Figure 3

Estimated adjusted prevalence rates of reported health outcomes. The prevalence rates are displayed as bell-shaped functions of the residential distance from highway [m]. The underlying logistic regression models included sex, age, smoking status (current, former, never), pack years of cigarettes smoked, body mass index, community of residence and the binary variables exposure to environmental tobacco smoke (ETS), ETS-exposure at work, current occupational exposure to vapors, gas, dust, fumes, or aerosols, primary school education only, doctor diagnosed asthma, maternal atopy, and severe respiratory infection in early childhood.