Classroom, teacher, and peer interactions: Associations with preschool children’s salivary cortisol

• In Event: Unpacking the nuances of teacher-child interactions: Direct and indirect effects on child outcomes

Abstract

Young children’s production of salivary cortisol is affected by many factors, including relationships in early childhood (Gunnar & Quevedo, 2007); ideally, children’s cortisol levels show a decline across the day, following 30 minutes after wake. Some evidence indicates that positive, warm teacher-child interactions predict a decline in cortisol across the day or morning for preschool children (e.g., Hatfield, 2019), yet other work does not (e.g., Gunnar et al., 2010). A global classroom-level measure of interaction quality may not capture the nuances of individual children’s experiences, which may more directly influence a child’s HPA axis and thus production of cortisol. Additionally, certain child characteristics such as time in care (Roisman et al., 2009) and exposure to adverse childhood experiences (ACEs; Bevans et al., 2008) may also affect children’s cortisol. This paper examines classroom-level quality, children’s engagement, ACEs, and time in care as predictors of cortisol among 66 children within 20 classrooms who participated in a trauma-informed professional development program, utilizing baseline data.

Children displayed a range of characteristics (Table 1), and classrooms represented a variety of care settings (30% home, 70% center or Head Start). Parents reported on total number of ACEs using Center for Youth Wellness ACE-Questionnaire (Bucci et al., 2015); a categorical score (0, 1, 2, and > 3 events) was used in analyses. Parents also reported on child’s hours in care each week. Classroom observations using the CLASS (Pianta et al., 2008) and inCLASS (Downer et al., 2011) were conducted at baseline over multiple days by trained observers. Domain scores for CLASS Emotional Support (ES) and Classroom Organization (CO) were calculated. For the inCLASS four domains were calculated: Positive engagement with teachers, positive engagement with peers, positive engagement with tasks, and negative engagement. Salivary cortisol was collected twice over the morning on one of those observations days; assays were conducted in duplicate. Rate of diurnal change (RDC) was calculated to reflect change in cortisol across the morning at preschool. Descriptive statistics for these variables are presented in Table 2.

Six models were executed with three covariates (i.e., income, hours/week in care, and ACEs), accounting for children nested within classrooms, to predict children’s RDC from the two CLASS domains and the four inCLASS domains. First, lower hours/week in care predicted rise RDC for children in most of the models, and ACEs was not a significant predictor. Results indicate that classroom-level interactions (ES and CO domains) were not predictive of children’s RDC. Yet, two inCLASS domains predicted children’s RDC. First, children who demonstrated lower observed positive peer engagement showed an increase in RDC across the morning at preschool (B = -.19, SE = .09, p =.03). Second, children with a lower negative engagement also experienced a rise in cortisol across the morning (B = -.40, SE = .12, p =.002). Together, these indicate that children’s engagement is related to the activity in the HPA axis for preschool children. Future models plan to unpack these associations and consider role of additional child, classroom, and teacher characteristics.

Authors

• Bridget E Hatfield, Oregon State University - Corvallis
  Presenting Author

• Shannon Tierney Lipscomb, Oregon State University
  Non-Presenting Author

• Aubrey Sills, Oregon State University-Cascades
  Non-Presenting Author