Abstract
PURPOSE: To map neuropathic pain symptom networks across allostatic load levels in cancer survivors, quantify structural differences, and explore potentially influential symptoms using simulated network perturbations.</p>
METHODS: This cross-sectional study included 5923 UK Biobank cancer survivors stratified into relatively lower and higher allostatic load groups using the sample median. Seven interview-administered DN4 symptoms were analysed using Ising network models, network comparison tests, and simulated perturbations.</p>
RESULTS: Tingling, pins and needles, and numbness formed a stable sensory core across networks (edge weights: 1.24-2.42). Tingling showed the highest expected influence (EI = 6.26, 95% CI: 5.93-6.59) and ranked first in the simulated perturbations. Network structure differed between allostatic load groups (M = 0.78, P = 0.006). The lower allostatic load network showed stronger global connectivity (global EI: 11.67 vs. 8.59, P = 0.001), whereas the higher allostatic load network showed a stronger tingling-itching connection (P < 0.05).</p>
CONCLUSION: The organisation of seven neuropathic pain symptoms differed across relative allostatic load levels. Tingling was potentially influential within the fitted networks and simulations. These findings require longitudinal and interventional validation before clinical application.</p>