Cognitive training in neurological disorders: meta-analytic evidence for behavioral gains and neural correlates

Cognitive decline in neurological disorders substantially impairs daily functioning and quality of life, underscoring the need for effective non-pharmacological interventions. We aimed to quantify the behavioral benefits of cognitive training, characterize convergent patterns of task-related brain activation changes, and examine moderators of the neural responses underlying training effects. We conducted a meta-analysis of 21 task-based neuroimaging studies. Behavioral outcomes were synthesized using multivariate meta-analysis, while neural changes were examined with seed-based d mapping (SDM) to identify spatially consistent activation differences between training and control groups. Moderator analyses evaluated training parameters, study designs, and participant characteristics, and brain–behavior associations were assessed to link regional activation changes with cognitive improvements. Cognitive training produced a significant, moderate improvement in cognitive task performance (Hedges’ g = 0.451, 95% CI 0.207–0.696; p < .001). Neuroimaging meta-analysis showed increased task-evoked activation after training in the bilateral precuneus and the left precentral gyrus (L. PreCG) relative to controls. Only precuneus activation increases were associated with behavioral gains. Moderator analyses indicated reduced precuneus activation but increased L. PreCG activation during transfer relative to trained tasks, and reduced L. PreCG activation for adaptive versus fixed-difficulty training. Passive controls also showed stronger L. PreCG activation changes than active controls, whereas participant characteristics and training dose showed no significant effects. Cognitive training improved cognitive task performance and was accompanied by a convergent activation signature centered on the bilateral precuneus and the L. PreCG in neurological disorders. Precuneus engagement tracked cognitive gains, highlighting it as a candidate neural marker of training-related plasticity in this population.