Working memory in chronic pain: evidence for task-specific rather than global differences

Chronic pain is often accompanied by cognitive complaints, but evidence for global working memory problems is mixed. We tested whether working memory performance differences in chronic pain are global or task-specific and whether model-based analysis could help distinguish differences in processing efficiency, response caution or sensory/motor speed. In a preregistered online case-control study, 99 adults with mixed chronic pain conditions and 87 pain-free controls completed visuo-verbal, visuo-spatial and auditory-temporal n-back tasks at low (0/1-back) and high (2-back) load. Accuracy and reaction times were analysed with mixed effects models. Drift diffusion modelling decomposed performance into processing efficiency (drift rate), response caution (threshold separation) and non-decision (sensory/motor) time. Higher load reduced accuracy and slowed responses in both groups. There was no evidence of a global working-memory deficit in the chronic pain group. The clearest group difference was a larger load-related drop in accuracy in the auditory-temporal task (odds ratio 0.64, 95% CI 0.56 to 0.73), persisting after adjustment for mood, sleep and fatigue. Load-related slowing in visuo-verbal (6.7% slower, 5.1% to 8.2%) and auditory-temporal tasks (3.6% slower, 1.7% to 5.4%) were attenuated after adjustment. Diffusion modelling showed no evidence for sensory/motor slowing, but rather greater response caution in the auditory-temporal task and small efficiency (drift rate) reductions in low-load visual conditions. The results do not support a global working-memory capacity loss account in this mixed chronic pain sample. Rather, they suggest task-specific performance differences, most evident in auditory-temporal processing, with response caution as a plausible contributor.