To the Editors:

We were interested in the article by Moseley et al (1), published in a recent issue of Arthritis Care & Research, which demonstrated that motor imagery of the affected extremity increases pain and swelling in patients with chronic pain of the wrist/hand, including those with complex regional pain syndrome (CRPS) type 1.

We have observed a similar spectrum of responses across a heterogeneous CRPS population with imagined movements without pictorial prompts. Our protocol asks subjects to flex and extend their wrists (upper extremity CRPS) or ankles (lower extremity CRPS) to establish the active range of movement before proceeding to similar imagined movement, with their eyes shut, of the unaffected extremity followed by the affected extremity. The vividness and ease of the task and changes in sensation are noted. Finally, still with their eyes closed, the subjects are again asked to perform the imagined movement of the affected extremity, and this time to mirror the extent of the movement with their unaffected extremity. This allows measurement of the perceived range of movement.

We have observed that conscious imagined movement of the affected joint(s) takes longer than the unaffected side and presents with a more reduced range of motion than actual execution, as well as eliciting pain in ∼33% of patients (n = 6). Repeating this process for the contralateral unaffected joints does not elicit symptoms similar to Moseley (2), but in some cases, imagined movement of the ipsilateral unaffected joints increases pain at the affected joint. For example, imagining making a fist with the unaffected right hand increased right knee pain in a patient with CRPS affecting the right knee. This lends credence to the theory that it is the motor imagery rather than the ensuing attention on the affected extremity that drives symptom changes (3, 4).

Conscious motor imagery represents the second stage of Moseley's 3-stage motor imagery program, success for which relies on sequential cortical activation (5), suggesting that the negative responses to imagined movements would be alleviated by prior laterality training. This assumption is supported by the finding of a positive relationship between pain and swelling increases and reaction time to a laterality task, but it clearly awaits further study.

Our experience of laterality training has shown similar inconsistency of response to imagined movement across a cohort of patients with upper and lower extremity CRPS. Two of 15 patients reported a sustained increase in pain and other symptoms in the affected extremity when performing, and after performing, reported extremity laterality tasks that did not abate with task familiarity or prior relaxation training. Furthermore, these patients, when trialed on laterality training of their unaffected arms, especially when the pictures showed awkward postures, reported similar, though of a lesser magnitude, sensations of pain and dysesthesia in the affected regions. These effects occurred in the absence of any observed voluntary movement. At present, we are unable to predict which patients will have a positive or negative response to hand laterality or imagined movement tasks, but disease duration, extent of cortical reorganization as defined by greater pain intensity (6), presence of referred sensations (7), and body perception disturbance (8) appear to be significant factors. Further research is required and is currently underway.

The mechanisms underlying our findings may be related to autonomic activation, as Moseley et al (1) showed that changes in swelling were related to arousal, but such studies have yet to be completed for unconscious motor imagery (laterality task). The hand laterality task is considered to provide a measure of the integration between information processing, working body schema, and premotor processes activating a restricted parietofrontal network (9). In a particular subset of CRPS patients, Gieteling et al (10) showed significantly reduced cortical activation in areas spanning motor planning, sensory integration, autonomic function, and pain during conscious imagined movement (compared with healthy controls and compared with unaffected hand). Gustin et al (3) suggest that the pain arising from motor imagery is linked to the strength of motor cortex stimulation, on the basis that the analgesia from intense electrical stimulation of the primary motor cortex inhibits neuronal activity in the thalamus. Although laterality tasks do not activate the motor cortex, it is feasible that low-intensity stimulation of the cortex stimulates the pain neurosignature at increasingly lower thresholds as described by Moseley (11), in which smaller and seemingly less relevant inputs (diversification of threat) are sufficient to activate the individual-specific pain neuromatrix and thus produce pain. Therefore, imagined movement (conscious or unconscious), which activates similar cortical pathways to actual execution, is perceived as a threatening input for which the output is pain.

We concur with the findings of Moseley et al that responses to conscious imagined movement lack consistency, and report that unconscious motor imagery, sometimes of unaffected areas, can worsen symptoms in some patients with CRPS (1). This heterogeneity highlights the possibility of subtypes and underscores the need for close clinical supervision when undertaking a motor imagery program and for future studies to link patient characteristics to treatment and outcome.

Jane Hall PhD, MCSP*, David Blake FRCP, MB, CHB(Hons)*, Candida McCabe PhD, RGN†, * Royal National Hospital for Rheumatic Diseases Bath, UK, † Royal National Hospital for Rheumatic Diseases and University of Bath Bath, UK.