Change in electrophysiological correlates of recognition memory processes following mild traumatic brain injury

There is a considerable body of evidence showing the presence of long term change in electrophysiological correlates of attentional processes following mild traumatic brain injury (MTBI). As well there is a large body of research demonstrating impairment in cognitive functioning, including memory and attention processes, in the acute stage following MTBI. However, to date there has been little research investigating long term change in electrophysiological correlates of memory function following MTBI. More specifically, there is virtually no research investigating long term change in electrophysiological correlates of recognition memory function following MTBI. Therefore the present research aimed to investigate the electrophysiological correlates of recognition memory processes in males who had suffered an MTI3I. Event related potentials (ERPs) were collected during the encoding phases and the recognition phases from three recognition memory paradigm experiments. In Experiment 1, 22 males who experienced an MTBI 1 month prior, 21 males who experienced an MTBI 24 months prior, and 19 control group males completed two recognition memory tasks, one under deep and one under shallow Level-of-processing conditions, using low frequency nouns as stimuli. ERP data from the deep Level-of-processing encoding phase showed that the expected subsequent memory effect of greater mean amplitude (500-800ms) for subsequently remembered than subsequently missed studied words was present for the control group and the 24 month MTBI group, yet absent for the 1 month MTBI group. Recognition phase data showed that neither MTBI group displayed the implicit memory effect of greater mean amplitude (300-500ms) parietally for missed old words compared to unstudied words under shallow Level-of-processing conditions, whereas the control group did. Experiment 2 employed the same experimental paradigm as Experiment 1 and involved MTBI participants from both MTBI groups from Experiment 1. Fifteen participants from the 1 month MTBI group completed the experimental tasks at approximately 7 month following Experiment 1 (labelled 'Time 2'), and 12 of these participants completed the experimental task a third occasion after a further 10 months (labelled 'Time 3'). Sixteen participants from the 24 month MTBI group completed the same experimental task approximately 13 months after Experiment 1 (labelled 'Time 2'). The results from the encoding phase of Experiment 2 showed that the 1 month MTBI group displayed a subsequent memory effect at Time 2. However, at Time 3 for the 1 month MTBI group and at Time 2 for the 24 month MTBI group, no subsequent memory effects were present due to an apparent increase in mean amplitude elicited by subsequently missed old words. These results suggest that there may be impairment or change in processes involved in encoding meaningful information which, over months and years following the acute post-injury phase, may lead to adaption in these processes. Within recognition phase data results from both MTBI groups showed an absence of the parietal implicit memory effect (300-500ms) for missed old words encoded under shallow Level-of-processing conditions consistent with results from Experiment 1. Experiment 3 involved an MBTI group who experienced an MTBI approximately 24 month prior, and a matched control group of participants completing an associative recognition memory paradigm. ERP results from the encoding phase showed that both groups displayed the N400 effect for word-pairs sharing an association-relationship-only, and for word-pairs sharing both semantic-plus-association relationships in comparison to word-pairs sharing neither association or semantic relationships. However, while the control group displayed an expected N400 effect for word-pairs sharing a semantic-relationship-only, the MBTI group did not. ERP results from the recognition phase showed that as expected the control group displayed late parietal old/new effects for the three experimental word-type conditions of: association-relationship-only; semantic-plus-association relationship; and semantic-relationship-only. However, the MBTI group displayed late parietal old/new effects for semantic-relationship-only and for association-relationship-only, not for semantic-plus-association relationship word pairs. In addition, both groups displayed the expected FN400 old/new effects for association-relationship-only word-pairs however the MTBI group also displayed FN400 old/new effects for semantic-relationship-only word-pairs. These results suggest that the MBTI group may be applying a strategic encoding method, possibly interactive imagery, to accommodate for possible impairment in processes involved in encoding meaningful information. In all experiments MTBI participant groups displayed ERP results inconsistent with the control groups and/or inconsistent with previous relevant research. Results from the encoding phases indicated that MBTI may lead to impairment and/or long term change in the electrophysiological correlates of processes responsible for encoding semantic information. ERP data collected within the acute post injury phase from MTBI participants in Experiment 1 suggested the impairment in processes associated with encoding words under semantic conditions whereas ERP data in Experiment 2 suggested that MTBI participants may be employing adaptive strategies to encode words under semantic conditions. These adaptive strategies may reflect processes of a controlled nature where attempts are made to deeply and elaborately encode all presented words. Experiment 3 provided ERP data from MTBI participants, who were 24 month post-injury, to support results from Experiments 1 and 2 indicating long term impairment, or change, in encoding processes associated with semantic information. This participant group's data lacked the ERP indicator associated with encoding word-pairs sharing only a semantic relationship yet displayed the ERP indicator associated with encoding word-pairs sharing an association relationship. This suggests that impairment or change in processes at encoding may be specific to semantic information.