Motor Learning in Stroke: Trained Patients Are Not Equal to Untrained Patients With Less Impairment.

The study used a laboratory cursor task, not a real daily functional task, so it shows motor learning under controlled conditions rather than clinical recovery. Only 30 participants were studied, with 10 in each group, so the results may not apply broadly. Training lasted only four consecutive days, so the plateau does not prove that months or years of therapy cannot produce further gains. Participants had to pass a cognitive screen, have no cerebellar lesion or ataxia, and have a minimum biceps strength, so the findings do not cover people with cognitive deficits, cerebellar stroke, or very weak arms. The stroke sample included two people with a 66/66 upper-limb score, which may mean no measurable impairment on that scale, although resampling suggested they did not change the result. The paper does not describe random allocation; groups were defined by impairment level, so differences in learning are mixed with baseline impairment.

No conflicts of interest are stated in the supplied text; the article metadata lists NIH extramural research support.

Do not conclude that training makes a patient equal to a less impaired untrained person. The trained groups matched the untrained groups' pre-training performance, but they had already plateaued across training days, while the untrained groups had not yet trained and still had room for rapid gains.