Retraining the brain's map of the body.

What is graded motor imagery?

Graded motor imagery, or GMI, was developed largely by G. Lorimer Moseley and colleagues for conditions such as complex regional pain syndrome and phantom limb pain. It works in three stages that gradually increase the demand on the brain's movement networks without immediately asking for painful movement:

1. Left/right judgments. The person looks at pictures of hands or feet and decides, as quickly as possible, whether each is a left or right limb. This engages the brain's body maps implicitly. People with chronic limb pain are often slower and less accurate at judging the affected side.
2. Explicit motor imagery. The person imagines moving the painful limb without actually moving it. Imagined movement activates many of the same brain areas as real movement.
3. Mirror therapy. A mirror is placed so that the reflection of the healthy limb appears where the painful limb should be. Moving the healthy limb creates the visual impression that the painful limb is moving comfortably.

The logic is to approach painful movement in graded steps, much like exposure therapy, giving the nervous system safe experiences of movement before asking it to tolerate the real thing. Our guides to graded motor imagery and mirror therapy describe each stage in practice.

What do the trials show for graded motor imagery?

In an early randomized trial in 13 people with long-standing CRPS type 1, Moseley tested two weeks each of hand laterality recognition, imagined movements and mirror therapy (Moseley, 2004). The program reduced pain on a neuropathic pain scale by roughly 25 points compared with ongoing management, and the effect was repeated when the control group later crossed over to treatment.

A 2006 trial in Neurology extended the approach to people with CRPS type 1 or phantom limb pain (Moseley, 2006). On a 100 mm visual analogue scale, pain fell by an average of 23.4 mm in the motor imagery group compared with 10.5 mm in the control group, and the gains were maintained at six months. The author noted that the mechanism of the effect was not clear.

A 2013 systematic review and meta-analysis by Bowering and colleagues, including Moseley and Tasha Stanton, pooled the evidence for GMI and its components (Bowering et al., 2013). Comparing GMI with usual physiotherapy, they found a large pooled effect on pain, an effect size of 1.06, but the estimate came from only two studies with 63 participants. Results for motor imagery on its own were conflicting, while mirror therapy and full GMI showed positive effects. The authors concluded that GMI and mirror therapy may be effective, and that the evidence was limited and more rigorous studies were needed. A later review of rehabilitation for CRPS has examined these approaches among others (Shafiee et al., 2023).

What did the JAMA trial of sensorimotor retraining find?

A larger test of the body-map idea came in 2022 (Bagg et al., 2022). The researchers studied 276 adults with chronic low back pain and compared graded sensorimotor retraining with a sham procedure plus attention control. People in the intervention group attended 12 weekly sessions plus home training, focused first on education about pain and then on support to move and stay physically active despite back pain, building on the idea that the brain's processing of the back can be retrained. The control group spent a similar amount of time receiving sham laser, sham shortwave diathermy and sham non-invasive brain stimulation.

The between-group difference was 1.0 point on a 10-point scale, favoring retraining. The authors described the improvement as small and called for further research on how well the result generalizes, since the trial was run at a single center. That measured conclusion is worth noticing: retraining helped, the sham group improved too, and the extra benefit of the real program was modest.

What do these approaches have in common?

Graded motor imagery and sensorimotor retraining share three principles with the rest of therapeutic neuroscience:

• Pain involves the brain's interpretation of the body, not only the state of the tissues. Changing that interpretation can change pain.
• Grading matters. Starting with tasks the nervous system can tolerate and progressing step by step reduces threat and builds confidence.
• Experience teaches more than explanation. Education sets the stage, but repeated safe experiences seem to be where lasting change happens.

The same principles underpin newer tools that use virtual reality to change how people see and feel their bodies, which the next article in this series, on virtual embodiment, explores. Karuna's program carries the graded, experience-first approach into VR; see how it works.

When is body-map retraining worth considering?

For clinicians, graded motor imagery can be considered for CRPS or phantom limb pain, particularly when movement itself is highly threatening. For back pain, sensorimotor retraining is a reasonable option within a broader program, with realistic expectations about the size of the effect.

For people with pain, these approaches can feel unusual. Imagining movement or watching a mirror may not seem like real treatment. They are grounded in a serious body of research on how the brain represents the body, and they can offer a gentle way back toward movement when direct movement feels impossible.