Environmental contaminant body burdens and the relationship with blood pressure measures among Indigenous adolescents using Bayesian Kernel Machine Regression: Results from the Nituuchischaayihtitaau Aschii: Multi-Community Environment-and-Health Study in Eeyou Istchee, Quebec, Canada, 2005–2009

Abstract

Early onset hypertension among youth has important cardiovascular health implications later in life. Persistent organochlorine pollutants, polybrominated diphenyl ethers (PBDEs), and toxic metals have been associated with adverse cardiovascular outcomes. This study examined body burdens of environmental contaminants and blood pressure measurements (BP) among adolescents. We used data collected from the Multi-Community Environment-and-Health Study. First Nation adolescents with valid contaminant and BP measures were eligible. Principal Component Analysis (PCA) produced orthogonal principal component (PC) axes that were subsequently used as independent predictors in modified Poisson regression models, with robust variance estimation. Prevalence ratios (PR) with corresponding 95% confidence intervals (95% CI) were derived for elevated blood pressure (EBP), which was defined as systolic (SBP≥120 mm Hg or diastolic DBP)≥80 mm Hg according to new clinical practice guidelines for BP screening among adolescents between 13 and 19 years of age. Bayesian Kernel Machine Regression (BKMR) further examined individual contaminant contributions on continuous BP measures. In total there were ( n =148; 47 % females) First Nation adolescents. Prevalence of EBP, was 29%. Four PCs were extracted from the PCA analysis. The first axis (PC-1) explained 45% of the total variance, showing positive, primarily high loadings from polychlorinated biphenyl congeners (PCB). PBDEs loaded highly on PC-2, explaining 18% of the variation. Toxic metals (i.e., lead and mercury), and cadmium highly and positively loaded on PC-3 and PC-4, accounting for 11% and 10% of the variation, respectively. EBP was significantly associated with increasing PC-2 (i.e., PBDEs) across all adjusted models, fully adjusted model (PR=1.21; 95% 1.05, 1.40). Results from BKMR bolster the findings from the PCA and regression models. In this cross-sectional analysis, exposure to PBDE, specifically, PBDE-47 may adversely affect BP measures among young First Nation adolescents in Canada.