DISCUSSION
Our study showed that antibiotic use during infancy increased the risk of AR in childhood, whereas the presence of older siblings had a protective effect on AR. The risk of AR increased more than twice when the ERS was high compared to when it was low. Furthermore, although each SNP, including IL-13 (rs20541), CD14 (rs2569190), TLR4 (rs1927911), and GSTP1 (rs1695) did not increase the risk of AR individually, a high PRS was associated with an increased risk of AR. This result suggests that polygenic effects generated by multiple SNPs may contribute to the development of AR. Above all, high ERS and PRS were not only independently associated with the risk of AR but also showed a combined effect on the risk of AR. These results suggest that the interaction of susceptibility of multiple genes and hygienic environmental risk factors might be associated with current AR, and can be applied to early detection of high-risk groups of school-age AR and prevention of school-age AR.
Antibiotic use in early childhood and the presence of older siblings are associated with allergic diseases. The risk of AR is increased in school children who used antibiotics in infancy and within two years after birth.25,26The use of antibiotics during early childhood and presence of older siblings has the potential to affect the distribution of gut bacteria and induce changes in early immune system formation.9,27This study showed that antibiotic use had a dose-effect, an important hygienic environmental factor increased the risk of AR, and that AR risk was known to decrease as the number of older siblings increased.
Several SNPs have been mentioned as genetic factors affecting the development of AR, of which SNP rs20541 located in exon 4 of the IL-13 gene was found to be strongly associated with high levels of plasma IgE and AR development.1,12,17In our study, SNP rs20541 had a combined effect with ERS, and SNP rs20541 increased the risk of AR by interaction with mold exposure.1 These results suggest that the SNP of the IL-13 gene is involved in the risk of AR by gene-environment interaction.
The CD14 gene, which is associated with the innate immune response and located on chromosome 5q31.3, encodes a protein that functions as a co-receptor for TLR and releases pro-inflammatory cytokines.28 The effect of CD14 rs2569190 is conflicting and influenced by lipopolysaccharides- related factors and interactions with environmental microorganisms.28 In our study, the TT genotype has a protective effect on the development of AR. A previous meta-analysis reported that CD14 SNP rs2569190 did not affect AR risk in Asians.12 These results suggest that the CD14 gene alone does not increase the risk of AR, but has different effects on the risk of AR through interaction with various environmental factors.
Toll-like receptor 4 initiates the innate immune system when exposed to environmental factors, and antagonists of TLR4 have been shown to aggravate the symptoms of AR.14 The CC genotype of TLR4 rs1927911 was associated with a higher risk of AR.13 Our results are consistent with the results of previous studies in that the CC genotype of TLR4 rs1927911 alone did not increase the risk of AR.
Glutathione S-transferase P1 is the most common form of GST found in the respiratory tract lining fluid, and the GSTP1 genotype is known to be associated with the severity of airway dysfunction.15,29Although there have been no previous studies on the association between GSTP1 genotype and AR risk, several studies showed the association between GSTP1 and development of asthma.15,16In our study, the GSTP1 rs1695 SNP showed a weak association with increase in AR risk, but did not increase the AR risk alone.
Several studies have introduced the concept of ERS to identify the effects of multiple environmental factors on the development of allergic diseases in children. In a previous study, ERS was calculated with risk factors, such as cesarean delivery and antibiotics use during infancy, which showed that children with higher scores had a higher incidence of atopic AR at school age.13 A recent study in Lebanon showed that children with higher scores based on risk factors, including environmental factors showed a higher frequency of allergic diseases, which is consistent with our study.5
Following studies on ERS, recent studies have been conducted on the effects of multiple genetic factors on the development of allergic diseases. A cohort study from Netherlands calculated a weighted PRS based on 10 SNPs associated with allergies in adults and showed that high PRS increased parental-reported allergy at 5 years of age and diagnosis of allergies in childhood by a physician.18 From two birth cohorts with 135 SNPs, PRS was associated with an increased risk of atopic march, but having a weak association with allergic diseases characterized by the presence of a single symptom.21 This suggests that multiple SNPs may influence multiple allergic comorbidities by unknown interactions. In our study, weighted PRS was calculated with four SNPs, which were related to the hygiene hypothesis. Children with high-weighted PRS had an increased risk of developing AR, suggesting that multiple SNPs may be linked to the development of allergic disease.
Our study has some limitations. First, information on antibiotic use in infancy may be biased because it is based on parents’ memories after several years have passed. However, our study was designed to study the hygiene hypothesis, and several hygiene related environmental factors were investigated in detail. Second, AR was defined in the questionnaire without laboratory tests or skin prick tests. To supplement this, we used the definition of current AR, which reflects not only the symptoms but also the history of AR diagnosis by a doctor. Third, target SNP was selected with only four hygiene related candidate SNPs.1,13,14However, the selected SNPs have been associated with allergic diseases in our previous studies.1,9,13,14Further studies using genome-wide association study or prospective birth cohort will be needed. Lastly, this study had a relatively limited number of children in Korea; however, it is a general population-based study.
Our study has the following strengths. First, our study is meaningful in that it is a general population-based study. Finally, the ISAAC questionnaire was verified in many previous studies in Korea, and the response rate to the questionnaire was high at over 95% in this study.
In conclusion, polygenic susceptibility and exposure to multiple hygienic environmental risk factors during infancy increase the risk of AR at school age, which suggests gene-environment interaction. Therefore, it is necessary to decrease exposure to unhygienic environmental risk factors in infancy to prevent AR in school children, especially in susceptible children. Further studies are needed to elucidate the mechanism for this interaction between PRS and ERS that contributes to the development of AR