Background: An increasing number of studies have examined the effects of training of cognitive and other tasks on brain structure, using magnetic resonance imaging.
Methods: Studies combining cognitive and other tasks training with longitudinal imaging designs were reviewed, with a view to identify paradigms potentially applicable to treatment of cognitive impairment.
Results: We identified 36 studies, employing training as variable as juggling, working memory, meditation, learning abstract information, and aerobic exercise. There were training-related structural changes, increases in gray matter volume, decreases, increases and decreases in different regions, or no change at all. There was increased integrity in white matter following training, but other patterns of results were also reported.
Conclusions: Questions still to be answered are: Are changes due to use-dependent effects or are they specific to learning? What are the underlying neural correlates of learning, the temporal dynamics of changes, the relations between structure and function, and the upper limits of improvement? How can gains be maintained? The question whether neuroplasticity will contribute to the treatment of dementia will need to be posed again at that stage.