DISCUSSION
The role of environmental phthalate exposure on allergic airway diseases has been a topic of great interest. However, very few studies have examined the association of phthalate exposure with allergic inflammation and lung function. Our measurements of IOS in a population-based sample of children showed that an increased level of urinary phthalate metabolites was significantly associated with airway dysfunction, and that this association was partially attributable to an increased serum periostin level. Moreover, the associations we identified persisted after adjusting for multiple covariates (height, gender, BMI z -score, aeroallergen sensitization, secondary smoking, and vitamin D level).
Phthalate exposure may occur from ingestion, inhalation, dermal absorption, and parenteral administration.20,21 LMWPs are used in a variety of personal-hygiene and cosmetic products, such as nail polish and fragrances, as scent stabilizers.22HMWPs are used in plastic tubing, food packaging, containers, vinyl toys, vinyl floor coverings, and building products.21,22 Koch et al.14 showed that exposure to HMWPs was mostly due to dietary intake, and that exposure to LMWPs was mainly from non-dietary exposures, such as from personal care products, dust, and indoor air.14 In the present study, we showed that children with higher urinary levels of LMWPs had significantly greater airway resistance after adjusting for confounding variables.
Experimental studies demonstrated that exposure to phthalates increases the levels of Th2 cells and multiple cytokines, and thereby enhances airway inflammation.8,9 Clinical studies found an association between phthalate exposure and FeNO.5,8Serum periostin is a biomarker of type-2 inflammation in asthma.23 The exact function of YKL-40 remains unclear, but it consistently correlates with airway obstruction in studies of patients with asthma,24-26 and with measures of airway remodeling, such as bronchial wall thickness and subepithelial fibrosis.24,25 Little is known about the function of periostin and YKL-40 in patients with allergic inflammation related to environmental pollutants. We initially hypothesized that there may be some differences in serum levels of periostin and YKL-40 following phthalate exposure because type-2 and non–type-2-induced airway inflammation are involved in phthalate-related airway inflammation. We therefore evaluated the relationships of serum periostin and YKL-40 levels in children with phthalate exposure, and investigated their relevance to clinical characteristics and other type-2 biomarkers, including blood eosinophil counts, serum total IgE, and FeNO. We found that FeNO level was significantly associated with the quartiles of urinary Σ4HMWP metabolites, but not quartiles of urinary Σ3LMWP metabolites. Our multivariate linear regression analysis indicated that urinary Σ4HMWP and Σ3LMWP metabolites were both significantly associated with serum periostin level. These findings are clinically significant because they demonstrate an association between phthalate exposures with serum periostin, an established marker of Th2 inflammation. However, we found no significant associations in the levels of urinary Σ4HMWP and Σ3LMWP metabolites with serum YKL-40 level. These results suggest that serum periostin may be used as a biomarker for type-2 inflammation in children following phthalate exposure, but serum YKL-40 has less value for this assessment.
We found that serum periostin level was significantly associated with Rrs5 and Rrs20-5. This outcome is similar to that of a recent study which reported the relationship between periostin level and pulmonary function in asthma patients. This previous study found that patients with high periostin levels had lower FEV1/FVC values.27 Although there are reports of associations between pulmonary function and periostin level, there are only limited data on the relationships of periostin with small airway function. We therefore used an objective method — IOS — to evaluate small airway dysfunction our patients.
We found that periostin had a significant effect in mediating the relationship between urinary LMWP metabolites and airway resistance. To quantify this mediating effect, we performed a model-based mediation analysis by using the mediation package in R software.19 The algorithm uses a quasi-Bayesian Monte Carlo method to estimate the presence of mediation (average causal mediation effect/indirect effect) and the proportion of the link between phthalate exposure and airway dysfunction that is mediated by periostin.19
There are several limitations to the present study. First, because this was a cross-sectional study, we did not obtain any direct evidence for cause-and-effect relationships. Second, use of periostin as a dependable biomarker in growing children may be questionable because it is an extracellular matrix protein that is secreted by osteoblasts. However, the levels in our study subjects (10-12 years-old) were not higher than published values for adults28 and were not significantly associated with age (data not shown).
To the best of our knowledge, this is the first large-sample study of urban children to comprehensively investigate the role of periostin in mediating the relationship between phthalate exposure and airway dysfunction. Previous studies have investigated the association of phthalate exposure with other inflammation markers, but no previous studies investigated the role of periostin and YKL-40 on airway dysfunction in children exposed to phthalates. Another merit of this study is that we assessed small airway function using IOS.
In conclusion, we found that exposure to low-molecular-weight phthalates was significantly associated with airway dysfunction, and this effect is partially attributable to increased serum periostin level. Periostin appears to function in the Th2 cell-mediated inflammation that causes pulmonary dysfunction in children exposed to phthalates, but further studies are required to clarify this relationship.