Temporal Structures of Physiological Influence During Couples' Support Discussions: An Application of Continuous-time...

We use continuous-time structural equation modeling (ctSEM) to model physiological influence in couples during a support conversation, capturing the direction and strength of linkage separately for the support provider and recipient. ctSEM accounts for the fact that physiological systems react at different speeds, helping to identify how quickly influence occurs and when it peaks for each system.

80 Couples took part in a 7-minute support discussion where a support recipient discussed a personal goal with their partner (support provider). Physiological data were recorded using 7 electrocardiogram and 2 electrodermal activity sensors and couples filled out self-report questionnaires pre and post the discussion. ctSEM models were used to analyse physiological influence.

Partners showed significant physiological influence in inter-beat interval (IBI) and skin conductance level (SCL), but not in pre-ejection period. IBI influence peaked at a 2-3 s lag (only from support recipient to provider). In contrast, SCL showed mutual influence peaking at 15-16 s.

Our findings show that physiological influence unfolds at different speeds, with IBI co-regulation occurring faster than skin conductance-consistent with known autonomic response patterns. ctSEM proved useful in capturing the timing and direction of these effects, offering a more precise, flexible approach than discrete-time models by modeling the continuous nature of physiological processes. Its ability to identify peak time-lags can help guide future study designs and sampling strategies.