by David Dobbs
Editor, Mind Matters

Bad memories can seem to have their own power, as if they are independent agents infecting our thoughts and moods. Yet memories are creations of mind and brain, and the organ that makes them can also temper their power. A new study, reviewed here by Daniel Weissman and Clare Porter of the University of Michigan, sheds light on how we accomplish this vital and welcome feat.


Putting a Lid on Bad Memories:The Mechanics of Memory Suppression

by Clare Porter & Daniel Weissman
University of Michigan, Ann Arbor

Memories provide us a personal history and a sense of identity. There are times, however, when we'd like to forget a social blunder or other embarrassing incident -- or in some cases, a memory so traumatic that it is painful to recall. Soldiers who have experienced horrific events may suffer from post-traumatic stress disorder (PTSD), emotional distress stemming from an inability to stop recalling traumatic events. Our comfort, and sometimes our mental health, can depend on suppressing such memories. How do our brains manage this task? The Past in Many Parts An emotional memory has many components. For example, the memory of a car accident might be associated with the sound of tires squealing, the sight of two cars colliding, the smell of gasoline, and feelings of fear and panic that build as the accident unfolds. One might imagine that suppressing such a memory would require suppressing each of the individual components. Brendan Depue, Tim Curran and Marie Banich, all of the University of Colorado, explored this hypothesis in their study entitled "Prefrontal Regions Orchestrate Suppression of Emotional Memories via a Two-Phase Process." Human participants were trained to associate each of several female faces with a distinct photograph of an emotionally distressing event (such as a car crash). Later, they were shown each of the faces in turn and asked either to think or to not think about the associated photograph. While participants were performing this task, the authors used functional magnetic resonance imaging (fMRI) to measure their brain activity. (Functional MRI reveals where blood flows in the brain when a stimulus is presented, thereby indirectly indicating which regions become active). After the fMRI scan was completed, participants were given a memory test in which they were shown each face and asked to describe the photograph it had been paired with. Block That Memory The results suggested that participants did indeed suppress the face-photograph pairings that they were told not to think about during the fMRI scan. Specifically, when they were tested later, the participants remembered these pairings least accurately. They remembered with intermediate accuracy pairings they had been trained to associate but that were not shown during the fMRI scan. And they remembered most accurately pairings they had been asked to think about in the scanner. Consciously trying not to think about the association between a face and a photograph, then, seemed to weaken the association. With those behavioral indications of memory suppression in hand, the researchers examined the fMRI data to determine which parts of the brain were involved in squelching the associations. The fMRI data showed that a number of brain regions participate in suppressing emotional memories. First, there was increased activity in multiple regions of the frontal lobes. The frontal lobes are thought to underlie higher cognitive abilities --like attention, memory and inhibition -- that require coordination with other brain regions to be carried out. One of the activated frontal regions, the right dorsolateral prefrontal cortex, is thought to control the retrieval of information from memory. A second, the right inferior frontal gyrus, seems to play an important role in inhibition. A third, the frontopolar cortex, is relatively unexplored but may help to coordinate the activity of other brain regions. These findings appear to extend the established role of frontal lobe structures in higher cognitive abilities to suppressing emotional memories. The researchers also found that suppressing emotional memories reduces activity in brain regions that process sensory information (the thalamus and the visual cortex) and emotional/memory information (the amygdala and the hippocampus). This reduction in activity was measured relative to activity when subjects fixated on a dot in the middle of a blank screen -- an important baseline not included in previous studies. These results agree nicely with the conceptual picture of active suppression: when we are told to inhibit a memory, our brains shut down emotional and sensory processing that relates to the memory. A Timing Thing Given that frontal regions frequently coordinate other brain regions to accomplish a goal, frontal regions activated during memory suppression might be inhibiting activity in emotional and sensory regions in order to suppress these distinct components of emotional memories. How might one measure such an inhibitory effect? When brain regions communicate, their activity becomes correlated. Thus if one region is working to curb another, then whenever the first region's activity increases, the second's should decrease. Consistent with such inhibition, the authors found that greater activity in the right inferior frontal gyrus was associated with reduced activity in the visual cortex and the thalamus, which process sensory information. Furthermore, greater activity in the right middle frontal gyrus was associated with reduced activity in the hippocampus and the amygdala, which underlie memory and emotion. Finally, these restricting effects occurred earlier in the sensory regions than in the emotional/memory regions, suggesting that emotional memories are suppressed via a two-phase process -- first the relevant sensory information, then the emotional reaction to it. How is the timing of these two phases controlled? The authors found evidence to suggest that the frontopolar cortex boosted activity in the right inferior frontal gyrus just before the right inferior frontal gyrus suppressed activity in sensory regions. Slightly later, the frontopolar cortex boosted activity in the right middle frontal gyrus just before the right middle frontal gyrus suppressed activity in emotional/memory regions. Thus, the frontopolar cortex appears to control the order in which different components of emotional memories are suppressed. This study by Depue and colleagues raises several fascinating questions. First, might the sequence in which different components of emotional memories are suppressed change in various clinical disorders? For example, might patients with post-traumatic stress syndrome (PTSD) suppress the emotional aspects of traumatic memories earlier than healthy participants because their memories are so disturbing? Would treatment for PTSD be more effective before or after the emotional aspects of traumatic memories are suppressed? And how much specialization is there? Given that distinct subregions of the sensory cortex process different types of sensory information, can we suppress some sensory components of an emotional memory more easily than others -- visions or sound, for instance, more easily than odor? Finally, since different brain regions process emotional as opposed to nonemotional information, are different brain mechanisms involved in suppressing emotional versus nonemotional memories? The questions framed by this study open the way to a much more nuanced picture of how we contain and temper the power of the past.

Clare Porter is an undergraduate and Daniel Weissman an assistant professor in the psychology department at the University of Michigan, where Weissman heads the Attention and Cognitive Control Lab. -- Edited by David Dobbs at 11/30/2007 11:33 AM -- Edited by David Dobbs at 11/30/2007 11:35 AM -- Edited by David Dobbs at 11/30/2007 11:36 AM -- Edited by David Dobbs at 11/30/2007 11:41 AM