Abstract
OBJECTIVES: This study aims to construct a depression symptom network in elderly hypertensive patients, identify central and bridging symptoms, and explore the association between network structure and modifiable risk factors. METHODS: This study adopts a retrospective research design, reviewing the medical records and survey data of 562 elderly hypertensive patients from a tertiary comprehensive hospital from September 2022 to May 2023. The data was retrospectively collected from patient health records including a general demographic questionnaire, Insomnia Severity Index-7(ISI-7), 9-item Patient Health Questionnaire (PHQ-9), Generalized Anxiety Disorder-7 (GAD-7), and Connor Davidson Resilience Scale-25 (CD-RISC-25). Calculate centrality indices (intensity, betweenness centrality, and intimacy) to identify core symptoms. A comprehensive network model integrating GAD-7, ISI-7, CD-RISC-25, and demographic variables was constructed. RESULTS: A total of 562 patients were enrolled in the study. The average score of PHQ-9 is (10.69 ± 3.42) points. Network analysis shows that anhedonia (PHQ1) exhibits the highest intensity centrality. The strongest partial correlation was observed between Sleep problems(PHQ3) and PHQ1 (weight=0.40), fatigue (PHQ4) and depressed mood (PHQ2) (weight=0.29), and PHQ4 and PHQ1 (weight=0.29). There are two different symptom clusters: somatic affective clusters (PHQ1, PHQ3, PHQ4) and cognitive vegetative clusters (appetite problems(PHQ5), feeling of worthlessness (PHQ6), concentration problems (PHQ7)). Suicide ideation (PHQ9) exhibits the lowest centrality. The comprehensive network model indicates a strong positive correlation between depression and anxiety (PHQ-GAD), depression and insomnia (PHQ-ISI), and anxiety and insomnia (GAD-ISI). The dimensions of psychological resilience, including self reinforcement, resilience, and optimism, are negatively correlated with PHQ scores (all P<0.001), while GAD-7 scores are positively correlated. There are edge connections between exercise (EX) and ISI, disease course (DU), and gender (GD). Drink (DR) is positively correlated with GD, while degree of education (DOE) is connected within demographic clusters and has an edge with GD. CONCLUSIONS: Network analysis revealed that in the depressive network of patients with hypertension, anhedonia is the most central symptom, indicating that it may become a primary intervention target. The comprehensive network uncovered significant interconnections among depression, anxiety, and insomnia. Furthermore, the resilience dimension negatively correlates with depressive symptoms, while there are edge connections between exercise and both insomnia and demographic factors, highlighting modifiable protective factors.