Dimensions of cognition measured by the Tower of London and other cognitive tasks in head-injured children and adolescents

To determine the factor structure of executive functioning in head-injured (n = 81) and normal children (n = 102), we administered tests of concept formation and problem solving, plus planning, verbal fluency, design fluency, memory (to evaluate semantic organization), and response modulation using a Go/No-Go task. The children who sustained closed head injury (CHI) were divided into subgroups who sustained severe (n = 39) and mild/moderate (n = 42) injury. The CHI groups and normal controls were also grouped according to age at the time of testing (6-8, 9-12, and 13-16). The principal components analysis disclosed a five-factor solution that accounted for 79% of the variance: Conceptual-Productivity (Factor 1), Planning (Factor 2), Schema (Factor 3), Cluster (Factor 4), and Inhibition (Factor 5). Age had a significant effect on Factors 1, 2, and 5, whereas severity of CHI affected Factors 1, 2, 4, and 5. Using hierarchial regression in which the Glasgow Coma Scale score, age, and their interaction were entered first, the volume of frontal lobe lesion contributed significantly to predicting Factors 1 (Conceptual-Productivity) and 2 (Planning), whereas the volume of left frontal lesions also predicted Factor 3 (Schema). The volume of extrafrontal lesions augmented the prediction of Factor 3, supporting the general relation of left hemisphere abnormality to the cognitive variables loading on this factor. Pending replication in a different sample of head-injured children, caution is advised in interpreting the findings due to potential instability of the factor structure.