The aim of this study was to clarify which cognitive mechanisms underlie Trail Making Test (TMT) direct and derived scores. A comprehensive review of the literature on the topic was carried out to clarify which cognitive factors had been related to TMT performance. Following the review, we explored the relative contribution from working memory, inhibition/interference control, task-switching ability, and visuomotor speed to TMT performance. Forty-one healthy old subjects participated in the study and performed a battery of neuropsychological tests including the TMT, the Digit Symbol subtest [Wechsler Adult Intelligence Scale (Third Version) (WAIS-III)], a Finger Tapping Test, the Digits Forward and Backward subtests (WAIS-III), Stroop Test, and a task-switching paradigm inspired in the Wisconsin Card Sorting Test. Correlation and regression analyses were used in order to clarify the joint and unique contributions from different cognitive factors to the prediction of TMT scores. The results suggest that TMT-A requires mainly visuoperceptual abilities, TMT-B reflects primarily working memory and secondarily task-switching ability, while B-A minimizes visuoperceptual and working memory demands, providing a relatively pure indicator of executive control abilities.

The Trail Making Test (TMT) has been a useful assessment tool to investigate executive function. Several studies have recently improved the existing TMT norms by mean of large samples of healthy individuals stratified by a number of demographic variables from different populations. In contrast, criticisms have been raised about the utility of norms from healthy samples to detect changes across time in clinical samples where TMT performance used to be altered. In addition, few studies have compared groups of patients with deficits in TMT performance, making it difficult to decide whether a single set of norms is sufficient to assess different clinical populations. We provide normative data from three large samples of patients with traumatic brain injury (TBI) (n=90), schizophrenia spectrum disorders (n=127), and healthy Spanish speakers (n=223). Differences between healthy participants and patients in all TMT direct (TMT-A, TMT-B) and derived (B-A, B:A, B-A/A) scores were found. TMT performance was poorer in TBI patients than in schizophrenia patients except for the B:A and B-A/A scores, suggesting a similar underlying executive deficit. Normal ageing impaired both direct and derived TMT indices, as revealed by lower scores in the healthy elderly group (55-80 years old) as compared with young (16-24) and middle-aged (25-54) healthy participants. Three different sets of norms stratified by age, education, or both are presented for clinical use. Recommendations on TMT scores are made for future research.