Bibian van der Voorn

78 CHAPTER 6 Committee of the VUMC (protocol METc VUMC 2013/80 and 15/001), and written informed consent was obtained from all participants. Study 1. We recruited 10 mothers who delivered at term and 10 who delivered very prematurely within the first week postpartum. Mothers who breastfed their very preterm infant (GA <32 wks) or full-term infant (GA ≥ 37 wks) were eligible for inclusion. Exclusion criteria were multiple pregnancy, breast surgery, HELLP syndrome, pre-existing or gestational diabetes, autoimmune disease, life expectancy of the neonate of <72h, and/or major congenital anomalies of the neonate. Study 2. We recruited 10 mothers who delivered at term within the first month postpartum. Healthy mothers who breastfed their full-term infants were eligible for inclusion. To minimize confounding influences such as lifestyle and psychological factors, we excludedmotherswith the following: a prepregnancy BMI (in kg/m 2 ) <18.5 or >30 14 ; excess pregnancy weight gain of >16 kg for women with a prepregnancy BMI of 18.5 to 25 or >11 kg for women with a prepregnancy BMI of 25 to 30, according to 2009 Institute of Medicine/NRC guidelines 15 ; medication use; alcohol consumption of >7 IU/wk (1 IU of beer = ∼12 ounces, 1 IU of wine = ∼4 ounces) 16 ; fever > 38.5 ⁰C 17 ; and/or an Edinburgh Postnatal Depression Scale score >10 (of 30) in the third week postpartum 18-20 . SAMPLE COLLECTION Study 1. Milk samples were obtained weekly during the first month postpartum, at the same time of day, to avoid bias related to the timing of collection. Weekly collections within the first month postpartum were chosen, because previous studies showed that macronutrient milk composition differed between term and very preterm milk, especially in the early stages of lactation 21,22 . One milliliter of foremilk was collected with a breast pump just before every feeding occasion, although cortisol concentrations were found to be equal in fore- and hindmilk, irrespective of lactation stage 23 . Samples were stored in polypropylene vials at –20 ⁰C until analysis. Study 2. Paired foremilk and saliva samples were collected in the fourth week postpartum, over a 24-h period. Samples were collected at each feeding occasion (7 to 8 times/24h). Mothers breastfed their children on demand and were therefore asked to write down the exact time of sample collection. In addition, we asked mothers to collect 2 additional samples early in the morning (between 0600 and 1200) to obtain a more accurate assessment of peak cortisol and cortisone (see Figure 1 ). At each sampling occasion, 1 mL foremilk was collected according to the previously mentioned procedure. Simultaneously, with every milk sample, a saliva sample was