Signs of REM sleep dependent enhancement of implicit face memory: a repetition priming study

Abstract

Faces are processed and stored in distinct neuroanatomical systems. Based on evidence of a critical role of sleep in memory processes, we investigated the impact of nocturnal sleep on implicit memories for faces in healthy men. Face repetition effects in reaction times were compared across sleep periods early in the night, which are dominated by slow wave sleep (SWS), and late in the night, where rapid eye movement (REM) sleep prevails, as well as across corresponding nocturnal intervals of wakefulness. An inverse priming effect was found selectively across REM sleep rich late sleep, as indicated by distinctly prolonged response latencies to previously presented faces compared with novel faces after this period of sleep (P<0.05). We assumed this inverse priming to reflect a facilitated identification of previously presented faces after extended REM sleep periods, thereby producing interference with the response generation in our task which did not require face identification but rather required recognizing formal features of the faces. This interpretation was supported by a supplementary experiment where enhanced positive repetition priming was found across late, REM sleep dominated sleep in a task requiring face identification. Together, these findings indicate that implicit face memories particularly benefit from REM sleep associated brain mechanisms.

Introduction

Faces play an important role in human social life. In order to be able to interact appropriately with a familiar person, we must be able to store and recognize immediately his or her identity, which is commonly accomplished by looking at the face because the face includes the most reliable invariant features of a person and conveys additional emotional information. In fact, faces are processed differently from other visual stimuli (Farah, 1996), and specific face processing mechanisms have been developed in the human brain. This is most impressively demonstrated by neuropsychological case studies of patients suffering from prosopagnosia, who are selectively impaired in the recognition of familiar faces but not of other familiar objects (Damasio et al., 1982, McNeil and Warrington, 1993). Most likely, facial stimuli are processed in specific occipito-temporal brain areas, in particular the fusiform gyrus (Damasio et al., 1990, Haxby et al., 2000, Kanwisher et al., 1997, McCarthy et al., 1997). Such specific processing of faces in visual brain areas makes sense because recognition of faces implies recognition of visually invariant features.

Recognition of a known face is an automatic and unstoppable process (Young and Bruce, 1991) and thus represents a prototypical example of implicit memory, which—in contrast to explicit memory—is accomplished involuntarily and is not subject to conscious control (Squire, 1992). Numerous studies have provided evidence that sleep is critically involved in the consolidation of memories (Gais et al., 2000, Stickgold et al., 2000, Fischer et al., 2002). Moreover, sleep appears to be a condition most suitable for studying the distinction between different memory systems (e.g. Tilley and Empson, 1978, Plihal and Born, 1997, Plihal and Born, 1999, Wagner et al., 2001, Peigneux et al., 2001). Specifically, several studies have indicated differential effects of slow wave sleep (SWS) and rapid eye movement (REM) sleep depending on the type of task. While SWS exerted beneficial influences on tasks typical for explicit memory, like memory for paired-associate wordlists and for spatial relations (Barrett and Ekstrand, 1972, Ekstrand, 1977, Fowler et al., 1973, Yaroush et al., 1971, Plihal and Born, 1997, Plihal and Born, 1999), REM sleep appeared to be more supportive in tasks representing implicit memory processes, like mirror tracing, wordstem priming and visual discrimination (Plihal and Born, 1997, Plihal and Born, 1999, Karni et al., 1994). To our knowledge no study so far has dealt with the role sleep plays in face memory.

The present study aimed to investigate differential effects of SWS and REM sleep on implicit memory for faces. For this purpose, we used in two experiments (the main experiment and a supplementary study) the paradigm of repetition effects in reaction times, which addresses implicit activation of memories in the purest fashion since subjects are never informed that memory processes inferred from the reaction time data are the actual target of investigation. Repetition effects are typically reflected behaviorally in accelerated responses to repeated compared with novel stimuli. This phenomenon, called repetition priming, is well documented for a broad variety of stimuli including faces (Ellis et al., 1987, Schweinberger et al., 1995). Although face repetition priming is commonly studied over short time intervals of a few minutes, it can be also observed across considerably longer time intervals of several weeks (Flude, 1991). The extent of priming, however, usually fades with time, although the time course of fading appears to vary considerably depending on the type of priming, materials and task requirements (Squire et al., 1987, Schacter, 1987, Roediger and McDermott, 1993). Also, a prolongation of reaction times for repeated stimuli (i.e. an inverse or ‘negative’ priming effect) can be observed under conditions of interfering response demands (Tipper, 1985, Khurana et al., 2000).

Based on previous studies using non-facial learning materials, we hypothesized a selective beneficial effect of REM sleep also on implicit memory for faces in reaction time repetition tasks. To dissociate effects of SWS and REM sleep, we adopted an experimental protocol (Barrett and Ekstrand, 1972, Ekstrand, 1977, Plihal and Born, 1999) that compared the effects of early and late nocturnal sleep on memory. Since SWS prevails during early sleep whereas late sleep is dominated by REM sleep, conclusions concerning the functions of these sleep stages can be drawn. Although sleep duration remains limited, this study design minimizes the negative side effects of arousals accompanying selective sleep deprivation (Horne and McGrath, 1984, Born and Gais, 2000).