Bibian van der Voorn

172 CHAPTER 11 RESULTS A short overview of all articles is presented in Tables 1 through 5 . For a more in- depth summary of the articles, see Online Supplementary File 3 . Data on 14,591 subjects were included in this review, with an age range of 31 hours to 18 years. DIURNAL RHYTHM Twenty-nine studies (with the data of 8,971 subjects) described diurnal rhythmicity and/or decline of cortisol throughout the day in children, of which fifteen studies reported no significant sex differences. 11-25 Fourteen studies reported significant sex differences, of which twelve reported higher cortisol levels and/or a steeper decline over the day in girls. Both Adam et al. (2010) 26 (n=230, age: 17.04±0.36 years) and Williams et al. (2013) 27 (n=27, age: 9.13±1.41 years) reported a steeper diurnal cortisol curve in girls. Morin-Major et al. (2016) 28 (n=88, age: 14.5±1.8 years) found a higher area under the curve as measured from the ground (AUCg) in girls. Martikainen et al. (2013) 29 (n=252, age: 8.1±0.3 years) reported a higher cortisol level at awakening in girls, while there was no difference between sexes at nadir, suggesting a steeper cortisol decline over the day in girls compared to boys. This was also found by Rosmalen et al. (2005) 30 (n=1768, age: 11.08±0.55 years), who found this to be already present pre-pubertally, while age and pubertal status were not associated with diurnal rhythm. Fransson et al. (2014) 31 (n=157, age: 14-16 years) found a higher cortisol level at awakening and a steeper diurnal decline in girls. Kelly et al. (2008) 32 (n=2,995, age: 15.4±0.3 years) found a greater decrease in cortisol concentration in girls as compared to boys between +/- 9am and 9:30am. Ruttle et al. (2013) 33 (n=346, age: 11, 13 and 15 years) and Shirtcliff et al. (2012) 34 (n=357, age: 9, 11, 13 and 15 years) examined the same cohort. Ruttle et al. (2013) found a significantly steeper diurnal decline in girls aged 11 and 13 years. At age 15, gender differences in cortisol slope had disappeared, although girls had higher cortisol levels throughout the day. Shirtcliff et al. (2012) found similar differences, with higher cortisol and steeper slopes, as well as more curvature, in girls. Moreover, the circadian rhythm became flatter with advancing puberty, particularly among girls. Vaillancourt et al. 2008) 35 (n=154, age: 147±9.1 months) examined morning and evening cortisol levels on Monday, Thursday and Saturday. They only found a higher cortisol concentration in girls on Saturday morning. Moreover, after modeling the circadian pattern, they found that girls consistently had higher cortisol levels than boys throughout the day. Bae et al. (2015) 36 (n=138, age: 10.7±1.7 years) found higher cortisol levels in girls at awakening and 30 minutes after awakening, as well as a higher total daily output.