Brain Biomarkers May Help ID Those at Risk for Severe PTSD

Deborah Brauser

February 14, 2019

Brain biomarker patterns may help clarify why symptoms of post-traumatic stress disorder (PTSD) are more severe in certain individuals, new research suggests.

Ilan Harpaz-Rotem, PhD

A study of 54 combat veterans exposed previously to intense events showed that patterns of neurological and physiological responses to associative threat learning were different in those with severe PTSD compared with those with lesser symptoms.

Associative learning is the ability to distinguish between harmful and safe factors, such as people, objects, and even sounds. In the study, participants were shown pictures of two mildly angry faces.

Using fMRI and computational modeling, investigators found that the amygdala and the striatum in the participants with more severe PTSD symptoms were less able to track changes in threat level, leading to a mismatch between expectation and actual experience.

An example of this type of mismatch is looking at a trash can, which could have contained an explosive device in a war zone, and having difficulty realizing that it has little-to-no threat in a civilian setting, the researchers note in a press release.

"We are shedding new light on how people learn fear and unlearn it. There were pronounced differences in the 'learning rates' of those with severe symptoms and those without," co-principal investigator Ilan Harpaz-Rotem, PhD, associate professor of psychiatry at the Yale University School of Medicine and from the US Department of Veterans Affairs National Center for PTSD in New Haven, Connecticut, said in the release.

"People in our study with a higher level of post-traumatic stress disorder overcorrected their expectations the next time they saw the stimuli. They were unable to adjust their learning as well as those without PTSD," he told Medscape Medical News.

Harpaz-Rotem added that the investigators hope to eventually "translate the findings into clinical implications" through further research.

The findings were published online January 21 in Nature Neuroscience.

Many Questions

The researchers note that for quite a while, questions have swirled around why some military personnel develop symptoms of PTSD after returning from combat while others do not — and why the symptoms are more severe in certain individuals.

"There have been mixed results in [past] studies that looked at fear learning and distinction of fear in populations that have developed PTSD. So we wanted to look at the ways different people learn," Harpaz-Rotem said.

"With associative learning, you are learning through experience and informing through decisions based on the information that you get over time. And there are new computational models that allowed us to look at these individual differences in learning when associating fear or potential fear," he added.

The investigators enrolled 54 combat-exposed veterans (90% men; mean age, 54 years) with a range of PTSD symptoms, as measured on the Clinician-Administered PTSD Scale (CAPS). Of these participants, 24 had a clinical diagnosis of PTSD and the remaining 30 did not.

The researchers used a process that required continuous and flexible updates of responses to threats by the participants (a "threat reversal paradigm") plus computational modeling to determine learning patterns.

Photos of mildly angry faces acted as two different visual stimuli. In the first phase, the photos were presented on a screen in a somewhat randomized order; but during a third of the presentations, only one photo was paired with a mild electric shock to the foot. During the reversal phase that followed immediately, the other face photo was the only one paired with an electric shock.

"In this type of reversal learning, we showed that something was safe and something was unsafe. And then in the middle of the experiment, we switched it around and what was safe became 'dangerous' and the dangerous became 'safe,'" said Harpaz-Rotem.

"Clear Evidence"

Results at first showed that all of the veterans demonstrated successful reversal learning. After controlling for age and sex, there was no significant association between reversal learning and symptoms of PTSD (P = .089).

These findings led to the additional use of computational models, including a hybrid Rescorla-Wagner and Pearce-Hall model.

This "more fine-grained" analysis showed that "symptomatic veterans assigned higher weights to prediction errors," the investigators write. In fact, this higher prediction-error weight was significantly linked to higher symptom scores on CAPS (P < .001).

"We found that symptomatic veterans showed greater physiological adjustment to cues that did not predict what they had expected, indicating greater sensitivity to prediction errors for negative outcomes," the researchers write.

"This exaggerated weighting of prediction errors shapes the dynamic learning rate (associability) and value of threat predictive cues," they add.

A linear mixed model showed that decreased tracking of threat-level associability in the striatum was significantly and independently linked to increased PTSD severity. Decreased tracking "of value" in the amygdala, as well as smaller amygdala volumes, were also associated with higher symptom severity scores.

"These results provide evidence for distinct neurocomputational contributions to PTSD symptoms," the investigators write. However, they add that the results "do not allow us to draw causal inferences."

New Understanding

Still, "the findings of this study provide new and innovative understandings of the neurobiology of PTSD and a better understanding of learning processes in this population that might be useful in the future to refine treatment for the disorder," co-author Ifat Levy, associate professor of comparative medicine and neuroscience at Yale, said in the release.

"Post-traumatic stress disorder has a unique feature. People tend to overgeneralize fear, such as associating loud noises from an explosion with every loud bang or similar noise. How they learn that something is no longer dangerous in a new environment is really fascinating to us," Harpaz-Rotem told Medscape Medical News.

"In our study, their adjustment or learning was not well calibrated. They gave too much weight to the most recent presentation they received compared to the other people," he added.

Harpaz-Rotem agreed with Levy that the findings may help to shed a light on what needs to be better targeted in regards to treatment.

"Understanding how different individuals learn and update their fear and expectation from certain events is extremely important. We know that right now only about 50% of the people that receive exposure therapy improve. So trying to understand what those things are that interfere with fear learning in making exposure therapy effective is something that we are really interested in," he said.

The study was primarily funded by a grant from the National Institute of Mental Health and by the Clinical Neurosciences Division of the National Center for PTSD. The study authors have disclosed no relevant financial relationships.

Nat Neurosci. Published online January 21, 2019. Abstract

Follow Deborah Brauser on Twitter: @MedscapeDeb

For more Medscape Psychiatry news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: