Fig. 2

Path diagram of mediation model for TL. Path diagram of a mediation model showing the dependent variable «TL», the mediator variable «nDNA_changes» and two independent variables «Occup_PAHs» and «Environ_PAHs». An oval indicates the latent variable, square boxes indicate the observed variables, arrows specify the direction of causal flow, an arrowed route is a path, and the estimated beta coefficients with p-values appeared along the paths. The effect of one variable on another is called direct. Paths from each latent variable (nDNA_changes, Environ_PAHs) to each endogenous variable are also drawn. Abbreviations: TL = telomere length; nDNA_changes = Latent variable estimated by SEM (overall change of DNA) based on MN, Anti-BPDE-DNA adducts, p53, p16, IL-6, HICI, LINE-1, Alu, GSTM1; Environ_PAHs = Latent variable estimated by SEM (environmental exposure to PAHs) based on diet, home, indoor and outdoor activities; MN = micronuclei; PAHs = Polycyclic aromatic hydrocarbons; Occup_PAHs = (years of work in the cokery) × (1-pyrenol in µmoles/mol creatinine)