What if we could peer into our brains and find evidence capable of predicting the future?
What if this were possible for something even as complicated as knowing the likelihood of a relapse for those recovering from a substance use disorder?
At Yale, this is not a hypothetical. And, with a new Women’s Health Research at Yale-funded study on the biological intersection of pain and opioids, the future is now.
Dr. Sarah Yip, Assistant Professor of Psychiatry and Director of the Yale Imaging and Psychopharmacology Lab, is using a new technique developed at Yale by Dr. Todd Constable, Director of MRI Research, and colleagues. Dr. Yip aims to understand for the first time how patterns of brain organization might differ between women and men in the experience of pain and analgesia for the purpose of learning how to better treat chronic pain and avoid opioid addiction.
Known as connectome-based predictive modeling, the technique produces a “brain map of connections” associated with a behavior of interest, such as future substance use following treatment for addiction. The researchers first examine a set of data involving brain images and behaviors to build a predictive model in a training dataset. Then, the researchers see if they can use the connections that they have found to predict the same behavior in an entirely new group of people — the testing dataset.
This well validated “machine learning” method is designed to avoid over-fitting to any specific data set and thereby increases the likelihood their findings will be generalizable and applicable to real-world settings. Dr. Yip and her colleagues have already applied this technique to successfully predict cocaine abstinence during a 12-week treatment program.
“The goal of traditional statistical approaches is to explain the relationship between two variables,” Yip said. “The goal of this new technique, guided through machine learning, is to generate predictions in novel data. It’s the difference between looking backward versus looking forward.”
Yip, working with co-project leader, Dr. Sarah Lichenstein, and Dr. Declan Barry, is building on her prior work with colleagues Drs. Kathleen Carroll and Marc Potenza published in the American Journal of Psychiatry and in Molecular Psychiatry. Using magnetic resonance imaging (MRI), they aim to use this form of brain mapping to identify a “neural fingerprint” that shows pathways in the brain involved in the sensation of pain and pain relief through opioids.
This study will be one of the first studies to investigate the neurobiology of pain and analgesia at the same time as investigating sex differences. Fewer than 15 percent of participants in neuroimaging studies of opioid use disorder (OUD) have been female, and none of these studies sought to identify differences between females and males.
This is important because women account for 70 percent of Americans suffering from chronic pain and are more likely than men to become exposed to opioids when seeking medical treatment for pain. Once exposed, women are more likely than men to become addicted. Overdose deaths from opioids have grown for years and spiked substantially during the COVID-19 pandemic. The rate of increase in overdose deaths also has grown more for women than men. Altogether, a detailed understanding of how sex influences pain and pain relief promises a better way for women to treat their chronic pain.
Moreover, researchers have developed evidence-based treatments for substance use disorders that work for some individuals. But simple clinical variables such as baseline severity and years of past use do not consistently predict treatment success. Here too, Yip’s study could help identify sex-specific treatment targets for substance use based on how brain patterns of women and men process rewards and achieve abstinence. It could also identify individuals more likely to relapse into opioid use who can then be provided additional resources that would be cost effective, such as more frequent interactions with a counselor specializing in substance use prevention.
“The current standard of addiction treatment often involves multiple failed attempts,” Yip said. “Identifying the people most at need and designing treatments best suited for their individual needs would save money in the long run. More importantly, women are suffering, and science can show us how to help.”