The effects of imagery rescripting on memory outcomes in social anxiety disorder

Highlights

• Examined effects of imagery rescripting (IR) on memory outcomes in social anxiety disorder.

• IR was compared to imaginal exposure (IE) and supportive counselling.

• IR led to increases in positive memory details over time.

• IE led to increases in both positive and negative memory details.

• IR also facilitated the expression of updated core beliefs.

Abstract

Imagery rescripting (IR) is an effective intervention for social anxiety disorder (SAD) that targets negative autobiographical memories. IR has been theorized to work through various memory mechanisms, including modifying the content of negative memory representations, changing memory appraisals, and improving negative schema or core beliefs about self and others. However, no prior studies have investigated the unique effects of rescripting itself relative to other IR intervention components on these proposed mechanisms. In this preliminary study, 33 individuals with SAD were randomized to receive a single session of IR, imaginal exposure (IE), or supportive counselling (SC). Memory outcomes were assessed at 1- and 2-weeks post-intervention and at 3-months follow-up. Results demonstrated that the content of participants’ autobiographical memory representations changed in distinct ways across the three conditions. Whereas IR facilitated increases only in positive/neutral memory details, IE facilitated increases in both positive/neutral and negative memory details and SC facilitated no changes in memory details. Although memory appraisals did not differ across conditions, participants who received IR were more likely to update their negative memory-derived core beliefs. These unique effects of rescripting on memory representations and core beliefs enhance our understanding of the memory-based mechanisms of IR within the context of exposure-based learning for people with SAD.

Introduction

Cognitive models of social anxiety disorder (SAD) posit that negative self-imagery is central to the maintenance of the disorder (Clark & Wells, 1995; Rapee & Heimberg, 1997). Negative self-images often originate as a result of significant, socially painful autobiographical experiences (Hackmann, Clark, & McManus, 2000; Moscovitch et al., 2018), tying them thematically to negative core beliefs derived from these earlier ‘social failures’ (Çili & Stopa, 2015). Persistent experiences of negatively distorted self-imagery may therefore prevent socially anxious individuals from updating negative schema in the face of disconfirming evidence (Hirsch, Clark, & Mathews, 2006; Ng, Abbott, & Hunt, 2014; Wild & Clark, 2011). As such, therapeutic interventions that work by harnessing these affectively-charged memory-derived images may be particularly fruitful for targeting social anxiety symptoms (e.g., Iyadurai et al., 2018; McEvoy, Erceg-Hurn, Saulsman, & Thibodeau, 2015).

Imagery rescripting (IR) is a therapeutic technique that aims to modify negative mental self-representations (Holmes, Arntz, & Smucker, 2007; Morina, Lancee, & Arntz, 2017). In IR, which can be delivered within a single therapy session during the course of CBT, patients are guided to re-imagine past negative experiences in order to meet the needs of the younger self within the memory (Arntz & Weertman, 1999; Arntz, 2012). Patients’ rescripting of the memory typically involves actively imagining and guiding their younger selves to behave in ways they wished they could have at the time of the event, or having their older selves protect, nurture, or stand up for their younger selves in their moments of need. These interventions aim to make the event more positive or satisfying for the younger self such that new information is incorporated into the scene that can help to change the meaning of the memory. For example, by having the younger self stand-up to a critical other or receive compassion from the older self, the patient may observe that they did not deserve the harsh treatment they received at the time. By allowing the patient to change the memory in any way that they feel is right, IR can induce new perspectives on what happened at the time of the event (Arntz, 2011; Edwards, 2007).

Prior research has shown that IR delivered as a brief stand-alone intervention over the course of one or more sessions, either with or without cognitive restructuring, reduces social anxiety symptoms as well as the emotional impact and salience of negative autobiographical memories and associated memory-derived negative core beliefs (e.g., Frets, Kevenaar, & Heiden, 2014; Lee & Kwon, 2013; Nilsson, Lundh, & Viborg, 2012; Norton & Abbott, 2016; Reimer & Moscovitch, 2015; Wild, Hackmann, & Clark, 2007; Wild, Hackmann, & Clark, 2008). Though still empirically untested, a number of hypotheses have been proposed to account for the effects of IR, which correspond with differing views on what happens to the episodic memory itself as a result of treatment. Arntz has suggested that IR may modify the fear memory directly so that it becomes reconsolidated with a different meaning (Arntz & Weertman, 1999; Arntz, 2011). This account does not imply that IR erases the original memory representation, but rather that the meaning of the experience – particularly the automatic emotional processes associated with the memory – changes so that the original emotional memory no longer elicits the original emotional response. It is also possible that IR creates an alternative memory (or schema) that competes with the original for preferential retrieval (e.g., Brewin, 2006; Craske, Treanor, Conway, Zbozinek, & Vervliet, 2014).

At present, it is unknown whether the actual content of the rescripted autobiographical memory changes as a result of IR, or whether it is only its appraised meaning that changes. If memory content and/or meaning change during IR, is rescripting primarily responsible for this, or could similar changes occur by processing the memory in other ways, without explicit attempts to modify or rescript the memory itself? For example, imaginal exposure (IE) guides patients to relive a traumatic event repeatedly or over a prolonged period of time in as much sensory and emotional detail as possible (Foa & Rothbaum, 1998). Doing so is thought to facilitate emotional processing by activating the fear memory and enabling corrective information to be incorporated (Foa & Kozak, 1985, 1986; Lang, 1977). Research on IE for PTSD has shown that it promotes habituation to the fear response associated with the memory, reappraisal of the memory, and generation of a more coherent memory narrative (Foa & Cahill, 2001; Foa, 2011). Given that there are a number of similarities between Criterion A traumatic events and the socially painful or traumatic experiences described by socially anxious individuals (Carleton, Peluso, Collimore, & Asmundson, 2011; Erwin, Heimberg, Marx, & Franklin, 2006), IE may be similarly effective for modifying negative memories in SAD even in the absence of explicit rescripting (see Foa, Huppert, & Cahill, 2006).

The current study aimed to isolate the unique effects of rescripting on memory outcomes by examining the effects of single-session IR alongside two single session control conditions: (a) IE, in which patients were exposed deliberately and repeatedly to the negative memory content in their imagination without rescripting, and (b) supportive counselling (SC), in which patients were exposed to non-specific therapeutic components without repeated exposure to the memory narrative or rescripting. Though these brief interventions represent standalone protocols, they are best conceptualized as components of therapy rather than comprehensive treatments. As the modification of memory is thought to be central to the salutary benefits of IR, we were particularly interested in examining intervention-related changes in memory content, memory appraisals, and memory-derived schema (core beliefs). Given the relative brevity of the interventions and small sample size, we have conceptualized the present study as a preliminary investigation that would require replication and extension in future research.

We advanced several specific hypotheses. First, given that IR explicitly aims to rescript the content of episodic memory representations, we predicted that the content of participants’ autobiographical memory narratives would change uniquely in the IR condition such that positive and neutral memory details would increase over time, but that such enhancements in positive and neutral details would not occur in the other two conditions. Second, we anticipated that rescripting would facilitate significant changes in memory appraisals. Because Arntz has suggested that changes in meaning may be represented by changes in automatic emotional processes associated with the memory (Arntz & Weertman, 1999; Arntz, 2011), we operationalized memory appraisals as the amount of negative affect experienced upon memory retrieval as well as subjective ratings of memory intrusiveness and vividness. We expected that negative affect, intrusiveness, and vividness would decrease significantly over time for IR participants. Further, we expected that similar changes in memory appraisals would occur for those in the IE condition due to the effects of repeated imaginal exposure to the memory, but not for those in the SC condition in the absence of rescripting or exposure. Third, since IR is designed to facilitate new perspectives on the event, we hypothesized that rescripting would promote significant updating of participants’ negative core beliefs about the self and others, whereas the same would not be true of either IE or SC.