The association of gestational and childhood phthalate exposure with adolescent hair cortisol: The HOME Study

Abstract

BACKGROUND AND AIM: Phthalate exposure during gestation and childhood may disrupt the hypothalamic-pituitary-adrenocortical (HPA) axis to adversely impact children’s neurodevelopment and cardiometabolic health. We evaluated associations of serial urinary phthalate metabolite concentrations with hair cortisol, a measure of HPA-axis homeostasis, to identify periods of heightened susceptibility during gestation, infancy, and childhood. METHODS: We recruited pregnant women from Cincinnati, Ohio (2003-2006) and followed their children until age 12 years. We quantified urinary phthalate metabolites twice during gestation and up to six-times during childhood (1, 2, 3, 4, 5, 8 years). At age 12 years, we quantified children’s hair cortisol concentrations (pg/mg hair). We estimated covariate-adjusted associations of log10-transformed average gestational or average childhood (1-8 years) phthalate metabolites with log10-transformed hair cortisol using linear regression. We used multiple informant models to estimate associations of serial phthalate measures with hair cortisol and assess the heterogeneity of this association across ages (n=207). RESULTS:The median hair cortisol concentration was 2.2 pg/mg (25th,75th percentile= 1.3, 5.2). We did not find evidence of an association between gestational phthalate metabolites and hair cortisol. However, each interquartile range (IQR) increase in average childhood ∑DEHP, MCPP, and MEP was associated with 289% (95%CI: 38, 996%), 78% (95%CI: -9, 255%), and 78% (95%CI: 15, 175%) higher hair cortisol concentrations, respectively; these associations did not vary by exposure period (heterogeneity p-values=0.53, 0.41, and 0.35, respectively). Positive associations of MBzP and MCOP with hair cortisol were stronger for exposure at infancy than in later childhood (heterogeneity p-values=0.06 and 0.16, respectively). An IQR increase in MBzP at age one year and MCOP at age two years was associated with 164% (95%CI: 56, 349%) and 77% (95%CI: 22, 155%) higher hair cortisol concentrations, respectively. CONCLUSIONS:Higher urinary concentrations of several phthalate metabolites during childhood, but not gestation, were associated with elevated hair cortisol concentrations in adolescence. KEYWORDS: phthalates, cortisol, development, susceptibility