#SfN13 Getting rid of an unwanted memory for good

Poster 99.06/JJJ40 – Gradual extinction prevents the return of fear. SJ Gershman, CE Jones, KA Norman, MH Monfils, Y NIV. Brain and Cognitive Sci., MIT, Cambridge, MA; Psychology, The Univ. of Texas at Austin, Austin, TX;Neurosci. Inst. & Dept. of Psychology, Princeton Univ., Princeton, NJ

Poster 99.07/JJJ41 Gradual extinction prevents the return of fear in humans. JW Kanen, SJ Gershman, MH Monfils, EA Phelps, Y NIV. Dept. of Psychology, Ctr. for Neural Sci., New York Univ., New York, NY;Dept. of Brain and Cognitive Sci., MIT, Cambridge, MA; Univ. of Texas, Austin, TX; Nathan S. Kline Inst. for Psychiatric Res., Orangeburg, NY;Princeton Neurosci. Inst. and Psychology Dept., Princeton Univ., Princeton, NJ

Psychiatrists have a problem. Memories, especially fearful memories, are exceedingly hard to erase. Say you’ve learned that every time you touch a doorknob in the winter you get a painful electrostatic shock; fairly soon you might form an irrational fear of the doorknob. What can the good doctor do?


Ouch! Source: ashafullife.blogspot.com

The go-to therapy is extinction training. Here, you’ll repeatedly touch a doorknob that’s been treated to eliminate static – hence, no shock. After several sessions you loose your phobia. Great. Yet a few weeks later, you once again feel butterflies fluttering in your stomach at the thought of touching a doorknob. Somehow, the fear has returned.

The above scenario may seem ridiculous; yet for those suffering from post-traumatic stress disorder or debilitating phobias, the spontaneous recovery of a fear memory is nothing to laugh at. Scientists aren’t quite sure why this happens. Erasing a memory, or memory extinction, in theory “updates” the original memory trace, such that a fearful stimulus (eg doorknob) is now encoded as safe. Yet in practice, when the object of fear suddenly dissociates from harm (eg no more shocks!), the mPFC generates a large prediction error signal, such that the new information is treated as something entirely new and encoded in a separate memory trace.

Herein lies the problem. The original fear memory, alive and well, competes with the opposing new one for expression. Behaviourally, this often results in the return of fear. But here’s the silver lining: if you keep the prediction error signal small, the brain may opt to modify the old trace rather than encode an entirely new one, thus mitigating or erasing fear in actuality.

Researchers decided to test this theory out. First, they taught a cohort of rats to fear a tone by associating it with a shock. The rats were subsequently divided into three groups: the first received gradual extinction, in which the frequency of shock delivery declined gradually. In other words, as the trials progressed, rats more often than not experienced the tone without the shock. In a sense, researchers “weaned” these rats off the tone-shock association, thus triggering a small prediction error. The second group received the opposite treatment, with the frequency of shock delivery steadily increasing until the last 9 trials, in which all shocks were omitted to facilitate extinction. This is necessary as researchers wanted to observe the return of fear. The third group went through normal extinction; that it, only tones, no shocks. By the end, all rats lost their fear of the tone.

Fast-forward a month. Researchers returned the rats to the testing chamber and played the same tone. Rats that underwent gradual extinction showed significantly less fearful behaviour than the other two groups. In another experiment, researchers re-established the fear memory by giving the rats another shock. Once again, those that had undergone gradual extinction showed less fear. These results strongly suggest that gradual extinction is especially effective at persistently decreasing fear. But how do these results transfer to humans?

Using a similar fear-conditioning paradigm, researchers presented images of snakes to a group of volunteers. This was followed by an uncomfortable electrical shock to the wrist. Fear, as well as other states of arousal, increases sweating and skin conductance, the latter of which was carefully monitored to objectively measure the volunteers’ level of fear. Within a day, volunteers learned to fear the innocuous snake image.

One day later, researchers divided volunteers into groups and eliminated their fear memory through one of the three extinction processes as above. When tested on the third day, those who went through gradual extinction showed a trend towards less spontaneous recovery; that is, they sweated less at the sight of the snake images. However, the results are preliminary and due to the small number size (10-14 per group), the effect was not yet statistically significant.

Clarifying the conditions that facilitate persistent fear extinction may help clinical psychiatrists optimize extinction-based exposure therapies for the treatment of anxiety disorders and phobias. The evidence presented here – from rat to human – strongly suggest that minimizing prediction error through gradual extinction is a more effective way to modify and erase a memory, maybe for good.

Sometimes slow is a better way to go.


3 thoughts on “#SfN13 Getting rid of an unwanted memory for good

  1. This is interesting if you approach it from a human relationship perspective. As we know, people who experienced high levels of developmental fear within and as a result of their early relationships mature in to adults who have great difficulty maintaining healthy relationships and bring a lot of conflict in to these relationships. I would hypothesise that with their often serial approach to relationships and if those relationships themselves come with little danger, that gradually the implicit/intrinsic response to those relationships as a “threat to survival” might be lessened over time?

    • Very interesting point. I was thinking of the clinical implications of this study while writing it up (aka comparing it to exposure therapy and such), but more in terms of phobias. In terms of relationships, I suppose if each one is somewhat similar (ie if their SOs have similar triggers that act as a CS), but the triggers are paired less with the US (perhaps the negative response to those triggers) over time, according to this study it may facilitate removal of that negative response? Of course this is highly speculative: first the study will have to be replicated in humans, and second we’ll have to see if it applies to all types of US-CS pairings. And of course, it would always be nice to know what is happening at the circuit level…

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s