Bibian van der Voorn

128 CHAPTER 9 The cortisol awakening response (CAR) was calculated as the difference between levels at awakening and 15 min after awakening. We calculated the AUC increase (AUC i ; with reference to the lowest individual cortisol level) of all samples according to the formula by Pruessner 22 , and a time-weighted AUC i (divided by the total time of collection, AUC i /h). Analyses of salivary cortisol parameters in association with birth weight SDS, SGA-status and growth from birth to 6 mo. corrected age were performed with univariable linear regression analysis. Statistical significance was defined as a P-value of <0.05. Statistical analyses were performed with IBM SPSS Statistics version 22. RESULTS One hundred and fifty two subjects were included in STEP-1. Over time, 13 children were lost to follow-up, 21 children were excluded, and 39 children refused to participate in STEP-2, resulting in 79 subjects that could be included at age 8 y in STEP-2 ( Figure 1 ). No differences were found in baseline characteristics of STEP-2 participants vs. non- participants and excluded subjects (P > 0.3), except for parental ethnicity which was more often “non-Caucasian” in the excluded group compared to participants and non-participants ( Table 1 ). AGA and SGA subjects differed in gestational age, gender distribution, preeclampsia/HELLP syndrome and exposure to antenatal glucocorticoid treatment ( Table 2 ). We considered these variables as possible confounders and therefore tested their influence on the longitudinal analyses. TABLE 2. Perinatal and study characteristics that could potentially confound the research question AGA SGA P-value a n 109 35 PIH/Preeclampsia/HELLP 36 (33) 21 (60) <0.01 Antenatal glucocorticoid treatment 57 (52) 24 (69) <0.01 Male 49 (45) 25 (71) <0.01 Gestational age (wks) 30 [29 to 31] 31 [30 to 32] <0.01 STEP-1 human milk group 37 (34) 11 (31) NS NTISS score 21.4 ± 7.8 21.6 ± 6.9 NS Caucasian ethnicity 78 (72) 24 (69) NS Time of collection serum samples (24h) b 13:27 ± 2:05 13:29 ± 2:09 NS Data are expressed as mean ± SD, median [IQR], or n (%). Abbreviations: AGA, appropriate for gestational age (birth weight and birth length >–2SDS); NS, non-significant; NTISS, neonatal therapeutic intervention scoring system; PIH, pregnancy induced hypertension; SGA, small for gestational age (birth weight and/or birth length ≤–2SDS) a Groups were compared using independent t-test, Fisher’s exact test, or Mann- Whitney U test as appropriate b Mean time of collection of all sampling moments: term age, 3mo. or 6 mo. corrected age and age 8 y.