Understanding the Co‐occurrence of Depressive and Anxiety Symptoms Among 2021 Henan Flood Victims Through Panel Network Analysis: A 6‐Month, Three‐Wave Longitudinal Study

Previous studies have mostly used cross-sectional network to analyze the co-occurrence of depression and anxiety, but this method cannot capture the temporal influences between symptoms. This study uses longitudinal data to explore the dynamic structure of the cooccurrence of depressive and anxiety symptoms among flood victims in Henan in 2021. Data were collected at 3-month intervals from July 20, 2021, to January 30, 2022, and the final sample included 279 disaster victims reporting on items measuring anxiety and depression. We employ the generalized vector autoregressive model approach to estimate network models. The contemporaneous network results show that, within the same measurement occasion, all connections between anxiety and depressive symptoms are positive, with the strongest connection observed between "Sleep" and "Appetite". And "Nervous" is the most central symptom, while "Irritable" and "Motor" are the top two strongest bridge symptoms. The temporal network results indicate that depressive symptoms are more temporally causal and predictive, while the temporal associations between anxiety symptoms are rare. And depressive symptoms were found to predict anxiety symptoms. Additionally, "Suicide" and "Concentration" showed significant positive autocorrelations, indicating a self-sustaining capacity. "Anhedonia" has the highest in-strength centrality (incoming influence from prior time point), demonstrating the most downstream effect. In contrast, "Concentration" has an out-strength centrality (outgoing influence to the next time point) far exceeding that of other symptoms, suggesting it has the most outward influence. These patterns suggest testable, symptom-focused priorities for post-disaster care: early attention-focused strategies to curb downstream spread from concentration problems; behavioral activation for anhedonia; proactive safety planning and scheduled follow-ups for suicidal ideation; brief transdiagnostic modules targeting irritability and motor activation to reduce comorbidity; and integrated routines addressing the sleep-appetite dyad. Implications are observational and intended to inform hypothesis-driven trials and service planning.