• Jie Zhang, Liang Tian, Li Zhang, Ruidong Cheng, Ruili Wei, Fangping He, Juebao Li, Benyan Luo, and Xiangming Ye.
• Department of Rehabilitation Medicine, Zhejiang Provincial People's Hospital, People's Hospital of Hangzhou Medical College, Hangzhou, China.
• J. Neurol. Neurosurg. Psychiatr. 2019 Jan 1; 90 (1): 98-107.

ObjectiveTo investigate relationships between cognitive domains and white matter changes in different regions in patients with cognitive deficits after traumatic brain injury (TBI).MethodsDatabases including PubMed, Embase, Web of Science and CENTRAL were searched for studies published before 5 August 2017. Correlation coefficients between cognition and white matter integrity, measured by diffusion metrics, including fractional anisotropy (FA), were pooled from 49 studies including 1405 patients. The influence of demographic factors was assessed by meta-regression analysis.ResultsSignificant pooled FA-executive correlations (p<0.001) were found across various regions, including the corpus callosum (CC) (r=0.42, 95% CI 0.30 to 0.54), superior longitudinal fasciculus (r=0.50, 95% CI 0.41 to 0.59) and internal capsule (IC) (r=0.49, 95% CI 0.37 to 0.61). The fornix (r=0.62, 95% CI 0.45 to 0.78) and cingulum (r=0.57, 95% CI 0.34 to 0.81) particularly correlated with memory (p<0.001). The CC and IC also showed significant relationships with attention and processing speed (p<0.001). Demographic factors had no influence overall, except that studies with a greater proportion of males had stronger correlations between memory and white matter (p<0.05).ConclusionsFA is the most sensitive metric for detecting post-TBI cognitive decline across various domains. Representative white matter regions, such as the CC and IC, perform better than whole-brain white matter for reflecting a wide range of cognitive domains, including memory, attention and executive functions. Moreover, the fornix and cingulum particularly reflect memory function. They yield insights into particular imaging indicators that have neuropsychological value.© Author(s) (or their employer(s)) 2019. No commercial re-use. See rights and permissions. Published by BMJ.

25 keywords...

hide…