Reducing intrusive memories of real-world stimuli via memory reconsolidation

Abstract

After a distressing event, intrusive memories often persist and, for some, become pathological and debilitating (e.g., Brewin et al., 2010). Methods that enhance extinction learning may translate to improved exposure-based interventions that target intrusive memories. One possible opportunity for enhancing extinction is through memory reconsolidation (Nader, Schafe, & LeDoux, 2000; Monfils, Cowsanage, Klann, & LeDoux, 2009; Schiller et al., 2010). A retrieved memory reactivated by conditioned stimulus (CS) presentation is thought to enter a labile state as proteins are synthesized, and the effects of new learning that occurs within the reconsolidation window (about 10 min to 6 hrs post-retrieval) is more robust (e.g., Nader et al., 2000). To date, memory reconsolidation reseach in humans has been limited by fear learning paradigms that lack ecological validity (e.g., Elsey & Kindt, 2017), and parameters of boundary conditions (e.g., memory strength, retrieval cue specificity, prediction error) remain unclear (e.g., Treanor, Brown, Rissman, & Craske, 2017). In a two-study sequence, both behavioral and biological mechanisms underlying memory reconsolidation were examined, first in a non-clinical sample, and then in a sample of trauma-exposed individuals with and without current trauma-related intrusive memories. We used the film fear learning paradigm in order to elicit and then reduce film-related intrusive memories. Neutral and negative cues were used to explore differences in cue valence, given that previously, a negative CS retrieval cue elicited higher distress and more intrusive memories than non-retrieval conditions (Marks & Zoellner, 2014). Timing of cues were varied to examine any enhanced effects of extinction within the reconsolidation window. In Study 1, participants (N = 173) were randomized to one of four CS cueing conditions: Pre Neutral CS, Pre Negative CS, or Pre Scrambled cue, presented 10 min prior to extinction, or Delayed Neutral CS presented 10 min after extinction. Intrusive memories were assessed 24 hr and 72 hr after acquisition. There were no differences in intrusive memory frequency or distress 72 hr after acquisition between participants in the Pre Neutral and Pre Negative cue conditions, nor were there differences between the Pre Neutral and Pre Delayed conditions. Larger increases in sAA during acquisition, b = .23, and larger increases in cortisol and sAA together, b = .25, during acquisition predicted higher intrusive memory frequency 72 hr after acquisition. Larger cortisol increase, b = .28, and sAA increase, b = .25, during extinction also predicted intrusive memories 72 hr after acquisition, and a larger increase in sAA, b = .27, also predicted higher intrusive memory distress 72 hr after acquisition. Negative affect after acquisition predicted intrusive memory frequency and distress 72 hr after acquisition, b = .35 and b = .44 respectively. Boundary conditions of reconsolidation as they relate to more ecologically valid stimuli and intrusive memories remain elusive. Study 2 sought to extend this work to a clinical sample, characterized by persistent intrusive memories, and to better understand the specific type of new learning during extinction that may be required to initiate reconsolidation. Importantly, intrusive memories are a transdiagnostic construct present in a range of psychopathology (e.g., Brewin et al., 2010). Participants (N = 14) in the PTSD/MDD (n = 11) and control (n = 3) groups were randomized to one of three extinction conditions: an image extinction condition, where a brief 20 sec film segment that preceded the analogue trauma during acquisition is presented repeatedly in the absence of the analogue trauma, and a film extinction condition, where the acquisition segment is shown repeatedly, and an assessment only control condition, where participants do not engage in any kind of extinction procedure. All data from this study is preliminary. Patterns of intrusive memories 72 hr after acquisition suggest that, though intrusive memory frequency did not decrease d = 0.08, related distress did decrease, d = 0.85. Participants in the PTSD/MDD group reported more intrusive memories than the control group both 24 hr (d = 1.12) and 72 hr (d = 0.54) after acquisition. Intrusive memory frequency decreased in the assessment only (d = 0.89) but not in the extinction conditions 72 hr after acquisition (d = 0.07), but patterns of distress reduction from 24 to 72 hr post-acquisition appeared similar across conditions. Parameters of reconsolidation boundary conditions when more complex, ecologically valid stimuli and outcome measures are used remain unclear; neither cue valence nor timing of retrieval cue affected intrusive memories after extinction. Glucocorticoid and noradrenergic system activity predicted intrusive memories, illustrating the importance of these two systems in strengthening emotional memory. As efforts to push reconsolidation toward clinical settings continue, preliminary findings from Study 2 highlight the importance of capturing distressing and persistent intrusive memories and determining whether these intrusive memories are amenable to enhanced extinction, as these are the kinds of intrusive re-experiencing representative of psychopathology that are often missed in experimental paradigms.