413: Development of a network for interactions among biopsychosocial dimensions of chronic low back pain

Topic: Clinical Science
Subtopic: Pain category (ICD-11), Chronic primary musculoskeletal pain, Psychology, Developing methodology/technology, Treatment/management

Background and Aims: Network analysis describes complex interactions among features, enabling examination of system-level relationships across domains (1). Low back pain is a leading cause of disability, with approximately 40% of individuals experiencing pain longer than three months, termed chronic low back pain (CLBP) (2). Growing evidence indicates that multiple biopsychosocial factors are associated with CLBP and may interact to exacerbate its persistence (3,4). However, the nature of these interactions remains poorly understood. Here using a network analysis we aimed to uncover and characterize interrelations among biopsychosocial domains and identified the most influential features via centrality metrics in CLBP individuals.

Methods: This study utilized data from the Quebec Low Back Pain Study, a longitudinal prospective cohort initiated in November 2018 comprising 4,489 CLBP participants. We modeled relationships among baseline biopsychosocial features— extracted from Canadian adaptation of the NIH Minimum Dataset (5)— as networks, where nodes represent features and edges encode statistical or causal dependencies among them. Undirected network was inferred using distance correlation. Directed network was constructed using the Linear Non-Gaussian Acyclic Model (LiNGAM) (6), which estimates plausible causal directions. Influence maximization was performed using the Independent Cascade (IC) (7) model to identify the most influential features in each network.

Results: In undirected network, physical function and pain interference were the most central nodes, followed by depressive symptoms. In directed network, kinesiphobia, catastrophizing, and widespread pain were key downstream hubs receiving multiple causal inputs, whereas pain interference, physical function, and depressive symptoms were major upstream drivers exerting causal influence. IC model identified pain interference and physical function as the most influential features in undirected and directed networks, respectively. Bootstrapped confidence intervals indicated narrow variability for edge weights around top central nodes, and centrality stability coefficients remained above the recommended threshold (τ = 0.70) even after removing up to 50% of the sample.

Conclusions: Using undirected network, we identified pain interference, physical function, and depressive symptoms as the most central features shaping the overall system. The directed network further revealed that physical function, pain interference, and depressive symptoms exert causal effects on downstream variables, such as kinesiophobia, catastrophizing, and widespread pain. IC model confirmed the pivotal role of physical function and pain interference as the drivers of network-wide activation. Findings underscore that interventions targeting functionality and pain interference—rather than pain intensity alone—may yield broader improvements across psychological and functional domains. Adopting a network-informed, multidimensional assessment framework can enhance individualized pain management in CLBP individuals.

References: 1. Borsboom D, Cramer AO. Annu Rev Clin Psychol. 2013;9:91-121.
2. Hartvigsen J, et al. The Lancet. 2018;391:2356-2367.
3. Stevans JM, et al. JAMA Netw Open . 2021;4:e2037371-e2037371.
4. Chou R, Shekelle P. 2010;303:1295-1302.
5. Angarita-Fonseca A, et al. Pain. 2023;164:325-335.
6. Shimizu S, et al. JMLR. 2011;12:1225-1248.
7. Wang C, et al. Data Min Know Discov. 2012;25:545-576.

Presenting Author: Pouya Rabiei
Submitting Author: Pouya Rabiei

Senior Author: Hugo Massé-Alarie and Patrick Desrosiers shared senior author

IASP 2026 Global Year theme of Neuropathic Pain: No

• Hugo Massé-Alarie: PhD - Interdisciplinary Research Centre for Rehabilitation and Social Integration (Cirris), Laval University
• Patrick Desrosiers: PhD - CERVO Brain Research Center, Laval University