Aberrant error monitoring in traumatic brain injuries: A meta-analysis of event-related potential studies.

OBJECTIVE

Although individuals with traumatic brain injuries (TBI) often manifest altered error monitoring, evidence using event-related potentials (ERPs) to index these cortical processes is inconsistent. Therefore, this meta-analysis study aimed to comprehensively compare the error-related negativity (ERN) and error positivity (Pe) between individuals with TBI and healthy controls (HC) from the existing literature.

METHODS

Literature search was performed using PubMed/MEDLINE, Web of Science, and Cochrane Library. The effect sizes (Hedges' g) in the comparisons of ERN and Pe amplitudes between TBI and HC groups were employed by a random-effect, inverse-variance weighted model. The effects of age, TBI severity and experimental tasks on both ERP components were also examined.

RESULTS

Random-effect models showed decreased ERN (g = 0.361, p = 0.010), but intact Pe (g = 0.105, p = 0.443), in those with TBI compared to HC. A further analysis revealed that the adult patients (g = 0.326, p = 0.038), but not the youth patients, showed significant reduction of ERN as compared to the HC. However, we did not find moderating effects of TBI severity and experimental paradigms on either ERN or Pe.

CONCLUSIONS

ERN and Pe reflect separate neurophysiological mechanisms and different aspects of error monitoring in TBI. Our findings suggest that attenuated ERN amplitude may be an electrophysiological parameter of error monitoring deficits in TBI.