Craving in Addiction May Alter How the Brain Makes Decisions

For people with an addiction, craving—the strong desire for a substance—can affect their decision making, new research shows. And how craving affects a decision can depend on what's at stake.

The finding, published March 26 in Nature Mental Health, is important, say researchers, because understanding craving is key to understanding—and treating—addiction itself.

“Addiction is very hard to treat, and one of the reasons is that we don’t fully understand the craving linked with it,” says senior author Xiaosi Gu, PhD, professor of psychiatry and director of the Computational Psychiatry Unit at Yale School of Medicine (YSM).

While craving and making harmful choices are linked across different addictive disorders, it’s unclear how and to what degree they affect each other. In their new study, Gu and her team used computational models to gain a better understanding of the relationship between craving and decision making.

"If we get insight into what the brain mechanisms surrounding craving are, we can get a lot of insight into why people become addicted," says Kaustubh R. Kulkarni, MD, PhD, first author of the study and a psychiatry resident in YSM’s Neuroscience Research Training Program.

The study included 132 people who reported that they engaged in moderate to heavy alcohol or cannabis use.

The participants completed a task in which they had two slot machines to play with and a goal to maximize how much they won across the two machines. At a given time, one of the slot machines would allow a participant to win 80% of the time, while the other machine only offered winnings 20% of the time; the “winning machine” would then switch several times throughout the experiment. The participants were told one machine would yield more wins than the other, but not which machine was which.

When participants won a session, they would receive either “money” (an image of a coin) or be shown an image related to their substance of use, such as photo of beer for the alcohol group or a photo of a bong for the cannabis group. Participants also reported their craving levels and mood at various points throughout the experiment.

The researchers were interested in whether craving affected choice and what impact winning or losing had.

When participants played the game for money, they identified the winning slot machine quickly: After the participants won a monetary reward, they were more likely to choose that same slot machine again. If they didn’t win, they adjusted their choice, trying out the other machine.

But when the reward involved alcohol or cannabis images, the participants’ level of craving altered that learning process, the researchers found.

Through computational modeling, the researchers found that, among the alcohol group, stronger craving led participants to more quickly home in on a winning strategy. Among the cannabis group, however, greater craving slowed that learning process.

“There's something that distinguishes the mechanism of alcohol use vs. cannabis use,” Kulkarni says. "It could be the actual pharmacologic effect of alcohol on the brain, which is pretty different from cannabis, or it could be something else entirely."

The findings suggest that craving potentially rewires how people learn from their experiences. For someone with alcohol addiction, for example, intense craving might create a feedback loop, says Gu.

“This could explain why breaking the addictive cycle feels so difficult, as the brain is adapting constantly,” she says.

The research opens new possibilities for treating addiction, the team says.

“We've now demonstrated there's some computational dysfunction that has occurred," Kulkarni says. "Now we need to understand what part of the brain this is implemented in."

One way to do this is through neuroimaging. The team is now conducting follow-up studies using brain scans, funded by the Yale Biomedical Imaging Institute, to identify which neural circuits are involved.

“We need to understand not just the computational, but the neurocomputational framework of how particular parts of the disorder are implicated," Kulkarni says. "Then we can use our understanding of current pharmacologic treatments, brain stimulation treatments, or even psychotherapy to tailor those to particularly target the circuits that are implicated."

While the team cautions against inferring clinical applications at this stage—translating research into clinical applications takes time, says Kulkarni—their eventual goal is to use these computational fingerprints as biomarkers to assess patients and predict treatment outcomes.

"My hope is that these gamified behavioral assays will be used to assess patients in the clinic," Gu says. “Unlike brain imaging, which requires expensive equipment, behavioral tasks can be administered anywhere, even remotely during telemedicine visits.”

And this work could extend beyond addiction, she adds. "To take this intense subjective urge and be able to study it in this way, I think it is very important as it also impacts a lot of other conditions such as eating disorders and feelings of loneliness or social isolation."