Individual differences in face recognition memory: comparison among habitual short, average, and long sleepers.

Research indicates that habitual short sleepers show more rapid accumulation of slow-wave sleep at the beginning of the night. Enhancement in performance on declarative memory tasks has been associated with early NonREM sleep, consisting of the highest percentage of slow-wave sleep. Twenty-four subjects (eight short sleepers <or=7h, nine average >7 but <9h, seven long >or=9h) were tested. Subjects were presented with unfamiliar face stimuli and asked to memorize them for a subsequent test. Following sleep, the subjects were presented with the 40 "old/studied" items intermixed with 40 new and asked to indicate the previously presented stimuli. Event-related potentials (ERPs) were analyzed to verify the existence of the "Old/New" effect, i.e. amplitude difference [in ERPs] between the old and new stimuli. ANOVA on the scores revealed a significant interaction between the stimuli and group. Post-hoc test on the studied items revealed more accurate responses in the short sleepers compared to the average and long sleepers. Strikingly, the long sleepers failed to show significant retention of the old/studied items, with their recognition of old faces not different from chance. Reaction time (RT) responses were faster for the old vs. the new items. Pearson correlation revealed a significant negative correlation between accuracy and sleep duration in the short sleepers. However, long and average sleepers showed a positive correlation between the two variables. ANOVA performed on the ERPs revealed main effects of stimuli and site, and no interactions involving the group factor. In conclusion, our data show that individual differences in recognition memory performance may be associated with differences in habitual sleep duration.