Scientists Are Testing a New Way to Treat Phobias

Exposure therapy is a highly effective way to treat certain kinds of anxiety, specific phobias, and PTSD. It puts you face-to-face with the very objects or situations you fear most in the world, over and over, until they become less scary. This can happen gradually, with graded exposures that build up to confronting a phobia, by experiencing the worst of your fears all at once (a technique called "flooding"), or through systematic desensitization, when exposures are paired with relaxation techniques.

I went through graded exposures for two specific phobias related to my OCD: food poisoning and vomit. At the beginning, it was slightly manageable. I looked at images and words that reminded me of my fears and were triggering obsessive thought spirals and rituals in my everyday life. Things like spilled orange juice, splatters on the sidewalk, words like upchuck, chunk, puke, and barf, and even the throwing up emoji. It may not sound like it, but this was extremely difficult and laborious for me. After awhile though, I was better able to deal with these things, and while it’s still hard, I handle them a lot better than I did before.

Exposure therapy is thought to work in a few ways, one being habituation. The body can’t be in a state of extreme fear forever, and over time, the reaction to a phobia will decrease. Another is that facing a fear head-on in therapy, and seeing that nothing bad will happen, can change your association with that fear—especially if you’re scared of something because of a prior experience or trauma. A person can let go of their fear response by developing a new memory or belief that competes with the old one.

But exposure therapy, as helpful as it is, still sucks. Or, put more eloquently: “There’s an upper limit on its effectiveness,” Michelle Crankse, a clinical psychologist and anxiety treatment researcher at UCLA, tells me. "One of the barriers is that for a lot of people, it’s difficult to do. It takes a lot of bravery and courage to face these situations.”

Once my exposures started advancing–looking at actual vomit, spinning around in a chair until I felt sick, eating food off the floor, eating food close to its expiration date, reading in the car–I descended into a phobic blob and ran back to talk therapy. The exposure therapy dropout rates reflect my experience. For various anxiety and trauma disorders, they are up to 52 percent, and have been reported as high as 70 percent.

It's frustrating that a technique that works so well is so hard. What if there was a way you could expose yourself to what you’re scared of, alleviating the fear, without being consciously aware of it?

This is a concept Hakwan Lau, a cognitive neuroscientist at UCLA and the University of Hong Kong, has been toying with in the past several years. In a study from 2016 in Nature Human Behavior, he and others showed that after they conditioned people to be afraid of specific visual stimuli, they could lessen their fears later by pairing the same stimuli with a reward. In a new study in PNAS, Lau and his collaborators have taken the idea one step further: seeing if they could reduce a person’s already-existing phobia, without them even knowing it.

Vincent Taschereau-Dumouchel, the lead author on the new paper, said the idea was to provide a reward for when a person’s brain was thinking about their phobia, even when they didn’t know they were thinking about it. How could they know what a person was thinking? Each time you see something, a pattern of brain activity forms, he says. They put people in an fMRI and looked for the brain activity that corresponded with their fears. Then, they rewarded people for activating that specific pattern, or patterns that were close to it.

The person who came up with the pattern recognition technique they used is Jim Haxby, a computational cognitive neuroscientist at Dartmouth. He's not involved in this current work, and he tells me that he didn’t have phobias in mind at all when he developed this kind of imaging, called hyperalignment. He wanted to come up with a way to look, not just at what parts of the brain turned on and off when people looked at something, but the brain's overall pattern of activity, and its relationship to other patterns.

To do this, he had people look at pictures of all different things: faces, animals, scissors, shoes; everyday animals and objects. He found that the pattern of activity in the visual cortexes of the brain were very distinct, depending on what people were looking at.

But between each person, those patterns didn’t line up perfectly. This is where hyperalignment comes in: it’s a way of taking the patterns from each person, and turning it into a common space, or model, which is shared among all the people. It’s like a template, Haxby says, for what the specific response to an object, animal or person looks like in the brain.

“The key concept is this common space,” Haxby says. “By doing the hyperalignment, the features that are associated with visual information are now lined up. You have a common template for what the response to an object looks like. Once we figure out how to transform an individual’s cortical patch or field into this common model, which is shared across subjects, we can say: we can see the same pattern for shoes as compared to chairs. Or for monkeys as compared to lizards.”

In Taschereau-Dumouchel’s new study, they could use hyperalignment to come up with a prediction of what a subject’s brain was going to look like when they were looking at the thing that scared them. But they still wanted the whole process to remain unconscious. If they showed phobic people images of their fears again and again, to get the hyperalignment information, that wasn’t much better than an exposure. So instead, they showed healthy participants 3,600 images of 30 different animals and 10 inanimate objects. Since these people didn’t have any phobias, they acted as the “surrogates” for brain activity for the phobic people.

They used the surrogates' brains to come up with a "hyperalignment decoder" that provided the template for brain activity for certain fears, like snakes. They could infer what the people who were scared of snake’s brain would look like, if they were seeing them.

The experiment continued on 17 people who had high levels of fear for at least two of the animals in the images shown to the surrogates. For each participant, one of the feared animals was chosen at random to target; the other acted as a control—to see if what they were doing was specific enough to hone in on a single fear. First, the researchers measured the participant’s fear response by showing them images of the animals, as well as neutral images, and measured their skin conductance response, the minute physical changes that take place on your skin during stress, and the activity in their amygdalas, a part of the brain associated with fear.

Participants were then placed in an fMRI scanner and saw a circle, or disc in front of them. They were told to make the disc bigger just by thinking, but not told what to think about to do so. The bigger they could make the disc, the more money they would get after the experiment was over. Meanwhile, the fMRI scanner was reading their brain. If their thought patterns got closer to the template for what they were scared of, like snakes, the disc got bigger. If their thoughts moved further away from the template for snake, it got smaller.

Participants did this over the course of five days, and did effectively learn how to make the disc bigger. But they never consciously realized that they were supposed to be thinking about snakes, or that that was the goal of the experiment. They were just trying to get the reward. They learned to activate the parts of the brain that are usually activated when they thought of snakes, or close to them, but without the dreaded thinking-about-snakes part.

“Most of them didn’t understand very well what we’re asking,” Taschereau-Dumouchel tells me."We asked what strategies they used at the end and most reported very unrelated strategies. Like thinking about what they would eat for dinner, thinking about stuff that happened in their childhood. Almost all the strategies were absolutely unrelated to what we were trying to do.”

When they measured each person’s response to their fears afterwards, they found that their amygdala responses had decreased, and so did their skin conductance response—and only for the animal they were targeting, not the other one they were afraid of that they weren't being rewarded for.

Taschereau-Dumouchel says the reward association is one possibility. The other is that just by having people produce this pattern again and again, it acts as its own form of exposure therapy, but without the aversive experience. “It’s as if people learn to activate these patterns, and get used to them, and as a result their reactivity goes down,” he says.

It’s an exciting development, but there’s still a lot to learn. Their current results only measured physiological responses to a person’s fears. So, they saw that a person's body became less reactive to their phobia, but didn’t actually ask them: are you less afraid? This is something to be addressed in future work, but Craske says even just a physical response change might be meaningful. If you’re afraid of something, your body joins you: your heart pounds, your hands sweat, your head feels light and woozy. Even if the cognitive fear remained, it’s compelling to imagine how much easier it would be to try and face your fears if your body could remain calm while you did so.

“The way our brain works, our body responds, and how we behave are all connected,” says Carmen McLean, a clinical psychologist at at the Center for the Treatment and Study of Anxiety at the Perelman School of Medicine at University of Pennsylvania, who wasn't involved in the study. “It's really just looking at the same phenomenon from a different angle. For example cognitive therapy produces behavior change and behavior therapy produces cognitive change.”

Still, this study acts as more of a proof of concept than anything else. The next step is to apply it in a clinical trial setting—the authors, in collaboration with Craske, currently have a grant pending to do so. Because of the unconscious nature of the task, they’re hoping to create a double-blind placebo trial, which can often be difficult to design in psychological therapy studies.

I asked Taschereau-Dumouchel about more abstract fears like mine. My fear of vomiting might be easy to visualize (too easy) but what about germ contamination? Food poisoning? Those are less obviously visual, and could they be addressed too? He says they don’t know yet, but are developing an experiment targeting OCD right now, together with collaborators at Johns Hopkins. They’re going to try showing visual representation like, clean and contaminated toilets, or dirty and clean hands, to see if that has the same effect.

It’s these details that Craske and Taschereau-Dumouchel hope to nail in clinical trials. They want to know if this treatment can stand alone, or if it's something to be done before exposure therapy, or in conjunction with exposure treatment through virtual reality– an increasingly popular alternative to traditional exposure therapy. “What’s the best sequencing of events, is it sufficient on its own, or does it require additional components?” Craske says. "Right now, it’s too early to say. But it’s such an exciting development, let’s focus on that and see what happens."

McClean cautions that “the cost and availability of such a procedure is prohibitive,” she says. “This is not going to help us reach more people with phobias.” She's right. This is probably not something–as much as I wish I could try it–I could easily go to my doctor and get. I would need to recruit people to be my surrogate brain activators, and have access to an fMRI scan for multiple days.

Craske acknowledges this, saying “It’s going to be for those people who just cannot do the exposure therapy, in the direct, real life way. It’s going to be for those people who are so resistant to the idea that this is the only path they have.”

Nonetheless, understanding the basic neural mechanism of a fear and our responses to exposures to that fear can be helpful in unforeseen ways in the future. Haxby can attest to this now; he never envisioned that his method of visualizing the brain might be used for unconscious treatment of phobias.

“It's like science fiction,” he says to me, laughing. “It’s remarkable. I was just talking to my lab today and said, if anybody in my lab ever proposed this unconscious neural reinforcement, I would have said you’re crazy, that’s never going to work. But that’s how science works. A lot of the basic science seems kind of dry and it’s unclear how it might be useful. And then somebody else sees it and says, oh I know I can use that.”