Cognitive processing, memory, and the development of PTSD symptoms: two experimental analogue studies

Abstract

Memory deficits are implicated in the development of posttraumatic stress disorder (PTSD). Intentional recall of trauma memories is frequently disorganised or incomplete, whilst involuntary memory fragments are easily triggered by perceptual cues. Ehlers and Clark (Behaviour Research and Therapy 38 (2000) 319–345) propose that a predominance of data-driven processing (i.e., processing sensory impressions) during the trauma contributes to the development of this memory pattern, and therefore, predicts PTSD symptoms after trauma. Two experimental studies examined these hypotheses. Student volunteers viewed a distressing videotape as an analogue for a traumatic event. In Study 1, cognitive processing was manipulated; in Study 2, extreme scorers on a processing screening questionnaire were pre-selected. The results indicated that data-driven processing is associated with the development of PTSD-like memories and analogue symptoms.

Introduction

Posttraumatic stress disorder (PTSD) is characterised by disturbances of memory. The hallmark symptoms of PTSD are involuntary distressing and uncontrollable memories of the precipitating trauma. These unusually vivid memories are experienced in a variety of sensory modalities and generally constitute segments of the episode rather than the whole event. They are often triggered by perceptual cues that resemble those present at the time of the trauma (Ehlers & Clark, 2000; Ehlers et al., 2002; van der Kolk & Fisler, 1995). Conversely, individuals with PTSD may simultaneously experience problems in intentionally recalling the traumatic event (e.g., Tromp, Koss, Figueredo, & Tharan, 1995; van der Kolk & Fisler, 1995). The order of events during the trauma is often sequentially confused, and there may be amnesia for a significant part or parts of the experience.¹

How does the pattern of poor intentional recall and easy triggering of involuntary memories in PTSD develop? Many trauma theorists have proposed that the overwhelming nature of traumatic experiences prevents individuals from fully processing them at the time (e.g., Brewin, Dalgleish, & Joseph, 1996; Ehlers & Clark, 2000; Foa & Hearst-Ikeda, 1996; Horowitz, 1976; Siegel, 1995; van der Kolk & Ducey, 1989). In one such theoretical approach, Ehlers and Clark (2000) applied experimental psychology research on the relationship between information processing during encoding and the quality of subsequent memories to PTSD. The authors suggested that the memory pattern in PTSD may, in part, be the result of a predominance of data-driven processing during trauma (i.e., processing sensory impressions and perceptual characteristics rather than the meaning of the event) as opposed to conceptual processing (i.e., processing the meaning of the event, processing it in an organised way and placing it in context). Individuals who mainly engage in data-driven processing during trauma are expected to be at greater risk of developing PTSD symptoms than those who elaborate the contextual and meaning elements of the event (see also Siegel, 1995). This hypothesis is based on research indicating that (1) deeper, conceptually based processing results in elaborative encoding in memory and is required for intentional recall (Schacter, 1996), (2) data-driven processing results in a poorly elaborated, perceptually encoded memory trace that can be involuntarily retrieved by matching perceptual stimuli (Roediger, 1990), retrieval of information from memory being facilitated if retrieval conditions match those present during encoding (transfer-appropriate processing, Brown & Craig, 2000), and (3) elaboration of autobiographical memories inhibits unintentional retrieval (Conway & Pleydell-Pearce, 2000). Correlational evidence for the role of data-driven processing in the development of PTSD was found in three prospective studies of motor vehicle accident victims (Ehlers, Mayou, & Bryant, in press b; Murray, Ehlers, & Mayou, 2002; Rosario, Williams, & Ehlers, 2002), and a prospective study of assault survivors (Halligan, Michael, Clark, & Ehlers, 2002). The present studies sought experimental evidence for the role of data-driven processing in subsequent poor recall and PTSD symptoms.

A further goal of the present studies was to specifically investigate whether poor intentional recall directly predicts the development of PTSD. As noted above, work by Conway (Conway & Pleydell-Pearce, 2000) suggests that one of the functions of elaboration of autobiographical memories is to inhibit unintentional retrieval of memories. Thus, insufficient elaboration will render the memory trace more vulnerable to triggering by matching sensory cues, increasing the frequency of PTSD symptoms such as intrusions and arousal (Ehlers & Clark, 2000). Preliminary research has provided correlational evidence for a relationship between poor intentional recall of the trauma and PTSD symptoms. Studies have found that trauma memory disorganisation predicts subsequent symptoms of acute stress disorder and PTSD following road traffic accidents (RTAs) (Harvey & Bryant, 1999; Murray et al., 2002), and assault (Amir, Stafford, Freshman, & Foa, 1998; Halligan et al., 2002). Foa, Molnar, and Cashman (1995) reported that recovery during reliving therapy in rape victims was associated with the formation of more coherent narratives for the event.

The present paper describes two analogue studies investigating the hypothesised relationships between data-driven processing, intentional recall, and PTSD-like symptoms following exposure to distressing material. Study 1 attempted to experimentally manipulate processing during exposure to the analogue traumatic stressor, whereas Study 2 examined the effects of natural variations of this dimension by selecting extreme groups.