Detecting the onset of accelerated long-term forgetting: evidence from temporal lobe epilepsy.

Accelerated long-term forgetting (ALF) refers to a slowly developing anterograde amnesia in which material is retained normally over short delays but then forgotten at an abnormally fast rate over days to weeks. Such long-term memory impairment is not detected by standard clinical tests. This study analysed ALF in a temporal lobe epileptic, RY. Key issues addressed were: (i) the timeframe of ALF onset; (ii) whether disruption of memory consolidation during sleep is a necessary requirement for precipitating ALF; (iii) the effectiveness of repeated recall in limiting the impact of ALF. RY's memory for novel word-pairings was compared with that of matched controls using cued-recall and forced choice recognition (FCR) tests at multiple delays (5, 30, 55, 240 min). To investigate the impact of repeated recall some pairings were recalled at all intervals, and all material (repeatedly and non-repeatedly recalled) was tested again after a 24h delay. RY's initial learning and performance at 30 min were normal, but by 55 min both his cued-recall performance and the subjective quality of his recognition memory were significantly impaired. This suggests disruption of secondary consolidation processes occurring relatively soon after learning. It also raises the possibility of developing a standard test to diagnose ALF within a single clinical session rather than requiring multiple visits. Since RY remained awake it appears that disruption of memory consolidation during sleep is not a necessary condition for him to experience ALF. Repeated recall at multiple time-points within the first 4h sustained normal recall performance to 24h, indicating repeated recall could form the basis for a protective strategy.