Airborne Benzo[a]Pyrene may contribute to divergent Pheno-Endotypes in children

Abstract
Background
Asthma represents a syndrome for which our understanding of the molecular processes underlying discrete sub-diseases (i.e., endotypes), beyond atopic asthma, is limited. The public health needs to characterize etiology-associated endotype risks is becoming urgent. In particular, the roles of polyaromatic hydrocarbon (PAH), globally distributed combustion by-products, toward the two known endotypes – T helper 2 cell high (Th2) or T helper 2 cell low (non-Th2) – warrants clarification.

Objectives
To explain ambient B[a]P association with non-atopic asthma (i.e., a proxy of non-Th2 endotype) is markedly different from that with atopic asthma (i.e., a proxy for Th2-high endotype).

Methods
In a case-control study, we compare the non-atopic as well as atopic asthmatic boys and girls against their respective controls in terms of the ambient Benzo[a]pyrene concentration nearest to their home, plasma 15-Ft2-isoprostane (15-Ft2-isoP), urinary 8-oxo-7,8-dihydro-2′-deoxyguanosine (8-oxodG), and lung function deficit. We repeated the analysis for i) dichotomous asthma outcome and ii) multinomial asthma—overweight/obese (OV/OB) combined outcomes.

Results
The non-atopic asthma cases are associated with a significantly higher median B[a]P (11.16 ng/m3) compared to that in the non-atopic controls (3.83 ng/m3; P-value < 0.001). In asthma-OV/OB stratified analysis, the non-atopic girls with lean and OV/OB asthma are associated with a step-wisely elevated B[a]P (median,11.16 and 18.00 ng/m3, respectively), compared to the non-atopic lean control girls (median, 4.28 ng/m3, P-value < 0.001). In contrast, atopic asthmatic children (2.73 ng/m3) are not associated with a significantly elevated median B[a]P, compared to the atopic control children (2.60 ng/m3; P-value > 0.05). Based on the logistic regression model, on ln-unit increate in B[a]P is associated with 4.7-times greater odds (95% CI, 1.9–11.5, P = 0.001) of asthma among the non-atopic boys. The same unit increase in B[a]P is associated with 44.8-times greater odds (95% CI, 4.7–428.2, P = 0.001) among the non-atopic girls after adjusting for urinary Cotinine, lung function deficit, 15-Ft2-isoP, and 8-oxodG.

Conclusions
Ambient B[a]P is robustly associated with non-atopic asthma, while it has no clear associations with atopic asthma among lean children. Furthermore, lung function deficit, 15-Ft2-isoP, and 8-oxodG are associated with profound alteration of B[a]P-asthma associations among the non-atopic children.