In a Matter of Seconds: Comparative Modeling Of 8-Month-Old RSA During A Social Fear Task

• In Event: A Multimethod Symposium on Emotion Regulation from Infancy to Middle Childhood

Abstract

Developmental research largely relies on examining the rate, shape, and timing of change to provide empirical foundations for theories explaining patterns and mechanisms of change over time (Adolph et al., 2008). However, comparative models depicting these three components of change are widely understudied, particularly in developmental psychophysiology. Dynamic shifts in respiratory sinus arrhythmia (RSA) represent adaptive coping efforts to affective demands and behavioral response readiness (Porges, 2007). Individual differences in temperament predict toddlers’ quadratic change in RSA during a social fear task (Brooker & Buss, 2010). One regulatory aspect of temperament is soothability, which has been positively associated with infants’ physiological regulation (Huffman et al., 1998). However, we do not know if children exhibit different patterns of RSA during a social fear task when examining dynamic RSA at different change rates, and if individual differences in soothability are associated with these RSA changes.
This study assessed RSA using three different rates of change: static average score, 30-second epochs, and second-by-second epochs in a sample of 210 8-month-olds. RSA was measured across baseline and the Stranger Approach task (Goldsmith & Rothbart, 1996), and infant soothability was reported via the IBQ-R (Gartstein & Rothbart, 2003).
First, we calculated an RSA change score by subtracting the average task-based RSA from baseline RSA. A simple regression model indicated that RSA suppression was not significantly associated with soothability (SE = 0.09, t = .41, p =.68).
Second, we fit a multiphase growth model to examine RSA changes across baseline and three 30-second epochs, with soothability as a predictor. The model fit the data well, AIC = 1574.92, BIC = 1631.80, LL = -774.46. Across participants, there was an overall increase in RSA from baseline to the start of Stranger Approach, and a decrease from start to finish of Stranger Approach. Soothability did not significantly predict baseline RSA or increase in RSA from baseline to task, but lower levels of soothability were significantly associated with less RSA suppression throughout the task (b=-0.15, SE = 0.05, t=-2.55, p <.05).
Lastly, we fit a linear growth model to examine second-by-second changes in RSA, with soothability as a predictor. RSA was measured in second-by-second intervals using RSASeconds (Gates et al., 2015; Hansson & Jönsson, 2006). The model fit the data well, AIC = 8208.42, BIC = 8265.65, LL = - 4096.21. Across participants, there was a significant overall decrease in RSA throughout Stranger Approach (b=-0.01, SE = 0.00, t=-3.88, p <.001), but soothability was not significantly associated with RSA slope throughout the task (b=-0.003, SE = 0.002, t=-1.02, p =.31).
Individual differences in infant soothability were only associated with RSA change at the 30-second level. This finding suggests that infants who are less soothable via caregiver-directed regulatory efforts may show diminished psychophysiological regulation when experiencing fear. Additionally, while broader findings require replication, they may suggest a “sweet spot” for the rate at which we should examine RSA to capture individual differences at play that may not be otherwise detected in traditional static regression models or on a second-by-second basis.

Authors

• Elizabeth Youatt, Pennsylvania State University, University Park
  Presenting Author

• Anna M Zhou, University of Utah
  Non-Presenting Author

• Vanessa Lobue, Rutgers University
  Non-Presenting Author

• Kristin A. Buss, Pennsylvania State University, University Park
  Non-Presenting Author

• Koraly Elisa Perez-Edgar, Pennsylvania State University, University Park
  Non-Presenting Author