What can virtual-reality arm training do after stroke?

What the researchers actually compared

This was a systematic review that combined the results of 43 randomized trials involving about 1,893 adults who had a stroke. In each trial, people were randomly assigned to receive virtual-reality-supported exercise therapy for their affected arm, either alone or added to conventional rehabilitation, while the comparison group got conventional therapy or, in a few trials, no specific therapy. The VR sessions typically used interactive screen-based environments to guide arm movements; some programs were designed specifically for rehabilitation and others used commercial games. Sessions commonly lasted 20 to 75 minutes, several times a week, for roughly two weeks to two months.

What the combined findings showed

Across the trials, people who used VR-supported exercise had better movement control of their affected arm, as measured by a standard motor function scale, and a wider range of motion at the shoulder, elbow or wrist. They also had more muscle strength when tested by a therapist and greater independence in everyday activities on two common assessment tools, compared with those who received usual therapy or no therapy. These improvements were found when measurements were taken at the end of the planned therapy period.

What did not consistently improve

Not every aspect of arm recovery changed. Grip strength measured with a hand dynamometer and muscle tightness called spasticity were similar between the VR and comparison groups. Hand dexterity and several measures of arm-and-hand performance did not show a reliable overall advantage, and quality-of-life scores were also much the same. Some subgroup patterns pointed to possibly larger gains with longer VR training, but those were not consistent enough to be treated as a firm finding.

Why lasting benefit is still uncertain

Most trials stopped measuring participants when the therapy programme ended; only about one-third arranged any later follow-up assessment. In the studies that did look afterward, there was no evidence that the improvements were maintained after the intervention ended. This means the review can show benefit during the rehabilitation window, but it cannot yet tell how long the gains last or whether extra sessions would be needed.