Longitudinal hair cortisol in low-income young children: A useful biomarker of behavioral symptom change?

Early childhood is a developmental period characterized by significant plasticity, heterogeneity in behaviors and biological functioning. Yet, cumulative cortisol secretion, as measured by hair cortisol, has not been examined longitudinally in relation to change in behavioral problems in young children. The current study examined cross-sectional and longitudinal associations between hair cortisol and changes in behavior problems in a combined sample (N = 88) of two groups of young children from low-income families: 1) A trauma-exposed sample that participated in Child-Parent Psychotherapy (CPP) (n = 43; Mean Age = 4.31, SD = 1.16; 53% Female; 77% Hispanic), and 2) A community sample of children from families experiencing high stress (n = 45; Mean Age = 3.20, SD = 0.29; 67% Female; 58% Hispanic). Cortisol was assayed from hair collected from children at baseline and, on average, one year later. Mothers completed the Child Behavior Checklist at the same time hair samples were collected. Baseline hair cortisol in children was not associated with maternally-reported child behavioral problems at baseline and did not predict change in behavior problems over time. In contrast, increases in cortisol were associated with greater improvement in child behavior problems (b = -2.98, p < 0.05), controlling for group status and relevant covariates. Subgroup analyses showed that cortisol change across one year significantly differed between the two groups (p = 0.043): on average, community children exhibited a decrease, whereas CPP children demonstrated no change. Hair cortisol concentration was similarly related to improvements in motherreported behavior problems across both CPP and community groups over time. In summary, there were no cross-sectional associations with hair cortisol, whereas increases were associated with improved child well-being. Findings demonstrate an important link between this increasingly common biomarker and child health, but suggest that changes over time may be more informative than cross-sectional associations.