Disability is primarily predicted by pain intensity and avoidance in patients with chronic back pain

2520 

Abstract Submission 

Lejian Huang¹, Binbin Wu², Kenta Wakaizumi¹, Rami Jabakhanji¹, Apkar Apkarian¹, Marwan Baliki¹

¹Northwestern University, Chicago, IL, ²The Second Affiliated Hospital and Yuying Children's Hospital of Wenzhou Medical University, Wenzhou, Zhejiang

Lejian Huang  
Northwestern University
Chicago, IL

Binbin Wu  
The Second Affiliated Hospital and Yuying Children's Hospital of Wenzhou Medical University
Wenzhou, Zhejiang

Kenta Wakaizumi  
Northwestern University
Chicago, IL

Rami Jabakhanji  
Northwestern University
Chicago, IL

Apkar Apkarian  
Northwestern University
Chicago, IL

Marwan Baliki  
Northwestern University
Chicago, IL

Chronic back pain (CBP) and pain disability represent a formidable challenge to scientists and health professionals. A better understanding of the behavioral and neural factors affecting disability can provide a conceptual framework for the evaluation and treatment of CBP. In this study, we demonstrated that disability (ODI) is primarily predicted by pain intensity (NRS) and avoidance (PASS_avoidance) and then identified the unique brain networks mediating their effects by analyzing resting-state fMRI (rs-fMRI) data.

Three hundred and eighty six patients (256 + 130) with CBP (CBPs) were recruited (179 M, 207 F; 44.6 ± 11.9 years old), 130 of which (82 M, 48 F; 43.9 ± 11.6 years old) were scanned for rs-fMRI data with the following parameters: voxel size = 3.4375 x 3.4375 x 3.5 mm3; TR/TE = 2500/30 ms; flip angle = 90°; in-plane resolution = 64 × 64; field of view = 220 x 220 mm2; number of volumes = 230; slices per volume = 42. All CBPs completed a battery of self-reported behavioral questionnaires related to pain including Oswestry Disability Index (ODI), Numerical Rating Scale (NRS), Beck Depression Inventory (BDI), Pain Anxiety Symptoms Scale (PASS), Positive and Negative Affect Schedule (PANAS), pain duration, Pain Sensitivity Questionnaire (PSQ), and Pain Catastrophizing Scale (PCS).
The 256 CBPs without brain image were evenly allocated to either an exploratory group (Discovery) or a validation group (Validation) with matched demographics and behavioral characteristics.
The correlations between pain disability and the brain network properties and local functional connectivity were investigated. 264 regions of interest (ROIs) were used to construct 13 functional networks (Power, Cohen et al. 2011). Linear regression was used to estimate ßij (i, j = 1, 2, …, 264) for each brain functional connectivity between two ROIs in brain. The connections with their corresponding ROI coordinates and network assignment were recorded that exhibited significant association with ODI (p < 0.001) (80 significant connections were found at this stage).
To determine whether this connectivity pattern mediated the effects of pain intensity and avoidance on disability, we first used a principal component analysis (PAC) to decompose the 80 "betas" into 20 orthogonal factors that accounted for 80.1% of the variance. Following this, the factors were submitted to perform mediation analysis.

Only NRS and PASS-avoidance in behavioral factors showed significant regression across all folds (r = 0.621, p < 0.001 for Discovery group). This relationship was validated in the Validation group (r = 0.577, p < 0.001).
The 80 pair-wise connections associated with ODI mainly localized in default-mode (DMN), somatosensory, and salience networks defined.
PCA identified 20 components that accounted for 80.1% of variance of 80 connections that exhibited significant association with ODI across the 130 CBPs. Mediation analyses determined that the first component was the only one that significantly partially mediated both pain intensity and avoidance on pain disability.

Overall, our results show that disability in CBP patients is primarily predicted by pain intensity and avoidance. DMN exhibits most association between disability and functional connectivity. Besides, we found PASS-avoidance and NRS partially mediate ODI through a common DMN-dominated brain functional network. Taken together, these findings support a predicting role of pain intensity and pain-related avoidance for disability in CBPs, and further confirm the DMN including the mPFC may be the underlying biological mechanisms that mediate pain disability in CBP patients.

Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s)

fMRI Connectivity and Network Modeling ²

Perception: Pain and Visceral ¹

Computational Neuroscience

FUNCTIONAL MRI

Pain