The random dyadic interdependence model: Modeling variability in physiological covariation within dyads

The present study demonstrates a novel modeling strategy for capturing physiological linkage in dyads using techniques newly available in the Dynamic Structural Equation Modeling framework. Leveraging repeated physiological measures data from a sample of older parent-adult child dyads (204 individuals, N = 102 dyads) coping with early-stage cognitive impairment, we expected a wide range of physiological interdependence. This study demonstrates the substantial heterogeneity in dyadic interdependence in several physiological measures (systolic blood pressure, diastolic blood pressure, mean arterial pressure, heart rate, and respiration rate). Results provided evidence of a variable interdependence in all physiological outcomes, with a both negative and positive dyadic interdependence patterns estimated across dyads. Results provided preliminary support for the use of variable interdependence as a dependent variable. Family cohesion and open expression were found to be associated with more strongly positive interdependences in blood pressure outcomes, but not heart or respiration rates. Other predictors were not significantly associated with interdependence. One benefit of the random covariance modeling technique is its ability to simultaneously estimate a range of negative to positive interdependences in physiological data, both of which were represented in our sample of older parent-adult child dyads.