Dynamic interactions among anxiety, emotional stability, and mindfulness: a cross-lagged panel network analysis

Background: Anxiety is one of the most prevalent global mental health disorders, with its incidence having risen substantially since the COVID-19 pandemic and remaining elevated in the post-pandemic era. According to the process model of emotion regulation, mindfulness may serve as an early-stage cognitive change strategy that enhances emotional stability and ultimately reduces anxiety. While previous research has demonstrated robust interconnections among anxiety, emotional stability, and mindfulness, the majority of studies have relied on cross-sectional designs and aggregate scores, neglecting their longitudinal, item-level dynamic relationships.

Method: This study collected data from 547 college students ( M age ± SD = 20.11 ± 1.20) across three time points and utilized a cross-lagged panel network (CLPN) approach to analyze item-level temporal dynamics. The goal was to build network models of anxiety, emotional stability, and mindfulness, and to investigate how these constructs were temporally interrelated at the item level in an exploratory manner.

Result: Mindfulness items had higher out-Expected influence (OEI), whereas anxiety symptoms had stronger in-Expected Influence (IEI), with anxiety and emotional stability having the strongest bridge effects. Specifically, mindfulness components “Thought Acceptance” (M7) and “Nonjudgmental Awareness” (M8) exerted significant effect on anxiety and emotional stability. Anxiety symptoms were highly responsive to other symptoms, which undermined emotional stability. The network analysis revealed complex temporal and predictive associations among anxiety, emotional stability, and mindfulness.

Conclusion: This study is the first investigation to reveal the longitudinal dynamic connections between anxiety, emotional stability, and mindfulness at the item level, consequently provided longitudinal support consistent with the emotion regulation process model.