Science: How Brain Scans Can Detect Suicide Risks

Science: How Brain Scans Can Detect Suicide Risks

Just a week before his own suicide, Jeremy Richman flew down Florida to deliver a talk he called “The Neuroscience of Being Human(e).” In it, the 49-year-old neuropharmacologist examined how brain research might help identify people in crisis, so that we might better intervene and help those at risk of violence to themselves or others.

The topic could not have been more timely: Just that week, two students who had survived the massacre at Marjorie Stoneman Douglas High School in Parkland, Calif., had taken their own lives. And it was a subject he himself cared deeply about. On December 14, 2012, Richman and his wife Jennifer Hensel lost their little girl Avielle, 6, when a troubled teenager stormed Sandy Hook Elementary School and killed her and 25 others before shooting himself. Within 48 blurry hours, Richman would later recall, he and his wife had resolved to channel their grief into action, establishing a foundation in the name of their daughter to fund brain research to both prevent violence and build compassion. To help run it, Richman, who had earned a doctorate in pharmacology and toxicology from the University of Arizona, left his job working in drug discovery and focused full time on the issue.

In the area of suicide, there was plenty to talk about. In recent years, neuroscientists have made dramatic strides in their ability to recognize those who are most at risk—identifying distinct anatomical characteristics present in the brains of those more predisposed to take their own lives, and even distinct neural activation patterns elicited by specific words that can be spotted in the brains of those with suicidal ideation.

For all his efforts as a tireless promoter of his cause, though, in the end, nobody was able to prevent Richman from taking his own life.

“Tragically, his death speaks to how insidious and formidable brain health can be and how critical it is for all of us to seek help, our loved ones, and anyone who we suspect may be in need,” the Avielle Foundation posted on their website late Monday on behalf of Richman’s devastated family. “We are crushed to pieces, but this important work will continue, because, as Jeremy would say, we have to.”

The fact that an expert in brain health could be felled by one of the very conditions he had devoted himself to combat is a cruel irony. But researchers who study suicide and its connection to the brain aren’t surprised. In many ways, it gets to the heart of what they have discovered.

In recent years, many new insights have begun to emerge—and one of the key insights is that knowledge is no protector. Suicide, several noted, is one of the leading causes of death in medical students and young doctors—even those specializing in psychiatry.

Still, the effort Richman devoted his foundation to promoting is making progress—and has in fact made giant leaps in recent years. It was catalyzed more than 25 years ago, when psychiatrists and neuroscientists at Columbia University and the New York State Psychiatric Institute made a fortuitous discovery. In an effort to better understand the pathology of depression, a team of researchers had begun collecting from medical examiners the brains of individuals who had died of suicide, reasoning that all of them were likely depressed. But when the team began to speak to the families that donated the brains about their deceased loved ones, they made a surprising discovery: roughly half of the patients didn’t have depression as an illness.

Even more surprising, as the team began comparing the brains to each other and those of individuals who died of natural causes, there was, they found, an identifiable constellation of neurological characteristics seen to a much higher degree in those who had committed suicide, regardless of whether they were depressed or not.

“Nobody at that time had any idea that there was a separate phenomenom of brain abnormalities associated with suicide,” says J. John Mann M.D., a neuroscientist who made the discovery and is Vice Chair for Research in the Department of Psychiatry at Columbia University.

For the last two and a half decades, Mann and his collaborators have been working to characterize these differences by performing biochemical analysis on the neurotransmitter systems of donated brains, measuring connectivity, and using brain imaging techniques to study the neural patterns of individuals who have survived suicide attempts. A number of distinct differences have emerged.

In suicidal individuals, parts of the brain involved in producing neurotransmitters were fundamentally altered in some areas. In some parts there were fewer neurons and a thinner cortex. And the areas where these abnormalities were found weren’t necessarily the ones you might expect. Ninety percent of suicidal individuals, they found, were suffering from one form of mental illness or another at the time of their death. In depressed individuals and those suffering from other forms of mental illness, the emotional centers of the brain called the amygdala are hyperactive. And often it was the mental illness that activated a predisposition to suicide.

Yet the primary difference between those more likely to commit suicide wasn’t in the amygdala: the differences were in the areas of the brain known as the anterior cingulate cortex and the dorsal lateral prefontal cortex, which control how individuals subjectively assess the stress of their condition.

“Objectively they have the same severity of illness, but subjectively they experience it as much more depressing,” Mann says. “They seem to have difficulty regulating their emotions. And their subjective stress is more severe than other individuals who are not at risk of suicidal behavior. In other words, their depression sensor mechanism is hypersensitive.”

Nor was that the only difference; areas in the brain involved in decision making were also abnormal. Those who attempted or committed suicide were more likely to make risky decisions when presented with a task aimed at characterizing decision making. (These abnormalities were also seen in the interior anterior cingulate, a brain area that Mann calls the “CPU [central processing unit] of decision making.”)

In addition, the ability of suicidal individuals to learn and problem solve is abnormal in depressed people, but more so in suicidally depressed people. Finally, those prone to suicide seemed to have more problems reading social signals, were hypersensitive to negative social cues and had a reduced response to positive social cues. They perceived the world as less helpful, more critical and threatening.

“All of these contribute to suicidal behavior, you feel the depression more acutely, you are more prone to act on your feelings, you can see fewer options or solutions and finally you see the individuals around you as more critical and less helpful,” Mann says. “The unfortunate thing of course is that the person themselves doesn’t realize they are any different than anybody else. They are not aware of the risk.”

Todd Gould, a professor in the depart Departments of Psychiatry, Pharmacology, and Anatomy & Neurobiology at the University of Maryland School of Medicine, notes that those who study the neurological causes tend to break the problem into two parts and study them separately: there is the thought of acting, and the actual act itself.

The thought that life is not worth living often goes along with depression, Gould says. But the biological circuits involved in impulsivity and decision-making play a key role in the willingness to act on those feelings. Many individuals are prevented from acting on the desire to die because they think about how it might affect their families and friends, and after weighing the risk and benefits, they conclude the cost would be too great. But suicidal individuals often act before contemplating the implications of the act.

“It seems that certain personalities goes along with acting on suicidal thoughts,” he says. “In many cases, it’s an impulsive behavior to act. And if an individual had a different perspective, or thought about how it would affect their family or others in the long run they wouldn’t act.”

Aggression, too, seems to be involved. Gould notes that as far back as Sigmund Freud, theorists suggested that suicide is aggression turned inward. He has been studying the circuits involved in impulsivity and aggressiveness in animals and notes that a growing body of research has shown that the effectiveness of Lithium salts in reducing suicide amongst depressed patients likely stems from its ability to affect these circuits.

A number of studies have found that ketamin has an acute and immediate effect on reducing suicidal thoughts. But the mechanisms are not yet understood, he adds.

The best solution, most agree, is screening. Mann suggests everyone should be screened at least once a year—and those deemed at risk far more frequently. Even questions like “do you find life worth living” could go a long way towards identifying those at risk, he says.

Meanwhile, neuroscientists are finding new ways that may be even more powerful at detecting who is at risk. In 2017 researchers at Carnegie Mellon University and the University of Pittsburgh demonstrated they could use machine learning algorithms to identify patients suffering from suicidal ideation just by analyzing their brain scans.

“When you have a psychiatric condition it changes the way you think about certain things,” says Marcel Just, a cognitive psychologist and Director at the Center for Cognitive Brain Imaging at Carnegie Mellon University. “A paranoid person, for instance, will have different brain activation patterns if you say the word 'police.' The same is true with certain words for those with suicidal ideations.”

Science: How Brain Scans Can Detect Suicide Risks

Using fMRI, the researchers identified individuals with suicidal tendencies. Carnegie Mellon University

As his colleague David Brent read subjects a list of 30 words while they sat in brain scanning machines, Just analyzed their brain activation patterns. The subjects consisted of three groups: those who had already attempted suicide, those experiencing suicidal ideation who had not yet made an attempt, and a control group. Just and Brent found that brain patterns elicited by just six words were enough for their machine learning program to identify the patients with suicidal thoughts 90 percent of the time. (The words were death, cruelty, trouble, carefree, good and praise.) The program could also tell the difference between those who had attempted suicide and those thinking about it at 80 percent accuracy.

The activation patterns were systemic and distinct. Most notably, the areas of the brain involved in "self-reference" were far more active in those with suicidal ideation when read the list of six words than the control group. That is, the word “death” might evoke thoughts of Iraq in some people. But in people who were predisposed toward suicide, the areas of the brain that lit up were those involved in thinking about and defining oneself.

The technique could be used to guide therapy and measure whether it is working to reduce the risk, Just says.

Jeremy Richman, an accomplished pharmacologist and biologist who had an appointment as a faculty lecturer in psychiatry at the Yale School of Medicine, was probably aware of all this research. But it wasn’t enough to save him.

“Insight doesn’t necessarily lead to behavior change,” says Brent, who is the Endowed Chair in Suicide Studies and Professor of Psychiatry, Pediatrics, Epidemiology, and Clinical and Translational Science at the University of Pittsburgh. “Forget about what he knew. Think about what he was going through: he lost a six-year-old child. We know that people that are bereaved, particularly if their grief doesn’t subside, are at risk of suicide. If he was depressed; if he had PTSD; if his grief was not resolving—all of these things could have set him up for this. And the fact that he was an accomplished neuroscientist is immaterial.”