A Cognitive Network Approach to Exploring Semantic and Emotional Patterns in Anxiety and Depression Narratives

Background

Anxiety and depression are among the most prevalent mental disorders, frequently co-occurring and presenting challenges in diagnosis and treatment. Understanding their underlying cognitive and emotional organization is essential for developing accurate and effective interventions.

Methods

This study applied a cognitive network science approach to explore semantic and emotional patterns in narratives written by individuals reporting symptoms of anxiety, depression, comorbid symptoms, and no symptoms (asymptomatic). Using co-occurrence and correlation-based estimation methods, networks were constructed with and without stopwords to examine their structural differences. Key analyses included structural properties, emotional balance, and semantic frames of central words to identify group-specific patterns.

Results

Distinct cognitive and emotional profiles emerged. Anxiety and depression groups were characterized by denser but fragmented networks, dominated by negative emotional content. Narratives from depression group additionally showed a prominent presence of somatic-related concepts and higher emotional balance, reflecting the co-occurrence of positive and negative concepts within the same semantic structure. The comorbid group combined relatively high density with the lowest emotional balance, indicating greater emotional tension and ambivalent semantic organization. In contrast, the asymptomatic group demonstrated sparser yet more coherently organized and efficient networks, alongside the highest emotional balance, suggesting a more stable and regulated cognitive–emotional profile.

Conclusions

These findings demonstrate the utility of cognitive network analysis for capturing differences in semantic and emotional organization associated with anxiety and depressive symptom profiles. By integrating network topology, emotional balance, and semantic framing, this approach offers a nuanced framework for investigating cognitive and emotional processes beyond isolated word use, with potential implications for personalized therapeutic interventions and the monitoring of treatment outcomes.

‍