Abstracts

Traditional memory theory argues that the role of the hippocampus in the processing of autobiographic information is time-limited. That is to say, the retrieval of personal events becomes independent of the structure once memories have been integrated into neocortical stores. However, the fMRI evidence pertaining to this effect is inconsistent. The prevalent use of pre-scan interviews might be to blame, which are typically used to procure autobiographic information from each participant prior to scanning. As a consequence, memories are refreshed during this process and could alter subsequent brain activity during in-magnet retrieval. To explore this issue, we constructed an autobiographic memory recall test that did not require a pre-scan interview, but relied on a list of common life experiences (e.g., a job held in high school). Subjects were scanned twice, one-month apart, with the same set of memory probes, in order to evaluate the effect of prior reflection on retrieval activation. Twelve, right-handed participants (18 - 29 years of age) were recruited from the Yale University campus and scanned at 3 Tesla using gradient-echo planar imaging. The images were acquired on the coronal-oblique plane perpendicular to the hippocampus. Behavioral testing revealed that the quality of recall (i.e., vividness, level of detail) did not significantly differ across test sessions. The overall pattern of neural activity was also very similar across sessions, although the magnitude of the response was significantly more pronounced for the initial scan. Subtractions between sessions showed significantly more positive activation for scan one, relative to scan two, within the prefrontal cortex, inferior temporal gyrus, inferior parietal lobule, thalamus, and cerebellum (p [lt] .02). Although this subtraction did not show a significant difference in BOLD intensity within the hippocampus proper, the strength of hippocampal activation relative to baseline was considerably more robust for scan one than scan two. The study demonstrates that recent consideration of autobiographic events weakens the strength of the neural response during subsequent retrieval. While the reason for this effect is unclear, it is possible that participants may not engage in a new memory search during scanning, but rely on their previous (pre-scan interview) recollections of these events. These findings should discourage the use of pre-scan interviews as they complicate the interpretation of the data and can provide false support for traditional consolidation theory. (Supported by NIH: NS38467 NS40497)