Behavioral Improvement and Task-Related Brain Activation Following Cognitive Training in Children/Adolescents and Young Adults: A Three-Level Behavioral and Coordinate-Based Neuroimaging Meta-Analysis

Background/Objectives: Cognitive training provides a framework for examining how repeated practice relates to changes in cognitive performance and task-related brain activation, yet age-stratified patterns of neural effects remain unclear. Methods: We conducted a three-level behavioral meta-analysis and a coordinate-based Seed-based Mapping (SDM) meta-analysis of task-based neuroimaging studies to jointly examine behavioral improvement and task-related activation convergence following cognitive training. Results: Cognitive training was associated with reliable behavioral improvement at the aggregate level (Hedges’ g = 0.546, 95% CI [0.414, 0.678]) and within both children/adolescents and young adults. SDM results indicated that task-related activation changes were spatially heterogeneous and included both increases and decreases across distributed brain regions. In age-stratified analyses, spatial convergence within children/adolescents involved occipital and frontoparietal regions; within young adults, it involved cingulo-insular and posterior midline regions. Exploratory behavioral moderator analyses indicated outcome-dependent variation, with smaller effects for transfer outcomes and larger effects in studies with greater training exposure. Conclusions: Behavioral improvements across included designs were observed in both age groups, whereas task-related activation convergence exhibited spatially heterogeneous, age-stratified patterns. These findings suggest that training-related neural effects are better characterized as spatial convergence patterns across studies rather than as a single uniform activation change. The age-stratified maps describe spatial convergence within each age group and should not be interpreted as statistically established developmental differences.