Intercorrelations between serum-, salivary- and hair-cortisol and child-reported estimates of stress in elementary school girls

• Main Authors: Barbara Vanaelst, Inge Huybrechts, Karin Bammann, Nathalie Michels, Tineke de Vriendt, Krishna Vyncke, Isabelle Sioen, Licia Iacoviello, Kathrin Gunther, Denes Molnar, Lauren Lissner, Noellie Rivet, Jean-Sebastien Raul, Stefaan de Henauw
• Format: Article
• Language: English
• Published: Taylor & Francis Group 2012-09-01
• Series: European Journal of Psychotraumatology
• Subjects: cortisol; child; stress; questionnaire; biomatrices

To evaluate the impact of stress on children's well-being, it is important to have valid and reliable stress assessment methods. Nevertheless, selection of an appropriate method for a particular research question may not be straightforward, as there is currently no consensus on a reference method to measure stress in children. This paper examined to what extent childhood stress can be estimated accurately by commonly applied stress measures.Two hundred and seventy-two girls between 5 and 11 years old participated in this study as part of the ChiBS project. Child-reported estimates of stress were collected through the Coddington Life Events Scale (CLES). Serum, saliva and hair samples were collected for cortisol analyses. The intercorrelations of cortisol in the different biological samples were investigated by Spearman rank correlations and Bland-Altman plots. Next, CLES-scores, salivary and hair cortisol concentrations were compared triangularly with the true, but unknown childhood stress using the Triads method, based on pair-wise Spearman's correlation coefficients and the calculation of validity coefficients.Serum cortisol (free and total) was positively correlated with salivary morning and AUC cortisol. Hair cortisol correlated with salivary morning and AUC cortisol, but not with serum cortisol. In relation to recent childhood stress (0–3 months ago), the highest validity coefficients were observed for salivary cortisol measurements, while for periods more distant in the past hair cortisol measurements displayed the highest validity coefficients.This paper investigated the relationship between cortisol measurements in different biological samples, showing a lack of association and disagreement between measures of single-point, short-term cortisol versus long(er)-term cortisol. In addition, this paper examined to what extent childhood stress can be accurately estimated by stressor questionnaires and biological markers in girls. Salivary cortisol was shown to most accurately indicate true childhood stress for short periods in the past (i.e. last three months), whereas hair cortisol may be preferred above salivary measurements for periods more distant and thus for chronic stress assessment.