Early childhood traffic-related air pollution and risk of allergic rhinitis at 2–4 years of age modification by family stress and male gender: a case-control study in Shenyang, China

Allergic rhinitis (AR) is the most common chronic allergic disease, involving inflammation of nasal eosinophils caused by air allergen sensitized by IgE [1]. AR is also considered a global health problem. Recently, the prevalence and incidence of AR have been increasing worldwide, especially in developing countries [2, 3]. AR will not cause very serious damage to the body, and preschool children cannot clearly express their condition. Therefore, children with AR are easy to be ignored, especially for some Chinese children [3]. If AR continues to develop, it will cause more serious complications in the upper and lower respiratory tracts, including asthma and sinusitis [4], which will seriously affect the child’s physical and social activities [5]. The etiology of AR is related to genetic and environmental risk factors. Since the rapid increase in the prevalence of AR in children is unlikely to be attributed solely to genetic changes, environmental factors have largely contributed to the development and deterioration of AR in China in the past decade [3]. Environmental factors include physical and social environmental exposure. However, there is little information about the potential effect modification of the social environment (family stress) on the relationship of children AR to the natural environment (traffic-related air pollution).

As a physical environmental factor, traffic-related air pollution (TRAP) can reduce lung function, cause asthma exacerbation, and lead to the development and exacerbation of allergic diseases, such as eczema, asthma, AR, and sensitization [6‐8]. A meta-analysis showed that the prevalence of children AR increased with exposure to nitrogen dioxide (NO₂), sulfur dioxide (SO₂), and particulate matter (PM₁₀ and PM_2.5), but the relationship between SO₂/PM₁₀ and prevalence of AR was not obviously closely in Asia [9]. Air pollutants come from many emission sources in cities, such as automobile exhaust, household oil fume, and industrial waste gas. In the past 10 years, the air pollution situation in Shenyang has changed. Motor vehicle traffic emissions are the main source of air pollution [10]. TRAP is the main source of changes in the concentration of air pollutants in cities [11]. A cohort study showed TRAP exposure during pregnancy and the first year of life was related to the development of AR in preschool children [3]. In addition, a case-control study showed that for residents less than 75 m from the main road increased lifetime diagnosis and AR symptoms of children aged 6 to 14 [12]. However, previous studies showed no significant correlation between TRAP exposure in early years of life and AR prevalence in children aged 1–12 years [13, 14]. These inconsistencies in the epidemiological survey results are mainly due to differences in TRAP exposure assessment and the classification of the subgroups included. These studies assessed TRAP exposure through traffic measurements or modeling air pollutant concentrations, and most studies only considered the home address at birth [3, 12‐14]. Furthermore, these studies included children aged 1–12 years and rarely included a narrow subgroup. Therefore, we need to limit the air exposure window of children with AR and the key components of air pollution in order to develop more effective prevention and intervention measures.

As a social environmental factor, psychological stress is related to allergic diseases including rhinitis, rhinoconjunctivitis, allergic conjunctivitis, wheezing, and asthma in children [15‐18]. In recent years, little consideration has been given to the psychological impact of family stress on preschool children in China [19]. This is mainly because under the control of traditional Chinese culture, preschool children are unable or unwilling to express their feelings to their parents. Especially, in the face of stress, the depressive performance of parents will damage the entire family environment, thereby inhibiting the psychosocial functions of preschool children [20]. Family stress (including separation/divorce of parents, unemployment of parents, serious health problems, death of family members or close relatives) is a risk factor closely related to the occurrence or development of asthma [21, 22]. This may be due to being stimulated by antigen or stimulating immune cells in vitro [21, 23, 24]. However, few studies have explored the moderating effects of family stress and gender on the occurrence of AR in preschool children under TRAP. Therefore, this study aimed to evaluate whether preschool children are prone to develop AR during early-life TRAP exposure including PM₁₀, NO₂, ozone (O₃), carbon monoxide (CO), and SO₂ exposures, with emphasis on stress or other possible influencing factors.

In this case-control study, we explored the moderating effects of family stress on the relationship between early-life TRAP exposure and the risk of AR in preschool children. In the whole study population, PM₁₀ or NO₂ exposure was positively associated with the risk of AR in children aged 2–4 years, respectively. We also analyzed whether sex, parental allergy history, and eczema in the first year increased the risk of AR in preschool children, under PM₁₀ or NO₂ exposure. These factors were also positively correlated with AR. By exploring the potential moderating effects of numerous risk factors associated with AR (when a causal effect of intervening on one exposure, there was another variable stratification), we found that family stress or boys increased the risk of PM₁₀ or NO₂ exposure to AR.

The main findings of this study were to explore the relationship between early TRAP (PM₁₀ or NO₂) exposure and the risk of AR in preschool children. Family stress and male gender may play a moderating effect on the relationship. Using the multi-pollutant model, the risks of other air pollutants to AR were adjusted to estimate their correlation. PM₁₀, SO₂, NO₂, CO, and O₃ have been used as markers of traffic air pollution. We only found that preschool children exposed to PM₁₀ and NO₂ had an increased risk of AR by 70% and 85%, respectively.

The prevalence of AR among middle-school students in Taiwan was related to the concentrations of nitrogen oxides (NOx) and CO exposures [25]. In China, atmospheric CO and NO₂ are the emissions of automobile exhaust, which are caused by incomplete combustion inside motor engines [11]. We used a multi-pollutant model to evaluate the association between combined exposure (PM₁₀, O₃, SO₂, CO and NO₂) and the risk of AR in children, instead of choosing PM_2.5 as a marker of TRAP. Because PM_2.5 is highly correlated with PM₁₀, there may be some potential variance inflation and deviation [28, 29]. Since we did not know which model would cause the least bias in the estimation, we compared the results of the single-pollutant model and the multi-pollutant model. In most cases, in both models, the effect estimates show similar associations between exposure and dependent variables. And most of the effect estimates in the multi-pollutant model are higher than those in the single-pollutant model. This study showed the same performance. The reason was that the multi-pollutant model produced synergistic or additive effects, which was consistent with other reports on air pollution and childhood allergic diseases [3, 30]. Therefore, in future research, we need to accurately evaluate the effect and improve the evaluation of the concentrations of pollutants and covariates. The complex sources of exposure may cause PM_2.5 and PM₁₀ to be non-specific pollutants and reduce the correlation. But, this study showed that PM_2.5 and PM₁₀ are highly correlated. This aspect may be related to the current main source of air pollution in China is automobile exhaust [10]. On the other hand, it may be because PM can act as an allergen carrier, exacerbating allergic reactions, and smaller particles (PM_2.5) can enter the respiratory tract of underdeveloped lungs and anatomically smaller peripheral airways more quickly in children [9, 31, 32]. Moreover, in China, heavy metal elements (S, Zn, Cu, and Pb) account for a large proportion of PM₁₀, while S and Pb are relatively enriched in PM_2.5 [33]. The difference in chemical compositions of PM may cause different degrees of mucosal damage [9].

The levels of O₃ and NO₂ are coupled by chemical bonds. The formation of O₃ requires sunlight to consume nitrogen [34]. Thus, especially in areas with heavy traffic, when NO₂ exposure is high, O₃ exposure is usually low [35‐37]. And, NO₂ can directly cause inflammation in children’s sensitive airways, such as AR [38]. The association between SO₂ exposure and the risk of AR in children is controversial [9]. From the point of view of the decrease of SO₂ content or the mixing of multiple pollutants, due to the increase of concentrations in other air pollutants, we cannot fully explain the inconsistent ORs trend of SO₂ exposure between the single-pollutant model and the multi-pollutant model. Other pollutants may enhance the corrosiveness of SO₂ [39]. In Asia, most previous studies did not show a statistically significant correlation between SO₂ concentration and AR prevalence in children. On the contrary, some recent studies showed that SO₂ concentration was closely related to the prevalence of AR. The change of this trend may be related to different regions, races, and nationalities [9, 39], whereas this debate on the correlation between SO₂ and the risk of AR in children may be mainly due to allergen-related or neurogenic complex mechanisms [40, 41]. The difference between SO₂ single exposure and mixed exposure was due to the fine dynamic balance between pro- and anti-apoptotic genes in damaged airway epithelium and undamaged airway epithelium [40]. Unfortunately, we did not detect the related immune indicators including these anti-inflammatory cytokines. Other studies have shown that SO₂ exposure in multi-pollutant model (petrochemical exposure) had no or moderate association with children’s respiratory diseases [42‐44]. Thus, we need to expand the sample size in analyzing the correlation between SO₂ and children AR.

Because as the height of the building increases, TRAP exposure will decrease [14]. We chose children living below the 4th floor in order to more accurately estimate the actual situation of TRAP exposure. At the same time, to adjust the misclassification bias from TRAP exposure, we collected every child who did not change the family residence during the observation period, and his/her home was very close to the kindergarten, or did not even go to kindergarten. In addition, we adjusted the statistical model for other sources of children’s exposure to TRAP (such as smoking at home).

Respiratory and immune system develop and mature rapidly in early childhood, so TRAP exposure during this sensitive period is more likely to damage children’s health [40, 41]. Thus, we chose the early exposure as the observation window for TRAP exposure to help assess the associations similar to other study, which have shown that in the third trimester and the first year of life, TRAP was more dangerous for asthma, AR, and eczema than exposure later [42]. Moreover, TRAP exposure during early developmental period could trigger AR later (aged 4 years old) [43]. The correlation between TRAP and AR was more obvious in children aged 3–4 years old [3]. Allergic children exposed to TRAP are more likely to develop asthma and allergy than non-allergic children [7, 43]. However, we only found that stress and gender factors may exacerbate the risk of TRAP on AR and did not find the influence of genetic susceptibility on AR. This may be related to the multi-collinearity from the two variables (parental allergy history and infant eczema) in this study. A meta-analysis reports that psychological stress will cause pregnant women to continuously secrete cortisol, which in turn increases the risk of asthma and allergy in offspring [44]. This may be due to the differentiation of dominant T helper 2 (Th2)-biased cell, which causes Th1/Th2 homeostasis imbalance in the immune system [45, 46].Thus, stress may increase the risk of TRAP by disrupting the immune response [47‐49], even at lower doses of physical exposure to promote adverse consequences [50, 51]. Whether there is a difference between gender in the associations between TRAP exposure and respiratory diseases is still controversial. For instance, a French birth cohort study showed that there was a positive association between nitrogen oxides and persistent wheezing in 4-year-old boys [46], while a Swedish study showed similar performance in 4-year-old girls [52]. In addition, a meta-analysis of five birth cohorts showed that PM_2.5 exposure was positively correlated with asthma in boys aged 4–5 years [45]. A Chinese similar cohort study showed a positive correlation between TRAP exposure and AR in boys aged 3–6 years [53]. The above gender differences are due to the later appearance of male lung surfactants in newborns. From the prenatal period to the first year after birth, the respiratory tract of boys is narrower than that of girls [54, 55].

There are several limitations as follows: First, the small sample size may lead to insufficient power to assess an interaction [56]. In order to reduce the possibility of attributable to chance, the investigation needs to be repeated. Second, although we chose the straight-line distance between family and kindergarten to be less than 400 m, there may still be potential exposure misclassifications. Three, the number of monitoring points was scarce and the land use conditions were ignored; thus, the exposure measurement at fixed monitoring points may lead to exposure bias. Four, although in modern Chinese cities with similar scales and architectural styles, the results of this study could be extrapolated to assess the relationship between air pollution and childhood allergic diseases. But, as a northern city in China, the climate and humanistic characteristics of Shenyang are different from those of southern China, so further research is needed to infer these results. Five, this study only involved outdoor air pollution exposure. But because children spend 90% of their time indoors [57] and the concentration of PM₁₀ and NO₂ varies greatly between indoor and outdoor [58, 59], it may cause uncertainty in air pollution exposure [57, 58]. However, at present, urban residents in Shenyang basically use natural gas for cooking, which can reduce indoor smoke and dust pollution; moreover, Chinese families are paying more and more attention to air quality and increasing the time to open windows for ventilation, which also increases outdoor air pollution exposure [10]. It cannot be ignored that Shenyang is a heavily polluted city in the northern of China; we speculate that the serious outdoor air pollution may cover up indoor air pollution [60]. Further work is required to investigate the potential interaction between outdoor or indoor air pollution and AR. Finally, we did not collect full data at 0–1 year of age and only calculated the 2 years of age to the day of AR diagnosis. This made our study inadequate in the ability to identify the susceptible window of exposure in the development of childhood AR and possible effect modifiers. Thus, the next study should be to investigate the relation of early TARP during pregnancy and the first year of life with the development of AR in later life, and possible effect modifiers.

Our case-control study supports that family stress and male gender may increase the susceptibility of AR caused by early exposure to TRAP (PM₁₀ or NO₂) in preschool children aged 2–4 years.