Temporal bidirectional dynamics of momentary stress and affect: Dynamic structural equation modelling of ecological momentary assessments

The process model of emotion regulation highlights affect's ebb and flow in daily life in response to external events or internal processes, such as stress. Collectively daily stress and affect can shape daily experiences and influence long-term health. Understanding the dynamics of the stress-affect relationship requires examining intensity (average level), inertia (autoregression), and variability (residual variances), yet few studies simultaneously consider both arousal (high vs low, denoted as H vs L) and valence (positive vs negative, denoted as PA vs NA) dimensions of affect. Participants (N = 424; 72.4 % women, M_age = 22.79 ± 5.35) came from three longitudinal ecological momentary assessment studies (completed between 2017 and 2020) across 7-15 days. Dynamic structural equation modelling tested the dynamics between stress and affect across valence and arousal, and their bidirectional relationships. Results showed that individuals with higher stress intensity experienced higher HNA and LNA and, to a lesser extent, lower HPA and LPA intensity. Individuals with higher stress intensity also had higher overall stress variability and less stress in response to PA. Higher average HNA was associated with higher stress variability, whereas higher average LNA was associated with higher affect variability. Additionally, higher HNA inertia was associated with more stress in response to HNA and less HNA reactivity to stress. Our findings support the bidirectional and dynamic nature of stress-affect relationships and the value of incorporating nuanced aspects of stress and affect into future research across populations and clinical intervention frameworks.