Research report: Identifying target symptoms for mixed anxiety and depressive disorder in treatment-seeking patients: network analysis and computer-simulated intervention

Researchers analyzing the symptom networks of 1,455 treatment-seeking patients with mixed anxiety and depressive disorder have identified which individual symptoms may be the most promising targets for intervention. The cross-sectional study, published in BMC Psychiatry, used network analysis and computer-simulated interventions to map how anxiety and depression symptoms reinforce one another. The strongest connection linking the anxiety and depression communities ran between ANX1 (nervousness or shakiness inside) and DEP9 (worrying too much about things), suggesting that these two symptoms form a critical bridge between the two symptom clusters.

In the simulated intervention analysis, targeting ANX7 (spells of terror or panic) produced the largest alleviation effect, with a change rate of 11.13%, while DEP5 (feeling trapped or caught) showed the largest aggravation effect at 5.39%, meaning that increases in this symptom were associated with the greatest simulated worsening across the network. The study included 485 male patients (33.33% of the sample) and had a mean age of 30.59 years (standard deviation 11.79). The researchers calculated expected influence (EI) and bridge expected influence (bridge EI) indices to measure which symptoms were most central within their own community and which served as bridges between the anxiety and depression communities.

The nonparametric bootstrapping test indicated that ANX4 (feeling fearful), ANX9 (the feeling that something bad is going to happen to you), and DEP5 (feeling trapped or caught) were significantly more influential than more than 50% of the other symptoms in the co-occurrence network. The case-dropping correlation stability coefficients were 0.67 for edge weights, 0.52 for bridge EI, and 0.44 for EI, indicating adequate stability of the network estimates. The computer-simulated intervention analysis went beyond static network indices by modeling what happens when a single symptom is artificially increased or decreased.

The results showed that EI was significantly and positively associated with simulated alleviation rates, while the other zero-order associations involving EI and bridge EI were not significant. This means that symptoms with high expected influence were the ones whose simulated reduction produced the greatest improvement across the network. The researchers also found that the simulated intervention effects were closely tied to the network's threshold parameters, meaning that the outcomes depended on the activation thresholds built into the model rather than on the network structure alone.

However, the findings come with important limitations. The study is cross-sectional, so it cannot establish causal relationships between symptoms, and the intervention analysis is computer-simulated rather than a clinical trial, so the reported change rates reflect model estimates rather than observed treatment outcomes. The study provides a data-driven framework for selecting intervention targets in comorbid anxiety and depression, but the simulated nature of the evidence means that clinical application remains a hypothesis rather than a demonstrated result.

Subject of Research: Identifying target symptoms for mixed anxiety and depressive disorder in treatment-seeking patients: network analysis and computer-simulated intervention

Article Title: Identifying target symptoms for mixed anxiety and depressive disorder in treatment-seeking patients: network analysis and computer-simulated intervention

Article References: Tao, Y., Wu, Y., Niu, H., Ma, Z., Li, Y., Zhang, X., Liu, X., Liu, G., & Xu, H. (2026). Identifying target symptoms for mixed anxiety and depressive disorder in treatment-seeking patients: network analysis and computer-simulated intervention. BMC Psychiatry. https://doi.org/10.1186/s12888-026-08697-4

Image Credits: AI Generated

DOI: 10.1186/s12888-026-08697-4

Keywords: Identifying, target, symptoms, mixed, anxiety, depressive, disorder, treatment-seeking, patients, network, analysis, computer-simulated