Enhanced perceptual priming for neutral stimuli occurring in a traumatic context: Two experimental investigations

Abstract

Intrusive memories in posttraumatic stress disorder are often triggered by stimuli that are perceptually similar to those present shortly before or during the trauma. The present study aims to examine the possible role of perceptual priming in this phenomenon. It further investigates whether the degree of perceptual priming is associated with dissociation and whether both perceptual priming and intrusive memories can be reduced through elaboration.

Two experiments measured perceptual priming for neutral stimuli that immediately preceded a “traumatic” event. Volunteers (N=46, 92) watched a series of “traumatic” and neutral picture stories, and completed a blurred object identification (perceptual priming) memory task, and a recognition memory task. Participants in Experiment 1 were selected to score either high or low on the Trait Dissociation Questionnaire [Murray, Ehlers, & Mayou (2002). Dissociation and posttraumatic stress disorder: Two prospective studies of motor vehicle accident survivors. British Journal of Psychiatry, 180, 363–368]. They also completed a state dissociation measure in the session. Experiment 2 randomly allocated participants to an experimental condition designed to increase elaboration or to a control condition. This experiment also included a measure of intrusive memories.

Both experiments found enhanced perceptual priming for the stimuli that immediately preceded the “traumatic” stories compared to those preceding neutral stories. Participants with high trait dissociation showed relatively stronger perceptual perceptual priming. The degree of perceptual priming for stimuli from the “traumatic” stories also correlated with state dissociation (Experiment 1). Experimental manipulation of the elaboration of the stories showed that elaboration reduced the enhanced perceptual priming effect and the relative probability of reexperiencing symptoms (Experiment 2). The results support the role of perceptual priming in intrusions after traumatic events.

Introduction

Intrusive memories are considered the hallmark symptom of PTSD. They usually consist of brief sensory fragments of the traumatic event, and are accompanied by high emotional distress (e.g., Ehlers & Steil, 1995; Holmes, Grey, & Young, 2005; van der Kolk & Fisler, 1995). Even though it is not uncommon for intrusive memories to have several sensory components, studies investigating the modalities of intrusive memories have repeatedly shown that visual sensations seem to be the most common form (Ehlers et al., 2002; Hackmann, Ehlers, Speckens, & Clark, 2004; Mellman & Davis, 1985; Michael, Ehlers, Halligan, & Clark, 2005). Furthermore, intrusive memories in PTSD are accompanied by a sense of “nowness”, i.e., the impression that the sensations are happening now rather than stemming from the past (e.g., Ehlers, Hackmann, & Michael, 2004; Hackmann et al., 2004; Michael, Ehlers, Halligan, & Clark, 2005).

Cognitive theories of PTSD concur in assuming that intrusive memories are due to the way the trauma memories are encoded, organized in memory, and retrieved (Brewin, Dalgleish, & Joseph, 1996; Conway & Pleydell-Pearce, 2000; Ehlers & Clark, 2000; Foa, Steketee, & Rothbaum, 1989). However, surprisingly little is known about the mechanisms that lead to intrusive memories. The clinical and theoretical literature on PTSD has emphasized the remarkably wide range of external and internal triggers of reexperiencing (Brewin et al., 1996; Foa et al., 1989). Only some of these triggers seem to be semantically linked to the traumatic event, such as TV or newspaper reports about similar events. Interestingly, triggers are often stimuli that are perceptually similar to the stimuli featuring in the intrusion (Ehlers & Clark, 2000; Ehlers et al., 2004). For example, a car crash survivor kept reexperiencing the bright headlights that were coming towards him before the accident. In therapy, it became clear that bright patches of light on a darker background (e.g., sunlight on a lawn) triggered this intrusive memory. Frequently, trauma survivors are not aware which stimuli trigger their intrusive memories and thus experience them as coming out of the blue. Thus, the triggering of intrusive memories appears to be mainly cue-driven.

Ehlers and colleagues observed that the content of intrusive memories is also somewhat surprising. They asked patients with PTSD who had experienced a range of different traumas to describe the content of their intrusive memories. Interestingly, most participants reported that they had sensory memories of stimuli that were present immediately before the traumatic event happened or immediately before the moment that had the largest emotional impact (Ehlers et al., 2002). For example, a man who witnessed the suicide of a person who jumped in front of a train had intrusive images of the sight of the railway tracks as he had seen them just before the person jumped. A subsequent study, in which two independent raters judged the content of participants’ intrusive memories, corroborated these findings. The majority of intrusions (>80%) did not represent the worst moment of the trauma. The most common intrusive memories were stimuli signalling the onset of the event, followed by stimuli that signalled a particularly upsetting moment during the trauma (Hackmann et al., 2004).

Ehlers and Clark (2000) suggested that the easy triggering of intrusive memories in PTSD by perceptually similar cues is a function of strong perceptual priming for stimuli that occurred shortly before or during the traumatic event. The authors propose that perceptual priming, together with strong associative links between these stimuli (S–S associations), and between the stimuli and strong affective responses (S–R associations) (see also Foa et al., 1989; Keane, Zimering, & Caddell, 1985), facilitates cue-driven retrieval of corresponding aspects of the trauma memory.

Cue-driven activation of autobiographical memories is usually inhibited when the memory becomes incorporated into the autobiographical knowledge base (Conway & Pleydell-Pearce, 2000). On the basis of these findings, Ehlers and Clark (2000) suggested that in PTSD there is poor inhibition of cue-driven retrieval of aspects of the trauma because the memory for the traumatic event is poorly elaborated. In particular, it is insufficiently linked with its context, i.e. with things that happened shortly before or afterwards, and with other autobiographical information. The three memory processes (perceptual priming, associative learning, and poor elaboration) are thought to work in conjunction in producing reexperiencing symptoms, i.e., poor memory elaboration in itself is not sufficient for producing intrusive memories (Ehlers & Clark, 2000).

Perceptual priming is a form of implicit memory that refers to the facilitated identification of perceptual objects as a consequence of prior exposure. Many laboratory experiments have demonstrated perceptual priming by showing that degraded perceptual stimuli of words or objects are more easily identified if the person has previously been exposed to the stimuli (Schacter, 1992). Importantly, perceptual priming can occur independently of any conscious recollection of a previous encounter with the stimuli (Schacter, 1992). Hence perceptual priming is thought to be responsible for certain mind-popping phenomena in which perceptual fragments of experience spontaneously materialize in awareness (Schacter, 1996). With respect to the above-mentioned example of the car crash survivor, it thus seems probable that the triggering of the “headlight intrusion” by patches of light reflects the influence of perceptual priming.

The perceptual priming hypothesis received indirect empirical support from studies showing enhanced implicit memory for trauma-related material for people with PTSD compared to those without PTSD (Amir, McNally, & Wiegatz, 1996; Michael, Ehlers, & Halligan, 2005; but see Golier, Yehuda, Lupien, & Harvey, 2003; McNally & Amir, 1996). However, these studies have the limitation that they investigated perceptual priming for words or sentences, which are related to, but dissimilar to the sensory impressions that are reexperienced after trauma. As visual intrusive memories seem to be the most common form of intrusive memories, it would be desirable to study perceptual priming for visual stimuli from the traumatic context that trigger intrusions. This is, however, difficult to achieve in trauma survivors for several reasons. First, traumatized people are often not aware which stimuli trigger their intrusive memories. Second, even if the triggers are identified in therapy, it would be extremely hard to bring them under experimental control, especially as the stimuli differ from one trauma survivor to the other and as it would be next to impossible to match a control group to the traumatized group with respect to duration of initial exposure and the time between initial exposure and the perceptual priming task.

Ehlers, Michael, Chen, Payne, and Shan (2006) therefore developed an experimental paradigm to study visual perceptual priming for stimuli that occur in a traumatic context. The paradigm investigates perceptual priming for neutral objects that occur just before something “traumatic” happens, as clinical observations showed that intrusive memories often feature such stimuli (see the headlights and railway track examples above). Participants watched a series of “traumatic” and neutral picture stories. The content of the first picture is unemotional. It contains neutral preceding stimuli (e.g. a cushion) for which memory is later tested. During the second picture, either something “traumatic” (e.g., a man being attacked with a knife) or something neutral happens. The last picture focuses on the outcome of the story for the main character (e.g., the attacked man is being decapitated). In accordance with the enhanced perceptual priming hypothesis, the results showed that neutral stimuli preceding a “traumatic” event showed enhanced perceptual priming and predicted intrusive memories (Ehlers et al., 2006).

At this stage, it remains unclear how a subsequent “traumatic” event can produce enhanced perceptual priming for previously encountered stimuli. One possible explanation may be that the distress brought on by the “traumatic” event produces an encoding style that favors perceptual processing (e.g., Siegel, 1995). This change in encoding might affect the encoding and storage of previously encountered information, as this is held in consciousness for some time in order to help the organism to connect and make sense of subsequent events.

Further support for the enhanced perceptual priming hypothesis stems from a recent study about emotional memory by Arntz, de Groot, and Kindt (2005). Participants watched a series of slides that were accompanied by a spoken story, which was either neutral or emotional (distressing). A subsequent memory test revealed that participants in the emotional condition performed better on a perceptual identification and on a perceptual recognition task than participants in the neutral condition.

Although the above results point to enhanced perceptual priming as one mechanism underlying intrusions, they need to be extended in order to address several important issues that remain unanswered. It is, for example, not yet known if the encoding style during a traumatic event influences the extent of perceptual priming and if perceptual priming and reexperiencing can be reduced through elaboration, as has been suggested (Conway & Pleydell-Pearce, 2000; Ehlers & Clark, 2000). The present series of experiments aims to investigate these questions by using a modified version of the Ehlers et al. (2006) paradigm.