Gene-environment interaction and acute childhood bronchitis

Abstract

Gene-environment interaction studies have often focused on chronic diseases in children such as asthma. This study, for the first time, investigates the relationships of second-hand tobacco smoke (SHS), polycyclic aromatic hydrocarbons (PAH) and PM2.5, with physician diagnosed acute bronchitis in the first two years, stratified by genotypes. In the Czech Republic 1133 randomly selected children, born between 1994 and 1998, was followed-up prospectively from birth, of these 793 were genotyped. Pediatric records were abstracted for acute bronchitis episodes. Information on several confounders including SHS exposure (y/n) was obtained through questionnaires and urinary cotinine verified the latter. Air monitoring provided ambient PAH and PM2.5 data. DNA was extracted from mid placental tissues, opposite the umbilical cord. We examined six single nucleotide polymorphisms (GSTM1, GSTP1, GSTT1, CYP1A1 MspI, EPHX1 exon 3 and 4) and one diplotype (combining EPHX1 exon 3 and 4). To investigate effect-measure modification we constructed logistic models using generalized estimating equations (for repeated observations), stratified by genotypes. Owing to the daily follow-up we had about 550,000 observations, which permitted us to perform gene-gene-environment analyses. The genotype sub-sample resembled the follow-up sample with respect to baseline characteristics, assuring no selection bias. Also, genotype prevalences were comparable to Caucasian populations. Bronchitis incidence was 55.3 per 1000 child-months. Children with histidine alleles at EPHX1 exon 4 had a significantly higher association for SHS, OR=1.62 (95% CI: 1.20, 2.19) while for PAH and PM2.5 it was higher at exon 3, OR=2.05 (95% CI: 1.25, 3.36) and 2.42 (95% CI: 1.61, 3.65), respectively. The EPHX1 low activity diplotype had consistently higher association across all three exposures, for SHS it was 1.92 (95% CI: 1.25, 2.95), while for PAH/PM2.5 the OR was in the order of 1.5. In gene-gene-environment analysis, children with both low activity EPHX1 diplotype and CYP1A1 MspI genotype had higher association for PAH [2.43(95% CI: 1.23, 4.77)] and PM2.5 [2.52(95% CI: 1.30, 4.90)]. Children with EPHX1 diplotype and GSTT1 null genotype had substantially high OR for SHS [3.65(95% CI: 1.39, 9.62]. Several xenobiotic metabolizing genes modify the associations between three classes of pollutants and acute bronchitis with consistent results for EPHX1 gene with histidine alleles.