Can the trail making test black and white predict white matter hyperintensity on MRI?

Abstract

White matter hyperintensity (WMH) in the general population has clinical importance. However, MRI (magnetic resonance imaging) is not suitable for screening purposes due to high cost and restricted availability. We hypothesized that performance on the trail making test black & white (TMT-B&W) may correlate with WMH and provide a screening tool for patients with WMH. Intergroup comparisons between WMH (+) (a Fazekas score of ≥2) and WMH (−) (<2) were performed for time variables in type A, B, B-A, B/A, B*A and errors in type A and B. To determine an ideal cutoff value and confirm the diagnostic performance of TMT-B&W on the presence of WMH (+), the area under the receiver operating characteristics curve (AUC) was analyzed. Multiple linear regression analysis was used, adjusting for age and education, to correlate the normalized WMH volumes and TMT-B&W performance. We show that higher WMH correlates with a significant increase in completion time for type A (p < 0.001) and B (p = 0.006) TMT-B&W. In AUC analysis, the optimal cutoff time to complete type A was 75 s [sensitivity, 66.7%; specificity, 67.2%; %; p = 0.001; AUC, 0.729; 95% confidence interval (CI), 0.600–0.858]. Additionally, there were significant correlations between normalized WMH volume and TOTAL-A (B = 0.233, p = 0.016) and errors in type A (B = 0.353, p < 0.001). We identified the correlation between the performance in TMT-B&W and WMH burden, and propose its potential clinical utility in predicting brain WMH burden. © 2019 Elsevier Ltd