Reducing cognitive vulnerability to depression: A preliminary investigation of MEmory Specificity Training (MEST) in inpatients with depressive symptomatology

Abstract

Reduced specificity of autobiographical memory retrieval represents an enduring vulnerability factor for depression. The present study is a preliminary evaluation of a newly developed group-based intervention program to increase specificity of memory retrieval. The MEmory Specificity Training (MEST) was administered on a weekly basis for 4 consecutive weeks to 10 inpatients with depressive symptomatology. Whereas earlier studies found that memory specificity does not improve following treatment as usual, the present results showed that participants' retrieval style became significantly more specific following MEST. These results suggest that the MEST may offer a potential and promising intervention to tackle a core cognitive process involved in depression and depressive vulnerability.

Introduction

A large body of evidence indicates that depressed individuals find it difficult to retrieve specific personal events from their memory as compared to normal controls (for a review, see Williams et al., 2007). For example, when asked to recall a specific memory in response to a cue-word, depressed people respond more often than non-depressed controls with overgeneral memories rather than with the requested specific memories. The latter refer to personal events that happened on a particular day (e.g., ‘Tuesday last month, when I went running in the forest and was attacked by a buzzard’), whereas the former summarize memories of the same type without the specification of a single particular occasion (e.g., ‘the times I go running in the forest’). This repeated observation has been labelled ‘overgeneral memory’ or ‘reduced autobiographical memory specificity’ (Williams et al., 2007).

Interestingly, this reduction in specificity for personal memories is not just a concomitant of a depressed mood state. Several studies have shown that when patients are recovered or in remission from their depression, their memory specificity does not improve (e.g., Brittlebank et al., 1993, Raes et al., 2006). Likewise, formerly depressed individuals are equally impaired in their retrieval of specific memories as currently depressed patients, and both these groups are clearly less specific with respect to their memory retrieval as compared to never depressed controls (e.g., Mackinger et al., 2000, Park et al., 2002, Williams and Dritschel, 1988). Moreover, studies also show that this phenomenon of reduced memory specificity has predictive value for the course of depression: The greater the extent to which people are characterized by reduced specificity for memories, the slower their recovery is (e.g., Brittlebank et al., 1993, Hermans et al., in press, Peeters et al., 2002, Raes et al., 2006). And even in non-clinical groups or populations that are not suffering from clinically significant pathology, reduced memory specificity predicts increased emotional (depressed) reactivity, for example, to stressful life-events (e.g., Bryant et al., 2007, Gibbs and Rude, 2004, Mackinger et al., 2000, Van Minnen et al., 2005; for a review, see Raes, Hermans, Williams, & Eelen, 2007).

At least four potential routes have been proposed via which reduced memory specificity leads, or is linked to depression. A first way through which reduced memory specificity negatively affects depression is that it hinders effective social or interpersonal problem solving. Research has demonstrated such association between reduced memory specificity and less effective problem solving. Using the Means-Ends Problem-Solving procedure (MEPS; Platt & Spivack, 1975), a task in which respondents are asked to generate solutions to hypothetical social problem situations, studies have shown that (a) the less specific depressed patients are in their memory retrieval, the less effective the solutions are they generate to such MEPS scenario's (e.g., Evans et al., 1992, Goddard et al., 1996, Goddard et al., 1997, Raes et al., 2005), that (b) manipulation of memory specificity changes problem-solving performance (Williams et al., 2006) and that (c) memory specificity moderates the effect of negative mood on problem-solving performance (Williams, Barnhofer, Crane, & Beck, 2005). The idea is that an unspecific memory base is not very helpful to a person when generating strategies for solving interpersonal problems (e.g., Evans et al., 1992). Poor problem-solving capacities in turn might lead to depressed feelings or further worsen current depression because it inevitably leads to (more) negative social encounters, and further decreases the chances for positive social reinforcement (Hermans, Raes, & Eelen, 2005).

Another, second way in which lack of memory specificity might contribute to the maintenance of depression or vulnerability for relapse is that it may increase feelings of hopelessness through its impact on imageability of the future, that is, the ability to imagine specific images of the future. Williams et al. (1996) showed that the induction of a less specific retrieval style reduces the specificity of images of one's personal future (see also de Decker, 2001). The link between feelings of hopelessness in depressed patients and an unspecific and vague view on one's future is self-evident. The idea, thus, is that memory specificity, via its negative effect on the specificity level of the image of one's own future, fuels feelings of hopelessness, and thereby installs or further exacerbates depressed feelings.

A third way in which reduced memory specificity may render people vulnerable for depression or impede recovery from it, is through a spiraling, reciprocal relationship with depressive rumination (Williams, 1996, Williams, 2004, Williams et al., 2007). Rumination could be broadly defined as repetitive and unhelpful thinking about one's own depressed or sad feelings and about related problems or past issues. Both correlational and experimental studies have provided evidence for the idea that reduced specificity in the retrieval of memories and ruminative thinking are mutually reinforcing one another, further exacerbating depressed feelings (e.g., Park et al., 2002, Raes et al., 2006, Raes et al., 2006, Ramponi et al., 2004, Watkins and Teasdale, 2001, Watkins and Teasdale, 2004, Watkins et al., 2000).

A fourth way in which reduced memory specificity might lead to (more) depression, pertains to the relative lack of experiential exposure to negative memories and the attached negative emotions (see Hermans et al., 2005). Exposure to distressing memories and feelings, although disturbing and unpleasant in the short-run, is associated with long-term benefits to psychological well-being (e.g., Littrell, 1998, Pennebaker and Seagal, 1999). This naturally occurring exposure effect, then, might be relatively absent in individuals who tend to recall past events in a less specific form, in that they might be avoiding necessary exposure that way (‘experiential avoidance’; Hayes, Wilson, Gifford, Follette, & Strosahl, 1996). In the short-run, this experiential avoidance (or lack of exposure) might be advantageous. In the long-term, however, this may proof disadvantageous and put people at risk for developing or exacerbating emotional problems, for example, depression (see e.g., Hayes et al., 1996; see also Golden, Dalgleish, & Mackintosh, 2007).

Previous research in the domain of reduced autobiographical memory specificity (and depression) has mainly focused on the study of the correlates and mechanisms of this vulnerability factor (Williams et al., 2007, for a review). However, it goes without saying that the accumulated body of knowledge in this domain could provide an important input to clinical interventions, in particular for depression. For example, based on the above described state of the art, it seems reasonable to assume that the remediation or reduction of overgeneral memory retrieval may be an important therapeutic target. Notwithstanding the fact that the existing literature clearly underscores the (clinical) importance of reduced memory specificity, this evident translation into clinical interventions has hitherto been insufficiently explored. Our current study is precisely an attempt to start filling this lacuna.

The primary aim of the study is to explore the feasibility of a newly developed group-based intervention program to increase memory specificity. More particularly, our aims are (a) to explore whether memory specificity can be increased in patients with depressive symptomatology (that is, remediation of the targeted depressogenic vulnerability factor, i.e. reduced memory specificity) and (b) to examine whether the observed gains in memory specificity are accompanied by improvements in other key variables, like rumination and problem-solving performance, that are assumed to mediate the impact of memory specificity on the course of depressive symptoms (see our review above).

As mentioned, memory specificity does not typically improve in depressed patients when offered treatment as usual (e.g., Brittlebank et al., 1993, Raes et al., 2006). Broader therapeutic programs, such as Mindfulness-Based Cognitive Therapy (Williams, Teasdale, Segal, & Soulsby, 2000) or standard cognitive behavior therapy for depression (McBride, Segal, Kennedy, & Gemar, 2007) have, however, been found to increase memory specificity.

Only one study so far has explored to what extent overgeneral memory retrieval can be remediated using a directly targeted intervention, specifically aimed at increasing memory specificity. Serrano, Latorre, Gatz, and Montanes (2004) offered a newly developed life review therapy to 20 older adults (aged 65–93) with clinically significant depressive symptoms. The life review intervention was individually administered on a weekly basis for 4 consecutive weeks. The main focus was on assisting the elderly in the retrieval of specific, albeit exclusively positive, memories. Each week the retrieval practice focused on a particular life period (childhood, adolescence, adulthood, and summary) for which participants were prompted to recall specific positive recollections by a set of 14 questions (e.g., “What is the most pleasant situation that you remember from your childhood?”, “Tell me a time that you remember experiencing the most pride at work”). It was observed that, relative to a control group (n = 23), the participants that received the intervention showed fewer depressive symptoms, less hopelessness and improved satisfaction at post-test (i.e. 2 weeks after the last intervention session). Further analyses showed that the experimental group, compared to controls, became more specific in their retrieval of positive and neutral memories, but not for negative memories. Participants who increased the most in generating specific memories were also found to improve most with respect to depressive symptoms, hopelessness, and life satisfaction. A major restriction of this study and its intervention is, however, that it exclusively focused on the retrieval of positive episodes or memories. Hence, the possibility that the observed beneficial effects were due to this focus on pleasant events cannot be ruled out. Indeed, as mentioned, patients did not improve in their retrieval of specific negative memories. This makes it difficult to reliably interpret the observed effects as being caused – in terms of underlying processes – by improvements in specificity of memory retrieval per se. Besides this important methodological concern, the exclusion of working with negative material does not really fit well with theories in the field of reduced autobiographical memory specificity, which primarily focus on the cognitive-emotional processing of negative autobiographical memories.

In order to address these and other shortcomings in the program of Serrano et al. (2004), we developed a new group-based intervention program, specifically targeted at the improvement of specificity of memory retrieval for both positive and negative materials. Additionally, the program integrates some of the most recent findings from experimental psychopathological research that are of relevance for the issue of memory specificity (e.g., Moberly and Watkins, 2006, Neumann and Philippot, 2007, Watkins and Moulds, 2005).

Our intervention is not framed in, or does not form part of a larger therapeutic program (like, for example, life review therapy, MBCT, or standard CBT). Rather, it represents a stand-alone intervention specifically aimed at increasing specificity of memory retrieval. As such, our intervention nicely fits with Barlow's (2004) recent plea that psychological interventions should preferentially be tailored to target pathological processes, as identified in (experimental) psychopathological research. Furthermore, the fact that we focus on a factor of which research clearly shows that it is implicated in the maintenance of depressive symptoms (Williams et al., 2007) is in line with Barlow's recommendation that new interventions should primarily focus on the prevention of recurrence of future depressive episodes.

To summarize then, the present study is designed as a preliminary evaluation of a newly developed group-based intervention program to reduce overgenerality of memory retrieval. The study's first aim is to see whether this program indeed reduces this known vulnerability factor, in that it increases patients' specificity of memory retrieval. Second, we will examine to what extent the observed improvements in memory specificity are accompanied by improvements in variables that are assumed to mediate the impact of memory specificity on the course of depressive symptoms (reduction of ruminative thinking, feelings of hopelessness, and experiential avoidance; and increase in problem-solving skills). Our prediction here is that participants who increase the most in the production of specific memories will improve most on these other variables. Thus, in the present pilot study, we will focus on the short-term effects of the memory training on these mediating variables. Possible longer-term effects on the course of depressive symptoms, established via changes in the mediating variables, are not examined in the present pilot study, which does not include longer-term follow-up (see Section 4). Third, we will examine whether improvements in memory specificity also generalize to improvements on other tasks than the standard AMT cue-word task. Fourth, we will check to what extent improvements in, for example, memory specificity, are not simply due to a decrease in depressive complaints.

We used patients experiencing significant current depressive symptomatology as participants in the current training study, since prior work shows that (a) this population suffers from reduced memory specificity and that (b) the extent to which such patients are characterized by reduced specificity for memories predicts slower recovery from their depression (see above).