Prevalence and risk factors of asthma among 6-12-year-old schoolchildren in a metropolitan environment – a cross-sectional, questionnaire-based study

Background A cross-sectional, questionnaire-based study was conducted in primary schools located in the capital of Hungary. We aimed to evaluate the prevalence of asthma and its risk factors within a 6-12-year-old population. Methods After meeting selection criteria, 3836 eligible parent-reported questionnaires were evaluated. The survey included the ISAAC phase three core questions for asthma and assessed the association with other atopic conditions and various environmental, lifestyle and nutritional risk factors. Results Cumulative asthma had a prevalence of 12.6% among the sampled population, with a girl-boy percentage of 37.4–62.6%. The proximity of any air-polluting factories, heavy-vehicle trac and weedy area associated with greater risk for asthma while a suburban residence showed lesser odds (odds ratios were 1.3319, 1.2883, 1.3939 and 0.6390 respectively). Indoor smoking, visible mould, and keeping a dog were dened as risk factors for asthma (odds ratios: 1.6509, 2.1282, 1.4362), while the presence of a plant in the bedroom and pet rodents associated with lower odds ratios (0.7884, 0.7231 respectively). The consumption of fast food, beverages containing additives and margarine were signicantly higher in asthmatics (odds ratios 1.7488, 1.2669, 1.3549), while we found frequent sport activity (0.6883) and cereal intake (0.5403) had favorable odds ratios for asthma. Conclusions Some of the obtained results have conrmed the outcomes of previous similar epidemiological studies, while in other cases, differences were found. These differences could have been a natural consequence of regional variability and other undiscovered factors; however, the correct evaluation of these controversial results require further investigation in the future. The current data can serve as a milestone for a prospective trial. Having the ndings of this study, we can articulate recommendations for the public regarding the most common risk factors of asthma that should be avoided at and in our and how to transform our daily habits for a prosperous future.


Background
Asthma is a heterogeneous disease with time-varying respiratory symptoms (including wheeze) and is often associated with airway hyperresponsiveness and in ammation (1). Taking the history of its characteristic symptoms is the key to the diagnostic work-up, reinforced with evidence of uctuating air ow limitation. Questionnaire-based studies are amenable tools to estimate asthma prevalence, especially for large-scale surveys in pediatric focus groups, with whom a physician's o ce visit is impractical (2). The International Study of Asthma and Allergies in Childhood (ISAAC) included three research phases between 1991 and 2012 and developed various questions to investigate asthma and allergy (3). The rationale behind the ISAAC initiative was to conduct epidemiologic research concerning the more recent symptoms that covered the past 12 months, to minimize recall errors (1). The core questions of the ISAAC surveys are still favorable tools for such epidemiologic research in respiratory medicine.
Our questionnaire-based study was conducted in Budapest, Hungary, in a metropolitan environment. We embedded the ISAAC core questions into a form to evaluate the current asthma prevalence in the 6-12- year-old population. A secondary aim was to reveal which environmental risk factors are the most common in the region of interest and which habits endanger a primary schoolchild's health by predisposing to asthma.

Methods
Ethical considerations: The study was endorsed by the Ethics Committee of Heim Pál National Pediatric Institute, Budapest (KUT-19/2019). Designing the study and collecting, handling and processing the scienti c data was carried out according to the principles of the Helsinki Declaration. Informed consent was obtained from all responders, parents and/or legal guardians as children under 16 years of age were involved in the study.

Study design:
This cross-sectional, questionnaire-based study was carried out in September 2019 in Budapest, Hungary.
A total number of twenty-one primary schools in eight districts of Budapest were randomly selected from the listings provided by the Central Data Processing and Registration O ce of the Hungarian Ministry of Interior. Parents of 6-12-year-old children, at the rst teacher-parent meetings of the school year, were asked to complete the survey. The teachers provided detailed instructions before completion. The questionnaires were collected at the end of teacher-parent meetings or within a week timeframe.

The questionnaire
The enrolled parents received the multi-aspect questionnaire. In the present study, we summarized the results related to asthma and its putative risk factors. To address the prevalence of asthma and its risk factors, we included the core questions for asthma according to the phase three manual of the ISAAC supplemented with self-designed queries on physician-diagnosed asthma and risk factors (4).
Subjects who responded "Yes" to the ISAAC core question "Has your child had wheezing or whistling in the chest in the past 12 months?" constituted the "current wheezing" group (CW). The prevalence of "physician-diagnosed asthma" (PDA) was determined based on the answers to the question "Has your child had asthma diagnosed by a physician?". The union of CW and PDA sets de ned the "cumulative asthma" group (CA), representing the overall prevalence of asthma among the focused age group of pupils (Fig. 1).
We also included questions dealing with associating atopic conditions (A), including food allergy, asthma and allergic rhinitis.
Environmental (E), lifestyle (L) and nutritional (N) factors were evaluated in association with CA group.
The presence of a given environmental factor was addressed in the form of a yes-no question (E01-E34).

Page 4/25
ZIP codes were registered to distinguish whether a student had a residence in the capital or the suburb.
The frequency of lifestyle factors was determined by questions expecting answers of three prede ned choices (L01-L02), except for evaluating smoking behavior (L03), when a polar question was used. The frequency of weekly consumption of nutritional factors was initially planned to be evaluated upon three prede ned choices: "≥3 times", "1-2 times", "Rarely [< 1 times]" (N01-16). Despite the given options, parents often responded with other intervals (e.g. 2-3 times a week) covering two choices. To overcome this, answers were rearranged into two intervals during evaluation: "Frequently [≥ 1 times per week]" or Rarely [< 1 times per week]. These derivative answers were marked with the letter "d" in the report (E01d-E16d). We asked for anthropometric data of the subjects, including heights and weights, and age-speci c body mass index (BMI) percentiles were calculated.

Statistical analysis and data visualization
The data were characterized by standard descriptive statistics: frequencies (percentages) and means for categorical and quantitative data, respectively. Binomial logistic regression and chi-square were used to compare frequencies, and the t-test was used to compare the means of groups. Results were considered statistically signi cant at p < 0.05. In the case of categorical variables, odds ratios (OR) and 95% con dence intervals (95% CI) were calculated to establish how much more likely it was that someone who had the risk factor would develop asthma than someone who did not have it. Prevalence-estimates were calculated by dividing positive responses to the given question by the total number of completed questionnaires. Percentages were calculated by dividing the frequency by the total number of observations, excluding missing answers, and then multiplying by 100. All analyzes were performed with the R-3.6.2 for Windows statistical program software (5).

Results
Data acquisition, re nement and demographic parameters of the study group: A total of 6869 questionnaires were distributed in the 21 primary schools. Altogether 3885 forms were returned, of which 49 had to be ruled out due to technical reasons or the subject's inappropriate age. Of the 3836 eligible students 51.6% (n = 1979) were girls and 48.4% (n = 1857) were boys. The mean age of the children was 10.33 years +/-1.68 (Fig. 2).

Prevalence of Asthma
In the last 12 months according to the ISAAC core questions, 9.3% of responders (n = 356) experienced wheezing (Table 1). This set of subjects represented the current wheezing (CW) group (Tables 1 & 2). The symptoms' frequency and characteristics are also included in Table 1. Physician diagnosed asthma (PDA) was 6.5% (n = 248) of the pupils regardless of being symptomatic or not in the previous year (Tables 1 & 2). Thus, cumulative asthma (CA) had a prevalence of 12.6% (n = 484) among the 6-12-yearold schoolchildren in the sampled environment (Table 2). Girls had signi cantly lower chance to develop The Severity of Asthma and Associating Atopic Diseases The ISAAC core questions also describe the severity of a subject's disease. Table 3 demonstrates the frequency of such symptoms (including sleep disturbances, speech limiting wheezing and wheezing related to physical exercises) in the CA group. As parts of the atopic march, eczema, food allergy and allergic rhinitis had signi cant associations with CA (Table 4). Physician-diagnosed allergic rhinitis had the greatest odds for asthma (OR = 6.2110, CI: 4.9013-7.8591, p < 0.0001) out of this sequence. Thirty-four questions aimed to analyze any possible relationships between the prevalence of cumulative asthma and local environmental risk factors (E01-E34; Additional le 1). Figure 3 visualizes the statistical association between CA and the risk factors.
Schoolchildren who lived in the suburbs had a remarkably lower chance to develop asthma than those who had their residence in the capital (OR = 0.6391, CI: 0.4255-0.9260, p = 0.0236).
Among indoor causative agents, smoking at home had a crucial role in the prevalence of asthma. Whether the child had been exposed during the rst year of life   (Fig. 4).
Based on the available data, we were able to calculate the BMI-for-age percentile of a child in the case of only 3295 participants. Among them, we identi ed an association (Chi-square = 16.26 df = 5 p = 0.0062) between the percentile values and cumulative asthma (Fig. 5

, Additonal le 3). Children with higher BMI
were more likely to have asthma.

Discussion
In the present study the prevalence of current wheezing based on the corresponding ISAAC core question "Has your child had wheezing or whistling in the chest in the past 12 months?" was 9.3% among the 6-12-year-old children in the metropolitan area. Physician-diagnosed asthma had a 6.5 % of prevalence (Tables 1 & 2). Union of these two sets of students de ned a cumulative asthma prevalence at 12.6 % ( Table 2) An almost population-wide screening of primary school students was carried out in Tyrol, Austria (8). The mean age of the participants was 8.4 years (SD ± 1.2). Besides the current wheezing, de ned by the ISAAC manual, they used an extended de nition for asthma by massing subjects with current wheeze or use of an asthma spray ever or recurrent wheezy bronchitis or a doctor diagnosis. 10.3 % of the total study population had current wheezing, which is 1 % higher than in our subjects. Doctor-diagnosed asthma was at 3.4 %. Based on their residence, Tyrolean students were sub-classi ed into farm children, rural children and Innsbruck-town children, and their exposure to particular agents (e.g. hay-loft, animal shed or farm milk). They found living on a farm as protective but only for those with regular exposure to farming agents. In our study, children commuting from the suburbs had a lower risk of asthma. With no history of exposure to particular agents 21.3 % of Innsbruck-town children ful lled the extended criteria of asthma. This is 1.70 times higher than our cumulative asthma results, which can be due to either the regional differences or the different sampling protocol.
In a study from Romania, asthma-like symptoms (dominantly dry cough) were reported among 20% of the participating students (9). 48% of the subjects were exposed to environmental tobacco smoke, one of the expected triggers responsible for the high rate of symptoms.
Out of our participants, 9.65% (n = 370) shared their home with a smoker relative (Additional le 1), which is a remarkable difference compared to the Romanian data. Sixty-eight (18.15%) out of them t into the CA group and veri ed tobacco exposure as a major risk factor for asthma (Fig. 3). Indoor tobacco fume also impacts other respiratory outcomes in children, including pneumonia, night cough and croup (10).
Increasing evidence has con rmed mould as another predisposing factor in children (11)(12)(13)(14)(15). Visible mould can be a source of fungal spores and other volatile organic compounds. These viable and nonviable particles are su cient enough to increase the risk of asthma (13). In our sample, mouldy surfaces doubled the chance of asthma development (Fig. 3). Fagbule and Ekanem found that mould was only harmful in the bedroom, and on the contrary, it had a somehow protective effect elsewhere at home (16). The authors noticed that children could also bene t from indoor plants (16). Our study also revealed this favorable correlation between the presence of plants in the bedroom and asthma prevalence (Fig. 3). It is challenging to explain this phenomenon because potted plants' soil could serve as an origin for fungal and other biological agents.
The literature is controversial on the role of pet allergens in asthma prevalence (17)(18)(19). We found that overall pet ownership did not associate with asthma in accordance with current meta-analyses (19); however, pet-speci c risks differed. A UK birth cohort showed that early childhood ownership could have a prophylactic effect, but surrounding rabbits or rodents could contribute to non-atopic asthma with a higher-odds (20). Exposure to mouse antigens could associate with wheezing (21). Among the subjects of the current study, a rodent's presence resulted in lower odds of asthma (Fig. 3). Inconsistency might be a result of a non-species-speci c investigation. In addition, rodents were subjects of leisure pet ownership in our study, but the exposure to their antigens can also be from pests invading households. Such circumstances may also associate with other pollutive agents and a lower socioeconomic environment.
Our observations concluded dog-keeping as a potential risk factor of asthma, but cat-ownership did not (Fig. 3). An explanation could be the so-called cat paradox (20,22,23). The increased odds associated with dog ownership argue with the contemporary view (19). The latter statistical relationship seems plausible but might be casual; a more detailed assessment of pet ownership should be conducted in the future.
Outdoor air pollution also plays a role in asthma development. In the current setting, the proximity of any air-polluting factories or heavy vehicle tra c associated with the risk of asthma (Fig. 3). Although numerous reports support these ndings, from an environmental health point-of-view the particular composition and concentration of airborne pollutants instead re ect the association (24)(25)(26)(27).
The protective effect of suburban living can also be a surrogate reference of this association between air pollution and asthma because exposure to ubiquitous atmospheric agents is suspected to be lower at those sites.
Living in a weedy area is considered a risk factor of asthma development with an OR of 1.3319 (Fig. 3). It is also associated with allergic rhinitis prevalence in the same environment, where common ragweed Ambrosia artemisiifolia is the most widespread cause of allergy-associated symptoms (28). However, a higher level of pollen load showed a non-signi cant association with allergy risk in Budapest (29).
Despite the geographical distance, it is worth noting that 14% of asthmatic children of the 3-11-year-old population sensitized against ragweed in a US-based study (30).
Measurement of physical activity is challenging to carry out: besides the duration and frequency, the intensity is supposed to be a question of interest. Questionnaire-based research is a limiting factor of good quality and comparable data. Our current survey reported a signi cant association when children realizing vigorous activity more than three times a week were less likely to be in the cumulative asthma group (Table 5). Visual display time as a reference for sedentary lifestyle correlated with the odds ratio of asthma, but the relationship was not statistically signi cant ( Table 5).
The ISAAC Phase Three summarised similar tendencies, but the methodology was different (31). Other studies with different approaches reported an ambiguous correlation between physical activity and asthma prevalence (32)(33)(34)(35)(36). We must emphasize that asthma has also been reported as a barrier to physical activity; thus, in the future, we should analyze this relation through an interdisciplinary lens (32).
Emerging technologies, including wearable biosensors or smartphone applications, may serve as more accurate physical activity data resources in the future.
Dietary patterns in uence the risk of asthma development. A Mediterranean-style diet, rich in fruits, vegetables and whole grains while taking less meat and dairy in, and other plant-based foods have associated with reduced odds for asthma (37)(38)(39)(40)(41). A Westernized diet, including a predominant amount of animal products with higher fat intake and lower bre consumption, is a risk factor of asthma (39). It would be welcome to keep adherence to a healthier diet where it is already historically and geographically predisposed; trends showed an increased prevalence of asthma symptoms in the Mediterranean and Latin-America (39,42). In the current sampled population, such emblematic meals of urbanized living like fast-food and drinks with arti cial additives associated with cumulative asthma (Fig. 4).
Regular consumption of margarine also showed an association with cumulative asthma (Fig. 4). A high rate of fat intake is also attributed to the Westernized diet, but there is more emphasis placed on the composition of fat fractions. According to the lipid hypothesis, an increased intake of polyunsaturated fatty acids (PUFAs) over saturated fat can increase asthma prevalence and allergic sensitization, but controversial results are also available (39,40,43,44). Margarine can serve as a signi cant source of PUFAs. Therefore, we should be aware of its potential causative role in asthma and atopy development (43,(45)(46)(47).
Frequent intake of cereals showed an association with decreased odds of cumulative asthma (Fig. 4).
The literature has supported this nding, though the protective mechanism is not completely understood yet (39-41, 44). Our survey did not evaluate prior cereal consumption, although the early introduction of cereal grains into the diet is substantial to avoid sensitization (48).
High energy diet is a predisposing factor of obesity. Obesity is associated with worse asthma control and an increased risk of exacerbations in all ages (39). Besides the fact that obese and overweight children are more likely to have asthma, according to the current view, obesity-related asthma in childhood is a separate entity with a Th 1 cell polarization (39,49,50). Our observation reinforced the conception of the association between asthma and BMI-for-age percentiles, but the phenotyping of obese asthmatics was beyond our scope (Fig. 5).

Conclusions
Our study delivered data on the recent prevalence of asthma among 6-12-year-old children in Budapest, Hungary. In this urban environment, we identi ed an increased asthma prevalence compared to some previously published studies, but the cross-sectional design and the different methodology did not permit us to draw timeframe-dependent conclusions. Besides the outdoor and indoor environmental factors, we also analyzed the contribution of lifestyle and nutritional determinants. Though the statistical associations revealed are not numerous, but the majority t in the existing literature.   De nition of cumulative asthma: the union of current wheezing and physician diagnosed asthma sets. Association between cumulative asthma and BMI-for-age precentiles. (Chi-square= 16.26 df= 5 p= 0.0062)