Skip to Main Content

Our memories are immensely powerful. For a person with alcohol use disorder, a memory triggered by a simple cue — like walking by a favorite bar or spotting a beer billboard — can drive a desire for a drink.

But they’re also surprisingly pliable. And scientists are trying to curb harmful drinking by dredging up memories and rewriting them — with the help of a dose of ketamine, a longtime anesthetic which is also used recreationally and to treat certain mental health conditions.


The idea hinges on the rewriting of the memories associated with drinking. Those memories are often associated with environmental cues, like certain friends a person with alcohol use disorder always drank with. By manipulating those memories in a moment when they’re malleable, researchers hope to break the tie between memories and environmental cues — and in turn, prevent relapse. In a small, preliminary new study published Tuesday in Nature Communications, the approach was associated with less alcohol consumption and less of an urge to drink among people who drank heavily and often.

“We know these cues cause relapse. Could you disrupt the memory associated between the cues and the [alcohol], and stop the relapse from happening?” said Amy Milton, a University of Cambridge researcher who has studied memory retrieval in addiction.

“It’s a really, really simple question. Practically, it’s a bit tricky to do,” added Milton, who wasn’t involved in the research.


When you make a memory, it’s stored within a few hours. But our brains are constantly retrieving those stored memories — they’re what tells us how to open a Google Doc or which way to turn to get to the grocery store.

Your brain uses those memories to make predictions. On your route to the grocery store, your brain knows to turn right at a certain intersection. The road will be open, you’ll drive 2 miles, and arrive at the store. If everything goes according to plan and there’s no new information for your brain to absorb, the memory gets reshelved for the next time it’s needed.

But what if something unexpected happens — like a road closed for construction?

“That means the memories you were using probably weren’t correct and need updating,” said Milton. Experts think that unexpected surprise — and the subsequent need to update — makes a memory temporarily unstable.

The researchers wanted to swoop in during that period of instability and manipulate the memories associated with alcohol.

But to do that, they first needed to bring the relevant memories to the forefront of participants’ minds. They did so with a process called memory retrieval — essentially, cueing the brain to pull a memory back up. It’s similar to the process used in existing mental health treatments, like cognitive behavioral therapy or exposure therapy.

In the new study, researchers tested their approach on a group of 90 participants who hadn’t formally been diagnosed with alcohol use disorder, but drank at high and harmful levels.

On the first day, researchers handed the participants a glass of beer and told them they could drink it after looking at a series of images. Those images — pictures of beer — and the beer itself were designed to be memory retrieval cues. After rating the images and their own desire to drink the beer, the participants were allowed to drink it.

The researchers ran through the whole rigamarole gain two days later. But the second time around, participants weren’t allowed to drink the beer — a surprise twist designed to make their alcohol-associated memories malleable.

Then, the researchers quickly gave the participants either an infusion of saline or an infusion of ketamine. Another group of patients received a dose of ketamine without undergoing memory retrieval. The drug is known to act on the brain’s NMDA receptor. The receptor plays a key part in the process of restabilizing memories — also known as “reconsolidation” — after they’ve been updated.

“We’re leveraging that period when the memory is unstable and trying to get in there with a drug like ketamine to prevent it from restabilizing,” said Ravi Das, a psychopharmacology researcher at University College London and a co-author of the new study.

The preliminary results suggest it worked. The participants didn’t stop drinking completely, and objectively drank more than recommended. But about a week after the second part of the experiment, people in the ketamine and memory retrieval group had less of an urge to drink, drank less, and drank less often than their peers in other groups. Nine months out, those effects held up. All three groups had decreased their drinking by that time, but the decrease from initial drinking levels was most significant among people in the ketamine and memory retrieval group.

“It’s a really exciting development. This is the first time it’s been shown in a clinical population that this can be effective,” said Milton.

But the study comes with several notable caveats. It was a small study and was only single-blinded. The participants hadn’t been clinically diagnosed with alcohol use disorder and their behavior was self-reported. They were recruited through online ads, which outside experts said introduces the possibility that the participants had already considered reducing the amount they drank and would have done so without any kind of intervention.

There’s also a question of why drinking fell across all three groups — and what that might signal about how well the memory retrieval and ketamine combination approach works in the long run, said Mary Torregrossa, a University of Pittsburgh neuroscientist who has studied drug-associated memories.

“This is a good first study that lends support to doing more trials and similar trials in different populations,” said Torregrossa, who wasn’t involved in the research. Torregrossa said the approach could also be studied for other substance use disorders.

There are still big questions about the best way to dig up memories and disrupt them. What works well to retrieve a memory for one person — like showing a picture of people laughing in a bar — might not be the best approach to trigger memories for another. Das, the study’s co-author, said he hoped one day retrieval methods could be somewhat personalized to the patient.

And because mucking up memories is a “silent process,” there isn’t a biomarker for scientists to measure how well they’ve done. For now, with experiments such as the new study, they have to rely on behavioral cues, like how much less a person is drinking after reconsolidation.

It’s also still not clear how many memories need to be disrupted for such an approach to work.

“Real memories have been formed over years and hundreds of different contexts,” said Das.

People with alcohol use disorder would, presumably, have a string of memories associated with drinking. Does manipulating one retrieved memory have a sweeping effect?

“We are still getting a handle on how far it spreads if you did this kind of memory manipulation,” said Milton, the University of Cambridge researcher.

  • @kdn

    I agree with your concerns about using ketamine. I will suggest that an addiction is different from a bad habit. A bad habit involves a response that reduces unease, i.e. we feel better, but it increases difficulty, i.e. the response increases demands or reduces resources. The use of a substance can be a bad habit. An addiction occurs when the absence of the substance, or the behavior, increases unease that is only reduced by the substance or the behavior. Treatments that are effective for a bad habit will not work for an addiction. A paper on unease and its role in learning, stress and response can be found in Frontiers in Psychiatry.

  • Considering that the brain is very complex organ (with billions of neurons and trillions of synapses that connect and interact in complex ways) I wonder how much damage this type of administering of a drug like ketamine does to an individual’s normal day to day functioning. Bad habits can be broken by using simple techniques – I suggest watching the TED talk “A simple way to break a bad habit.”

  • A whole lot of caveats chief of which is that they are using a drug of abuse. The same kind of pharmaceutical reconsolidation of memory can be done with both propranolol and hydrocortisone. Far less risky from my perspective.

Comments are closed.