Is asthma in children still increasing? 20‐year prevalence trends in northern Sweden

In the present study, we describe prevalence trends of asthma and investigate the association with asthma symptoms, use of asthma medication, and asthma severity among 8‐year‐old children in Norrbotten, Sweden in 1996, 2006, and 2017.


| INTRODUC TI ON
Asthma is one of the most common chronic non-communicable diseases among children and a global health concern.During the second half of the 20th century, an increase in asthma was documented in many areas of the world. 1 The International Study of Asthma and Allergies in Childhood (ISAAC) has studied time trends of asthma symptoms in children since the 1990s.ISAAC found that after decades of increasing rates, the prevalence seemed to stabilize in the early 2000s, and even decreased in some high-income countries. 2[5][6] The Global Asthma Network (GAN), a development from the ISAAC study, reported that the overall prevalence of asthma symptoms in Europe remained at a similar level from the early 2000s until 2020. 7 have previously published data on asthma prevalence trends in Swedish children from 1996 to 2006 and found that the prevalence of asthma-related symptoms remained on a similar level, while physician diagnosis of asthma and use of asthma medication increased. 8A similar result was found in another Swedish study of time trends of asthma in 15-year-old between the years 2000 and 2008. 9Explanations for these diverging results on diagnosis and symptoms could be changes in guidelines for diagnostic methods and practice, more effective treatment regimes, or increased knowledge and awareness in society, resulting in more and earlier asthma diagnosis among children presenting with symptoms. 8Trends after 2010 are unknown and it is unclear to what extent the increase in asthma diagnosis corresponds to an increase in asthma morbidity.
In the present study, we describe prevalence trends of asthma and investigate the association with asthma symptoms, use of asthma medication, and asthma severity among 8-year-old children in 1996, 2006, and 2017.

| Study sample
The Obstructive Lung Disease in Northern Sweden (OLIN) studies have recruited three pediatric cohorts for studies about asthma, rhinitis, and eczema.[12] The response rates were high in all three surveys, cohort I (1996) n = 3430 (97%), cohort II (2006) n = 2585 (96%), and cohort III (2017) n = 2785 (91%).The studies have been approved by the Swedish Ethical Review Authority.The parents provided informed consent for their child to participate.

| Questionnaire
The questionnaires were distributed through the schools, completed by the parents at home, and returned to school by the child.The questionnaire included the core questions about asthma, rhinitis, and eczema from the ISAAC questionnaire 13 and additional questions about physician diagnosis, treatment, and severity of asthma, as well as potential risk factors. 14The questions were identically worded in all surveys.

| Definitions
The prevalence of asthma symptoms is presented as point prevalence in the last 12 months, reported at age 8 years.The wording of the questions about asthma symptoms is presented in Table S1 in the online supplement.Physician-diagnosed asthma was defined as an affirmative answer to the question 'Has your child been diagnosed by a physician as having asthma?' Current asthma was defined as physician-diagnosed asthma and either wheezing or use of asthma medication in the last 12 months.Troublesome asthma was defined as either speech-limiting wheeze or sleep-disturbing wheeze in the last 12 months.Asthma severity score was based on an arbitrary score ranging from 0 to 5 and included wheeze in the last 12 months, daily use of asthma medication, ≥1 night per week with disturbed sleep, at least one episode of speech-limiting wheeze, and >12 episodes of wheezing.Each item should have occurred during the last 12 months and yielded one point each. 8SABA: short-acting beta-2 agonist.ICS: inhaled corticosteroid.

| Statistical analyses
The SPSS software, version 27.0 (IBM Corp, New York, USA) was used for analyses.A p-value of <.05 was considered statistically significant.The chi-square test and test for trend were used for comparisons of proportions and the t-test or ANOVA for comparison of mean values between groups.Proportions with 95% confidence intervals (CI) are presented in the online supplement.Internally missing values (<3%) for single questions about symptoms, diagnosis, and medication were regarded as negative responses.

| Prevalence trends of asthma and asthma symptoms from 1996 to 2006 and further to 2017
Overall, the prevalence of current asthma symptoms in the last 12 months remained on a similar level and there were no statistically significant differences between 1996, 2006, and 2017, while

Key message
The prevalence of asthma among children increased during the second half of the 20th century but time trends of asthma prevalence after 2010 are unknown.This population-based study reports that the prevalence of wheezing was stable among 8-year-old children in Sweden between 1996 and 2017 while the prevalence of physiciandiagnosed asthma almost doubled but without an increase in asthma morbidity.the prevalence of "ever" wheeze increased significantly.Physiciandiagnosed asthma increased from 5.7% in 1996, to 7.4% in 2006, and 12.2% in 2017, p < .001.Also, ever asthma, current asthma, and use of asthma medication in the last 12 months increased significantly (Table 1).The prevalence trends were similar in boys and girls but the estimates were significantly higher in boys than girls (Figure 1).Among those with wheeze in the last 12 months, the prevalence of physician-diagnosed asthma increased from 42.3% in 1996, 44.3% in 2006, to 55.6% in 2017, p = .002and test for trend p < .001,as did 'use of any asthma medication last 12 months', 55.1%; 61.8% and 71.2%, p < .001and test for trend p < .001(data not in table).

| Prevalence trends of asthma symptoms and impact on daily life among children with asthma
Among those with physician-diagnosed asthma, the prevalence of asthma symptoms and use of any asthma medication in the last 12 months decreased significantly from 1996 until 2017, as did the impact of asthma on daily activities and absence from school due to asthma (Table 2).This pattern was similar among those with current asthma, but the improvement in terms of less impact on daily activities was less pronounced and not statistically significant (Table 3).
Among those with physician-diagnosed asthma, the proportion that neither reported wheezing nor use of asthma medication in the last 12 months, i.e. not having current asthma, increased from 8.6% in 1996 to 18.8% in 2006 and 30.0% in 2017 (p < .001and test for trend p < .001)(data not in table ).
The distribution of the asthma severity score in the three surveys shows an increase in the proportion of children with a lower score over time (Figure 2).The mean asthma severity score was lower in 2017 (0.7), compared to 2006 (1.1) and 1996 (1.4), with p < .001for both comparisons.
Among children without physician-diagnosed asthma, there were no significant differences in the prevalence of respiratory symptoms between the surveys.For instance, wheezing in the last 12 months was reported by 6.2% in 1996, 6.5% in 2006, and 5.2% in 2017, p = .149(Table 2).

| Trends in use of asthma medication in children with and without asthma
Among those with physician-diagnosed asthma, use of short-acting beta-2 agonist (SABA) decreased between the three surveys; 85.3% in 1996, 77.0% in 2006, and 67.1% in 2017, p < .001(Table 2).In contrast, there was a significant difference in the use of inhaled corticosteroid (ICS) between the years: 54.8% in 1996, 67.0% in 2006, and 58.8% in 2017 (p = .043)among those with physician-diagnosed asthma, but no significant trend (p = .607).Among children without physician-diagnosed asthma, use of any asthma medication in the last 12 months was low but increased from 2.1% in 1996 to 3.0% in 2006, and 3.4% in 2017 (p = .003),and ICS use increased from 0.4% to 1.3% and 1.6% (p < .001).

TA B L E 1
The prevalence of asthma-related symptoms and asthma among 8-year-old children in 1996, 2006, and 2017.(p < .001),as did the number of wheezing episodes, sleep-disturbing wheezing, troublesome asthma, and interference with daily life.
However, there was no significant difference in the prevalence of speech-limiting wheezing.

| DISCUSS ION
In this population-based study of prevalence trends of asthma in 8-year-old children, we found that asthma-related symptoms remained on a similar level while the use of any asthma medication increased and the prevalence of physician-diagnosed asthma doubled between 1996 and 2017.However, among children diagnosed with asthma, the prevalence of asthma symptoms, the impact on daily life, and asthma severity decreased, while the use ICS increased.
6][17] However, few updates on prevalence trends of asthma have been published during the last years so whether this positive trend remains into the 2020s has been unclear.We found that the prevalence of wheeze remained stable around 10%-11% between 1996 and 2017 which is in line with the rates in Europe presented in the latest ISAAC and GAN update of the worldwide burden of asthma in 2020. 7Although we did not find an increase in asthma symptoms, we found that the prevalence of physiciandiagnosed asthma had almost doubled from 1996 to 2017.This trend was seen in both boys and girls but the prevalence estimates were higher among the boys, as expected. 18More detailed analyses of the cohorts revealed that among those with physician-diagnosed asthma, there was no increase in asthma morbidity.For instance, the prevalence of asthma symptoms and the asthma severity score decreased significantly during the study period.Moreover, the use of any asthma medication decreased, mainly explained by decreased use of SABA, while ICS use increased.This trend is expected as children presenting with respiratory symptoms of asthma are more likely to be prescribed ICS rather than SABA now than before.Asthma treatment guidelines have been revised during the study period with increasingly stronger recommendations of anti-inflammatory treatment with ICS and avoidance of single asthma treatment with SABA.Thus, the decreased symptom prevalence is likely an effect of increased use of ICS among children with asthma, supported by a Cochrane Library Review, concluding that ICS use improved asthma control, forced expiratory volume in 1 second (FEV 1 ), and asthma symptom scores. 19rther, we found that while the prevalence of "ever" wheeze had increased, almost one-third of those with physician-diagnosed asthma neither reported current wheeze nor use of asthma medicine in 2017 compared to less than one-tenth in 1996.Thus, one explanation for the increased prevalence of physician-diagnosed asthma could be that those with mild asthma and transient wheezing were diagnosed with asthma more often in 2017 compared to the previous cohorts 20 and that they were in remission at 8 years  TA B L E 3 Changes in prevalence of asthma symptoms, use of asthma medication, and impact on daily life among children with current asthma.  of age.Previous studies of the 1996 cohort showed that having a less severe disease was associated with remission of asthma from childhood until late teenage. 213][24][25] Taken together, our findings suggest that the increased prevalence of asthma partly can be explained by increased diagnosis of asthma at an earlier age, particularly among children with mild or transient asthma symptoms.
During the 20-year study period, there have been changes in national guidelines for diagnosis and treatment of asthma among children that probably affect our results.Additionally, the general knowledge of asthma has most likely increased.We found that among those who reported wheeze, the proportion with an asthma diagnosis as well as use of asthma medication increased.This could be an indication of increased awareness that wheezing is a sign of asthma that warrants seeking health care.Other factors that could have affected prevalence trends of asthma in childhood are variations in risk factors such as environmental exposures and allergic sensitization.Even though we did not explore these factors in the current study, we have previously demonstrated significant decreases in exposure to passive smoking, house dampness, and respiratory infections between 1996 and 2017. 11,26In contrast, during the same time period, allergic sensitization to inhaled allergens increased substantially in the study area. 11,14The cases of asthma that could be attributed to allergic sensitization did, however, decrease between 1996 and 2017. 11Thus, in-depth analyses of changes in risk factor patterns over time are warranted.
The strengths of this study include the recruitment of three representative samples of the general population, with 10-year intervals in the same geographical area, and the use of identical methods in the surveys.The response rates were exceptionally high in all three surveys, >90%, which reduced the risk of selection bias.A limitation was that the physician diagnosis of asthma was self-reported and not clinically verified in all three cohorts.A clinical validation study was performed in the first cohort recruited in 1996, and it showed that the question about physician-diagnosis of asthma had 99% specificity while the sensitivity was lower, 70%, which indicates underdiagnosis of asthma in 1996. 27 conclusion, the prevalence of asthma-related symptoms was stable in Swedish 8-year-olds from 1996 to 2017, while the prevalence of physician-diagnosed asthma doubled.Among children diagnosed with asthma, the prevalence of asthma symptoms, the impact on daily life, and asthma severity decreased, while the use of ICS increased.Taken together, these findings indicate an increase in physician-diagnosed asthma in part explained by more frequent and earlier diagnosis among those with mild asthma.The increased use of ICS can explain the stable prevalence of wheezing in the cohort and the decreased prevalence of symptoms among those with asthma.

AUTH O R CO NTR I B UTI O N S
Linnea Hedman: Investigation; funding acquisition; writing -original draft; visualization; writing -review and editing; formal analysis; TA B L E 4 Prevalence trends of asthma symptoms among those with physician-diagnosed asthma currently using and not using inhaled corticosteroids (ICS).

1996, n = 3430 2006, n = 2585 2017, n = 2785 Difference by year, p-value Test for trend, p-value
3.4 | Trends in asthma symptoms among users and non-users of ICSAmong those with physician-diagnosed asthma who had used ICS in the last 12 months, the prevalence of wheezing in the last 12 months decreased significantly, from 84.3% in 1996 to 81.3% in 2006 and 66.5% in 2017, p < .001(Table4).Moreover, markers of more severe symptoms decreased, i.e., the number of wheezing episodes, sleepdisturbing and speech-limiting wheezing, troublesome asthma, and interference with daily life.Also, among non-users of ICS with physician-diagnosed asthma, the prevalence of wheezing in the last 12 months decreased significantly, from 61.8% to 31.7% and 19.3% Changes in prevalence of asthma symptoms, use of asthma medication, and impact on daily life among children with and without asthma.
1 The prevalence of asthma-related symptoms and asthma among 8-year-old girls and boys in 1996, 2006, and 2017.Chi-square test and test for trend among girls: study year versus wheeze last 12 months p = .518and .756;versus use of asthma medication p > .001and <.001; versus physician-diagnosed asthma p < .001and <.001; and versus sleep-disturbing wheeze p = .369and .349.Chi-square test and test for trend among boys: study year versus wheeze last 12 months p = .096and .475;versus use of asthma medication p > .001and <.001; versus physician-diagnosed asthma p < .001and <.001; and versus sleep-disturbing wheeze p = .085and .882.TA B L E 2