Comparing Associations Between Acute and Long-Term Cortisol Measurements and Anxiety in Preschool-Aged Children

• In Event: Poster Session 03
  In Poster Session: PS 03 Topic - Developmental Psychopathology

Abstract

Cortisol, the end product of the hypothalamic-pituitary-adrenocortical (HPA) axis, may be associated with stress reactivity and risk for anxiety in children. Given individual variation in the interactions between biology and the environment (Bates, et al., 2017; Shonkoff et al., 2012), reliable, objective assessment of children’s stress response is needed. Identifying associations between methods of cortisol measurement and anxiety can clarify methodological discrepancies and further understanding of the relation between stress reactivity and anxiety. This study compared methods of assessing cortisol by examining acute (salivary) and long-term (hair) HPA-axis functioning in association with risk for anxiety in 3-5-year-old children (N=141;44.4%Female;56.74%Caucasian/White,31.91%Hispanic,5.67%multi-ethic/other,4.96%Asian,<1%African-American/Black). Children’s fear/inhibition (M=31.52,SD=11.11,Min=1.00,Max=64.00) was observed and coded during a structured laboratory stress task with a “stranger”. Four saliva samples timed around the laboratory stress task measured children’s acute cortisol reactivity; chronic cortisol concentrations were assessed through children’s hair samples (indexing hair cortisol concentrations [HCC]). Salivary cortisol was analyzed by computing both the area under the curve (AUC)and slope for each child. AUC represents the total amount of cortisol released from baseline to 40 minutes after the laboratory task while slope represents the change in cortisol across all 4 time points. Parents also reported about children’s separation (M=2.02,SD=0.80,Min=1.00,Max=4.40,α=.77) and social (M=2.02,SD=0.74,Min=1.00,Max=4.43,α=.83) anxiety using the Preschool Anxiety Scale (PAS). A linear regression demonstrated that children showing higher levels of fear/inhibition during a laboratory stress task showed an elevated positive change (slope) in salivary cortisol in response to the task (F(1,139)=7.38,p=.007,R2=.05). Specifically, higher levels of fear/inhibition during the laboratory task was associated with a more positive cortisol slope across the four timepoints (b=.28,SE=.10,t(139)=2.72,p=.007,95%CI=[.08,.49]). A mean split of laboratory fear/inhibition was computed to compare the salivary cortisol slopes of children with high and low fear. The slopes of the high and low fear groups were significantly different (t(139)=2.51,p=.013; See Figure 1): those with more fear in the laboratory task showed a steeper, positive slope across time. Those who displayed lower fear in the laboratory showed little change from baseline through the third sample and a decrease in cortisol in the last sample. Fear/inhibition was not associated with AUC (F(1,138)=1.28,p=.26). Parent-reported anxiety was not significantly related to AUC(r=-.01,p=.891), slope (r=-.02,p=.849), or to HCC(r=-.05,p =.566). Anxiety during the laboratory task was not significantly related to AUC (r=-.10,p=.260), or HCC (r=-.08,p= .373), but was significantly positively related to increasing cortisol slope (r=.23,p=.007). Thus, when comparing both cortisol assessments, this study suggests that change in salivary cortisol in response to an acute stressor may best predict behavior relevant to anxiety. If results are replicated in future research, these data could distinguish objective biological assessments that can aide in identifying children at risk for anxiety. Given the other nonsignificant results, replication is needed to confirm the lack of associations or to determine if there are associations between anxiety and cortisol not identified in the present study.

Authors

• Gillian Acedo, California State university San Marcos
  Presenting Author

• Natalie L. Gemmell, California State University San Marcos
  Non-Presenting Author

• Kimberly D’Anna-Hernandez, California State University San Marcos
  Non-Presenting Author

• Sara Bufferd, University of Louisville
  Non-Presenting Author

• Lea Rose Dougherty, University of Maryland, College Park
  Non-Presenting Author