A surprise makes memories wobbly

An element of surprise may be the key to whitewashing a painful memory. People who encountered something unexpected were better able to shake a troubling association, a new laboratory study finds. The results, published in the Feb. 15 Science, bring scientists closer to being able to weaken traumatic memories with help from a drug.

Understanding how the brain forms and reforms traumatic memories might lead to treatments that would help people who suffer from post-traumatic stress disorder and other anxiety disorders. “The idea that an original memory could have the sting taken out of it — that’s been very appealing,” says psychiatrist Roger Pitman of Harvard Medical School and Massachusetts General Hospital, who was not involved in the research.

Memories are not written in neural stone. Recent results in animals and humans have shown that once called to mind, painful memories’ emotional edges can be blunted. Experiments have used certain drugs to weaken associations between a memory and a negative response. But the details of how and why those drugs work haven’t been clear.  

The new result may have uncovered a previously underappreciated step in that weakening process: In order for the emotional response tied to a memory to wither, something unexpected must happen while the person is recalling the memory. This mismatch between what a person expects and what actually happens — called a prediction error — puts a memory into a wobbly, vulnerable form that can be washed out, says study coauthor Merel Kindt of the University of Amsterdam.

In the experiment, people learned that an ominous picture of either a gun or a spider came along with a loud noise and an unpleasant shock. This taught the people to fear one of the pictures — either the gun or the spider — more than the other. The next day, some people reexperienced the same connection (a spider picture always came with a noise and shock while a gun picture did not, for instance) while other people dealt with curve balls (a spider picture sometimes came with a noise and shock, but other times didn’t). Electrodes placed underneath the eye picked up people’s flinches, allowing the researchers to gauge involuntary fear responses.

After the second day’s training, every participant received propranolol, a drug that interferes with memory formation. The drug didn’t affect the people who relived the original experience, the team found. But for people who experienced something new, the drug wiped out the response to the original memory. These people reported that they were less likely to flinch in anticipation of a shock that, from the first day of training, they should have expected was coming. For these people, Kindt says, “it is not just a reduction of the fear response. The fear response is erased.”

“The results are very convincing in showing that prediction error is important,” psychologist Thomas Ågren of the Uppsala University in Sweden says. “However, to the claim that it is necessary, I remain humbly skeptical.” More experiments with different designs must be done, he says, to conclude that prediction error is always required to weaken memory.

Pitman cautions that it’s too soon to say whether the results will prove useful in treating people with anxiety disorders such as PTSD, in which traumatic memories become debilitating. “This is pretty basic research,” he cautions. “There’s an enormous leap” to the rich, complex memories that people form in their lives.