Bibian van der Voorn

132 CHAPTER 9 DISCUSSION We showed that differences in growth between birth and 6 mo. corrected age are related to the pattern of serum cortisol decline during infancy. In our longitudinal analyses, the growth-restricted groups (i.e., low BW, SGA, AGA with GR and SGA without CUG) all had a lower cortisol over time, compared to non-growth-restricted infants (AGA infants without GR). These differences were not explained by gestational age, antenatal glucocorticoid treatment or gender. Not surprisingly, statistical correction for pregnancy induced hypertensive disorders reduced the strength of our associations (Aufdenblatten et al., 2009; McCalla et al., 1998). In cross-sectional analyses at age 8 y, the differences between the growth-restricted and non-growth- restricted groups were no longer present. In addition, salivary cortisol at age 8 y was not different between these groups at any sampling moment throughout the day. In the first weeks of life, the HPA axis of preterm newborns is still immature. Among the impairments are insensitivity of the pituitary gland to synthetic CRH, decreased 11β-hydroxylase activity in the adrenal cortex, and a cortisol-cortisone shuttle favoring cortisone 3 . In animal studies, adverse events occurring in early life have been associated with permanent alterations in HPA axis activity 8 . In line with our results, in previous studies among infants of whom the majority were born at term, a blunted cortisol response to painful procedures was found in those born SGA 23,24 . In contrast, from childhood onwards, lower birth weight has been associated with increases in glucocorticoid metabolite excretion (Clark 1996), basal cortisol 25 and the cortisol response to psychosocial stress 26 . Preterm infants were found to exhibit altered responses to different kinds of stressors, as compared to their term counterparts 27,28 . Furthermore, prematurity has been associated with a lower cortisol response, in spite of a higher pretest cortisol level, during a psychosocial stress test 27,29,30 . These findings suggest that the HPA axis is hypoactive after being born SGA and/or preterm, and becomes hyperactive with age, although not indisputable. Similar shifts were observed in extremely preterm infants (gestational age <29 wks) compared to very preterm (gestational age 29-32 wks) and term infants 31 . Our study suggests that these longitudinal patterns in HPA axis activity of preterm infants are augmented by poor intrauterine and early-postnatal growth, although our follow-up might have been too short to demonstrate a subsequent increase in HPA axis activity. We found no differences in salivary cortisol parameters at age 8 y, which included diurnal rhythmicity and CAR, between growth-restricted and non-growth-restricted subjects. At school age, a lower CAR has been described in preterm born children compared to termborn control subjects 32 . Studies regarding the diurnal rhythm linked