Bibian van der Voorn

54 CHAPTER 4 INTRODUCTION Anxiety or depressive disorders are associated with alterations in HPA axis activity 1,2 , which, when present during pregnancy, have been associated with permanent alterations in offspring’s HPA axis activity and neurocognitive development 3-5 . During pregnancy, anxiety and depressive disorders are frequent, with numbers ranging from 1 in 10 to 1 in 5 pregnant women 6-8 . Although many observational studies described associations between prenatal stress and unfavorable infant outcomes, caution must be exercised in the interpretation of these findings due to the use of subjective measures of stress, while indices of HPA axis activity are lacking 4,9 . As part of physiological changes during pregnancy, both maternal and fetal glucocorticoids exert a positive feedback effect on the placenta by stimulating the synthesis of placental CRH. Due to this physiological feed forward response, maternal cortisol increases during gestation 3 . In addition, pregnancy is also characterized by high levels of estrogens, which have a well-known stimulating effect on CBG 10 . CBG, in turn, binds the majority of free cortisol. Moreover, placental 11βHSD type 2 converts maternal cortisol to inert cortisone. Accordingly, the fetus is protected from overexposure to maternal cortisol 11 . Hair cortisol and cortisone levels are indices of long-term glucocorticoid (GC) exposure, at least in adults and children above the age of 4 years 12-14 . Newborn hair GC levels are thought to offer a retrospective view on GC exposure during the last part of pregnancy 15,16 , although various other factors could explain inter- and intra-individual differences 17,18 . Kapoor et al. studied hair GC levels in the offspring of rhesus monkeys that were randomized to receive exposure to frightening noises while being in a darkened room for 10 minutes a day 5 days a week during 20% of their pregnancies, and found decreased hair cortisol in prenatally exposed offspring, but no difference in hair cortisone 19 . In humans, Hoffman et al. studied neonatal and maternal hair cortisol directly postpartum (pp), and showed that neonatal cortisol increased with advancing gestational age and birth weight 20 . Unfortunately, in this study stress exposure was not taken into account, and hair cortisone levels were not measured. Nonetheless, these data suggest that determination of neonatal hair GCs might offer valuable insights into intra-uterine GC regulation. Therefore, in the present study we assessed whether exposure to maternal stress pre- and postnatally is associated with neonatal and maternal hair cortisol and cortisone levels directly pp. To study this, we used data from a cohort in which women with severe stress during pregnancy were overrepresented.