Capturing real-life forgetting in transient epileptic amnesia via an incidental memory test

Transient epileptic amnesia (TEA) is an epileptic syndrome characterized by recurrent, brief episodes of amnesia. Patients with TEA often complain of interictal (between attacks) retention deficits, characterised by an 'evaporation' of memories for recent events over days to weeks. Clinical tests of anterograde memory often fail to corroborate these complaints as TEA patients commonly perform within the normal range after the standard 10-30-min delay period. Modified laboratory tests that include a 1-3 week delay period frequently reveal clear evidence of 'accelerated long-term forgetting' (ALF). However, they are not used routinely and lack ecological validity. In the present study we examined whether 'real-life' ALF can be captured via a controlled incidental memory test in TEA patients. To this end, the experimenter told 27 TEA patients and 32 controls a well-rehearsed amusing story, apparently as a way of making light conversation before starting a set of research experiments. Without prior warning, the experimenter subsequently probed the participants' memory of this story via tests of free recall and forced choice recognition after 30 min or 1 week. After 30 min retention was comparable in TEA patients and controls. After 1 week TEA patients retained significantly less story material than controls, and significant ALF was revealed in the TEA patients in the recognition test. Our data show that ALF in a 'real-life' situation can occur even when standard memory tests indicate normal memory function. Moreover, our data suggest that incidental memory tests can capture real-life ALF, and that forced-choice recognition tests might be more sensitive than free recall tests for the detection of real-life ALF.