摘要
Rare diseases often entail psychological distress and functional impairment, yet their interconnections with Quality of Life (QoL) remain poorly understood. Most studies have examined these domains separately, missing the systemic nature of mind-body comorbidity. Network analysis offers a novel framework to identify symptom-level central nodes and bridge pathways for precision intervention. This study applied network analysis to (1) map the interrelationship structure between psychological symptoms and QoL, and (2) identify central and bridge symptoms for effective targeted intervention.
A cross-sectional survey was conducted among hospitalized patients with rare diseases or undiagnosed conditions in a tertiary hospital between February 2025 and August 2026. Psychological symptoms and QoL were assessed using the Symptom Checklist-90 (SCL-90) and the MOS 36-Item Short-Form Health Survey (SF-36), respectively. Network analysis was performed using R software. Following redundancy testing via the goldbricker function, seven SCL-90 factors and seven SF-36 dimensions were retained as 14 network nodes. A regularized partial correlation network was estimated using the Gaussian Graphical Model via the EBICglasso procedure. Node centrality was evaluated using strength; bridge strength was computed to identify nodes connecting the psychological symptoms and QoL domains. Predictability was calculated via the mgm model, and network stability was assessed using case-dropping bootstrap (1,000 resamples), with CS-C > 0.5 indicating excellent stability.
Among the 204 eligible participants, the positive detection rates of the seven SCL-90 factors ranged from 6.9% to 19.6%, with "Obsession-Compulsion" (19.6%), "Somatization" (14.7%), and "Depression" (14.2%) being the most prevalent. The network analysis results are as follows: (1) Network structure and node centrality. The network comprised 56 non-zero edges. "Depression" (Strength=1.249) exhibited the highest centrality, followed by "Interpersonal Sensitivity" (Strength=1.229) and "Social Functioning" (Strength=1.194), indicating that these domains exerted the strongest influence within the overall network. (2) Bridge centrality. "Somatization" as the strongest bridge node connecting the psychological and QoL domains (Bridge Strength = 0.702), substantially surpassing all other nodes, followed by "Bodily Pain" (Bridge Strength=0.397) and "Depression" (Bridge Strength=0.381). This pattern indicates that somatic complaints serve as the primary conduit through which psychological distress translates into functional impairment in this population. (3) Edge strength and cross-domain connections. The top three strongest edges were "Physical Functioning—Social Functioning" (r=0.362), "Interpersonal Sensitivity—Paranoid Ideation" (r=0.333), and "Interpersonal Sensitivity—Phobic Anxiety" (r=0.310). Notably, the strongest cross-domain edges were "Somatization—Bodily Pain" (r=-0.335) and "Depression—Vitality" (r=-0.226), confirming that greater psychological symptom severity was associated with poorer QoL. (4) Network stability and predictability. Bootstrap analysis demonstrated excellent stability (CS-C=0.672). The mgm model yielded a mean R² of 0.255, suggesting that neighboring nodes explained a moderate proportion of variance in each node.
The psychological symptom–QoL network exhibits a tight mind-body coupling, wherein "Somatization" in the psychological domain serves as the critical bridge linking emotional distress to QoL decline. "Depression" and "Social Functioning" maintain this system as core hubs. Clinically, somatic complaints should be recognized as intervention entry points, social functioning prioritized alongside medical treatment, and routine assessment of these domains integrated into inpatient care. An integrated approach addressing both psychological and physical health offers the most promising pathway to restoring meaningful QoL.
