Fig. 2
From: Immune biomarkers link air pollution exposure to blood pressure in adolescents
V-plots generated from continuous PLS models for air concentrations in PM2.5, polycyclic aromatic hydrocarbons, carbon monoxide, ozone, nitrogen oxide and nitrogen dioxide. Influential biomarkers (VIP > 1.4) are named. PLS models explained 8.5% (PM2.5); 21.5% (PAH), 5.8% (O3), 62.7% (CO), 41.2% (NO) and 72.8% (NO2) of the pollutants’ variance. Correlation coefficients were scaled and centered. Analyses accounted for the variability in age, sex, BMI, asthma and smoking. Mono, monocytes; Treg, T-regulatory cells; Th1, T-helper cells; CRP, C-reactive protein; MCP, Methyl-accepting chemotaxis protein; ADAMTS, A Disintegrin and Metalloproteinase with Thrombospondin motifs; MPO, myeloperoxidase; PLS, partial least squares; VIP, variable importance in projection, CRP, c-reactive protein; CO, carbon monoxide; NO, nitrogen oxide; NO2, nitrogen dioxide; O3, ozone; PAH, polycyclic aromatic hydrocarbon; PM, particulate matter