Differential Anxiety–Depression–CRP Network Structures Across Insomnia Severity Levels: Evidence From UK Biobank

Abstract

Background: This study investigated the relationships between anxiety, depression symptoms, and C-reactive protein (CRP) across insomnia severity levels using network analysis and examined the structural differences within these networks.

Methods: Gaussian graphical model network analysis with Least Absolute Shrinkage and Selection Operator (LASSO) regularization was conducted on UK Biobank data (N = 143,027). Depression and anxiety symptoms were assessed using the 9-item Patient Health Questionnaire (PHQ-9) and 7-item Generalized Anxiety Disorder Scale (GAD-7), respectively. CRP was quantified using immunoturbidimetric-high-sensitivity analysis. Participants were categorized by insomnia frequency (never/rarely, sometimes, and usually). The strength symptoms and expected influence identified core symptoms, while bridge expected influence (bridge EI) determined bridge symptoms. Network comparison tests (NCTs) were performed pairwise across the three groups to assess differences in global strength and edge weights.

Results: Across all networks, “Depressed mood” demonstrated the highest strength centrality, while “Irritability” exhibited the highest bridge EI. “Depressed mood” had the highest expected influence centrality in the never/rarely insomnia group and “Uncontrollable worry” in other groups. NCTs revealed significant differences in global strength (S = 0.178, p < 0.01) and edge weights (M = 0.062, p < 0.01) between the never/rarely and usually insomnia groups, with stronger connections between depressive symptoms (energy/appetite) and CRP in the usually insomnia group (p < 0.001).

Conclusions: The central roles of depressed mood, uncontrollable worry, and irritability in the anxiety–depression–CRP network across all insomnia severity groups suggest that these symptoms represent potential targets for future intervention research. Notably, network structure differed across insomnia severity; the strengthened associations between depressive symptoms and CRP in the usually insomnia group suggest that insomnia severity may be an important factor to consider in understanding the relationships between affective and inflammatory processes.