Bibian van der Voorn

62 CHAPTER 4 A trend towards a higher prevalence of elevatedHAS subscores ppwas seen inmothers who used SSRIs, as compared to mothers who did not use SADs (26.1% vs. 13.4%, p = 0.07). Mothers who used a combination of SADs had significantly (p <0.01) more often clinically elevated HAS and HDS subscores, as compared to mothers who did not use SADs (80.0% vs. 13.4% and 80.0% vs. 7.6%, respectively). DISCUSSION In this study, exposure to excessive pre- and perinatal maternal stress was associated with a decrease in neonatal hair GC levels, with the largest decrease seen in children of mothers with persistent stress throughout pregnancy. In addition, elevated maternal stress-scores around birth were associated with increased maternal hair GC levels, while elevated stress-scores earlier in gestation had no effect on maternal hair GCs. These associations could not be explained by SAD use during pregnancy. One possible explanation for the discrepancy in the association between maternal stress with maternal vs. neonatal hair GCs could lie in the difference in growth rate between fetal and maternal hair. Whereas adult hair is known to grow one centimeter per month 16,26 , fetal hair growth velocity and the timing of transition into terminal hair strands is thought to vary significantly between and within neonates 18,25,29 . In an attempt to minimize the influences of these inter- and intra-individual variations, we chose to analyze the entire hair instead of the most proximal centimeter. Accordingly, it is possible that the analyzed hair segments of mother and child reflected different windows, i.e., the last month vs. a substantial part of the second half of pregnancy 17,18 . Presumably as a result, similar to that what was seen in rhesus monkeys exposed to prenatal stress 19 , stressful events early in pregnancy were not reflected in the most proximal maternal hair segment while they were in the neonatal hair segment. In addition, the negative association between maternal stress and neonatal hair GC levels might be interpreted as a suppression of fetal HPA axis activity due to overexposure to maternal GCs. In adults, it has previously been shown that exposure to supraphysiological levels of GCs is indeed associated with suppressed hair GCs 30 . The supply of glucocorticoids to the fetus comes primarily from the placental transfer of maternal glucocorticoids, and in addition, from the fetal adrenal itself, with the latter contribution expanding throughout gestation, due to maturation of the fetal adrenal. Because evidence is lacking concerning which part of intra-uterine GC regulation is reflected in neonatal hair GCs, it has been hypothesized that neonatal hair GC levels might reflect amniotic fluid GC levels 19 . This seems unlikely, since in our study neonatal hair cortisol was higher than cortisone, whereas due to conversion of