the ‘asthma epidemic’ of the 1980s and 1990s, the reversed trends in the prevalence of childhood asthma are difficult to explain. Although the decline has been generally attributed to lifestyle changes or/and to improved diagnostic and management practices12,15,25, such causal relationships have not been proven to date. Our study was not designed to evaluate the influence of various risk factors on asthma prevalence. However, it is highly unlikely that lifestyle and environmental parameters (allergen exposure, viral epidemics, nutrition, overweight/obesity, physical activity, atmospheric pollution, etc)26 have shifted during the last decade in a manner that would favor a substantial decline in asthma prevalence. The increase in disease awareness in association with the broad dissemination of pediatric asthma guidelines and the establishment of controller therapy, may have led to the significant reduction of asthma admissions since the early 1990s in Athens, Greece,27 i.e. in an urban environment similar to ours. However, the decrease in admission rates occurred almost two decades before the decline in asthma prevalence that we report herein and was noted in a period during which childhood asthma rates were in fact still increasing.
Based on our analysis, we maintain that the course of childhood asthma prevalence over the past decades largely reflects the changing perceptions regarding the disease among the physicians and/or parents28, especially in the case of younger children with troublesome respiratory symptoms. The diagnostic transfer at younger ages most likely reflects the ‘asthma fashion’ of the last two decades of the 20th century29, when broad dissemination of asthma guidelines among physicians occurred30. To the best of our knowledge, the present study is the only one to evaluate physician-diagnosed wheeze and asthma, both separately and in relation to their onset and duration (i.e., recent-onset, non-current, and persistent) over a period of three decades. Furthermore, in the 2018 study, all children with at least one positive answer to wheeze/asthma questions (recent-onset, persistent, and non-current wheezers/asthmatics) and a random sample of those with negative responses (healthy controls) were invited to perform spirometry. Participants with recent-onset wheeze/asthma had significantly lower FEV1 and FEV1/FVC (i.e. an obstructive spirometric pattern) as compared to all other groups; their FEF25-75 values were also lower than those of healthy controls. These results, together with the fact that the diagnosis of asthma was less frequently assigned to recent-onset wheezers, suggest that asthma which occurs at late preschool age may have been underdiagnosed and thus remained under suboptimal treatment in our population. On the other hand, children with persistent wheeze/asthma also had lower FEV1/FVC compared to healthy controls, although the FEV1 did not differ between the two groups. However, the distribution of FEV1 values in the persistent wheeze/asthma group was wide, thus including many cases with low FEV1 (Figure 4). Indeed, when only children with relatively low FEV1(i.e. less than -1 z-score) were considered, we found that both persistent and recent-onset wheezers/asthmatics were overrepresented. The latter finding is important, because it suggests that a considerable percentage of preschool wheezers in whom the symptoms persist at school age may exhibit decreased lung function. The cross-sectional design of our study does not permit to draw further conclusions on the trajectories of lung function in relation to the natural history of wheeze/asthma in this group.
The limitations of studies based on written asthma questionnaires have been addressed repeatedly; still, such studies remain the single most useful tool to obtain information from large numbers of participants13,19,24,31,32, It has also been established that the understanding of the term ‘wheeze’ differs greatly between parents and physicians33,34. In our surveys, however, parents were asked to report not on the symptom as perceived by themselves but on wheeze diagnosed by a physician. The reporting of physician-diagnosed asthma was almost always accompanied by reporting of physician-diagnosed wheeze; equivocal answers were resolved by telephone contact and, if still doubtful, they were excluded from analysis. Cough was not included in our questionnaire; it is our conviction that in order to maintain the results between surveys comparable the phrasing of the previous set of questions (1991, 1998, 2003, 2008) should not be altered. Moreover, the inclusion of persistent or recurrent cough without wheeze in the diagnosis of childhood asthma has been challenged35,36.
In conclusion, the results of seven identical surveys over a 40-year period show that the prevalence of childhood wheeze/asthma in Greece declines continuously since its peak during the 2003-2008 period. This reversing trend is most likely the result of changing asthma perceptions among physicians and/or parents, especially in the case of preschool children with troublesome respiratory symptoms. Concurrently, however, true asthma may be underdiagnosed and thus remain under suboptimal treatment. The lower lung function of children with recent-onset wheeze/asthma and of a considerable fraction of those with persistent disease, further strengthens the above hypothesis.