Fig. 3
Modeling of estimated changes in cord blood steroids levels for IQR increments of trimester-specific ambient air pollution exposure at maternal home address(es) during pregnancy. Estimated changes in cord blood steroid levels for trimester-specific interquartile range (IQR) increments in black carbon, NO2, or PM2.5 exposure (in µg/m3) during pregnancy: (A) 17α-hydroxypregnenolone (indicated as 17-OH-pregnenolone), (B) 17α-hydroxyprogesterone (indicated as 17-OH-progesterone), (C) DHEA, (D) pregnenolone, (E) androstenedione, and (F) testosterone. The models are adjusted for the other two trimester-specific mean air pollutant exposures, sex of the child, birth weight, smoking during pregnancy, gestational age, age of the mother, the season of delivery, maternal education and interaction between sex and exposure for the trimester under study; n = 397 for PM2.5 and NO2, and n = 378 for BC. All steroid levels are expressed in ng/dL. Asterisks indicate significant estimates of the modeled associations taking into account multiple testing (CI 0.99%)