This story was originally published in our July/August 2022 issue as "Redefining Suicide." Click here to subscribe to read more stories like this one.
Just speaking the word suicide — or reading this story headline — might make your stomach drop. There’s no gentle way to broach the subject. But the sheer number of people who are taking their own lives demands a closer look.
At least 700,000 people die by suicide each year, according to World Health Organization (WHO) data. In the U.S., well over 45,000 people become victims annually. That’s more than the yearly count of deaths from breast cancer, leukemia or prostate cancer.
These numbers fit into a concerning trend: Americans are now killing themselves at a rate roughly 30 percent higher than just 20 years ago.
The broader factors driving the uptick over the past two decades are vast and complicated. And, on the individual level, so too is diagnosing anyone as “at-risk for suicide” before it’s too late.
The WHO describes a prior suicide attempt as “by far the strongest risk factor” for an individual. Yet more than half of the suicide fatalities in the U.S. result from first-time attempts, according to multiple studies. This can leave many victims off the radar for support.
Psychologists, neurologists and other experts are turning to expansive data sets and machine-learning algorithms to pinpoint risk indicators and trends in suicidality that previously went unnoticed. In broad application, this work could address the sobering reality that suicide is now the second-leading cause of death in the U.S. (behind “unintentional injury”) for anyone between 10 and 34 years old.
“People are increasingly killing themselves, and young people particularly,” says Igor Galynker, director of the Suicide Research and Prevention Lab at Mount Sinai Beth Israel. “I don’t think it’s going to get better, given the stressors that humanity is facing."
What could improve: our detection of suicide risk and the corresponding treatments. One targeted solution is establishing a proper condition of suicidality in the Bible of the mental health world, the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM). Other efforts are looking for indicators outside the realm of psychology. Beyond the mind, physiological biomarkers might even lead the way to new, timely interventions.
Cracks in Treatment
Julie Goldstein Grumet has worked at the forefront of mental health and suicide prevention efforts since she examined the topic in her 2001 dissertation to become a clinical psychologist. She later launched and became director of the Zero Suicide Institute at Education Development Center in 2016, based in Waltham, Massachusetts. The agency operates from the core belief that all suicide is preventable and treatable. Given the data today, that leaves an urgent, constantly evolving question: Why are we seeing so many incidents?
In part, people are falling through the cracks of a fragmented health care system, according to Zero Suicide. Goldstein Grumet says a lack of standardized understanding and training, starting with graduate programs, ripples into the wide range of support options and care received. “It’s a little bit of the Wild West,” she says.
Education in medical schools has improved over the past few years, she adds. But that doesn’t suddenly reverse the vast system’s deficiency in knowledge and methods.
As one troubling example, Goldstein Grumet points to “no-suicide contracts.” These non-binding agreements (often signed on paper by a patient) served as a popular treatment response to suicidal ideation, starting in the 1970s and lasting well into the early 2000s. They’re still sometimes used, despite limited evidence of effectiveness. Essentially, physicians or therapists were asking people to promise they wouldn’t act on suicidal thoughts. “You’re not offering them any kind of alternative or intervention or wellness plan and coping skills,” she says.
Some shifts began between 2000 and 2010 with a federal push toward suicide awareness. Efforts often encouraged everyday mentors like teachers, coaches and pastors to notice warning signs and refer at-risk people to medical support and resources. “But health care wasn’t all that prepared,” Goldstein Grumet says. “You’re shooting people to the emergency room or hospital with a very well-intentioned but pretty ill-prepared workforce.”
While dynamic treatments and therapies have advanced in recent years — including magnetic stimulation and ketamine — many of them are centered on mental conditions like anxiety or depression. It’s now clear this scope runs too narrow, according to Goldstein Grumet and experts who are evaluating the likely contributors and circumstances preceding a suicidal crisis.
Seeing the Signs
In suicide prevention work, experts have frequently zeroed in on ideation, a broad term referring to thoughts of suicide. This is one reason psychiatric evaluations often ask a patient if they have considered self-harm. But ideation may not reliably foreshadow fatal consequences or even an attempt, according to data published over the past decade.
A 2021 review of suicidal ideation published in StatPearls flagged that “there is no clear association” between someone contemplating killing themselves and acting upon that thought. Machine-learning studies, which use sophisticated computer algorithms to process and analyze data, have also measured a slight but ultimately insignificant correlation between ideation and suicidal behavior in the near future.
Out of every 31 Americans with suicidal ideation, only one will actually attempt it, according to a review of a comprehensive national survey conducted between 2009 and 2014. And out of the people who follow through, some research suggests up to 75 percent “emphatically denied” an intention to commit suicide in communication shortly before the event, according to Galynker.
These findings challenge the prominent role that ideation has played in suicide-risk assessment. They also expose the need for more reliable measures.
“If you step back and think about it, would you ever rely on a patient with schizophrenia to diagnose themselves with schizophrenia? … [Or] a patient with bipolar disorder to diagnose themselves as bipolar?” Galynker asks. “And both of these conditions are less lethal than pre-suicidal mental state.”
Cultural context might also inform the person-dependent warning signs preceding suicidal action. Mental health researcher Michael Lindsey worked on multiple 2021 studies that highlighted rising suicidal behavior and particular patterns among Black youth. Namely, Black children and teens in the U.S. are at least twice as likely as their white peers to make a suicide attempt without voicing suicidal thoughts or plans, according to a sample of nearly 7,500 incidents reviewed in Prevention Science.
“I think we’ve been too reliant on these traditional risk factors or warning signs,” says Lindsey, who is executive director of the McSilver Institute for Poverty Policy and Research at New York University. “We’re able to distinguish patterns of risk that look very different, per race and ethnicity.”
Data analyses have helped catch these nuanced trends, and are just beginning to unpack potential contributors, Lindsey says. An AI hub that launched last year at his McSilver Institute will dig for possible explanations.
Lindsay says that some overlooked variables in suicide evaluation include trauma, limited access to resources and a reticence to acknowledge sadness or depression within many Black communities. These factors might create numbness, with no apparent ideation, in an individual while the risk for suicide grows. “It could be that there’s this kind of ticking bomb,” Lindsey says, “[that] ultimately will fester up and explode in the form of hopelessness or self-harm behavior.”
The intricate relationship between race and suicide gained some exposure in U.S. media last year, after reports revealed that the overall U.S. suicide rate in 2020 actually dropped 3 percent — a surprise to many, given the stressors of a national pandemic.
Parsing the numbers, however, may reveal varied impacts across racial groups. For example, Maryland’s average suicide count nearly doubled among Black residents, but decreased by half among white residents shortly after the pandemic emerged, according to a research letter published in JAMA Psychology.
Meanwhile, numerous reports showed how the pandemic disproportionately harmed Black populations, Indigenous communities and other people of color. And for many years, American Indian and Alaska Native populations have experienced the highest rate of suicide in the U.S.
“We can no longer afford to have a cookie cutter approach to how we assess risk,” Lindsey says.
Most suicide attempts take place less than one hour after crossing a person’s mind, according to various studies from the past two decades. Coupled with what we’re learning about ideation, some researchers emphasize that suicide prevention efforts must operate within a short time frame and be tailored to specific symptoms.
That’s one reason why Galynker’s team aims to establish a suicide-specific diagnosis in medicine, similar to depressive disorder or generalized anxiety disorder. This would legitimize the pre-suicidal mental state as its own condition, rather than an outcome of another mental illness. “Our approach is defining the symptoms, so you can find the treatment,” Galynker says. “It is a critical change.”
Suicide Crisis Syndrome (SCS), as Galynker calls it, is very short lived, spanning minutes or hours, like a panic attack. The criteria, or symptoms, include frantic hopelessness, extreme panic, loss of cognitive control — or “going down the rabbit hole” — and acute social withdrawal. As for common triggers of SCS, Galynker’s research places things like financial or career failure, a romantic breakup, bullying and homelessness high on the list, along with intractable mental illness.
Notably, ideation is absent from Galynker’s five criteria for identifying suicide risk in a person. This approach could equip medical professionals to recognize a pre-suicidal mental state without relying on ideation. Galynker says specific and short-term medication should treat acute, urgent symptoms of SCS. Then, tools such as psychotherapy and counseling can address other long-term mental health matters.
After various promising studies, SCS was submitted to an APA steering committee for preliminary review in 2020. Galynker’s team is now collecting additional data requested in that process, including a study seeking consistency in the model across 14 different countries.
Regardless of its clinical classification, suicidality and its indicators surely involve more than mental health. “Historically we said it was a mental health issue,” Goldstein Grumet says. “That’s not the case anymore. People are far more complex.”
For starters, the majority of U.S. suicide victims (54 percent) have no known mental health condition when they carry out the act, according to CDC research conducted in 2016. While interpretations of that statistic vary, it might suggest that the contributing factors in suicide are not always psychological. The data could also point to shortcomings in care and screenings for people in distress.
We know that patients vulnerable to suicide routinely slip through the cracks of the health system after seeing physicians, whether that is an emergency room visit or primary care appointment. In fact, over 60 percent of suicide attempts follow a health care visit within the past month.
Preventive measures might simply involve patient follow-up care, or more advanced evaluation of personal records and patterns. A 2016 American Journal of Psychiatry study applied computer models to 15 years of historical data in patients’ electronic health records to assess for suicide and create sophisticated risk profiles. The work demonstrated the potential of an early-warning system to flag patients for further screening and enhance predictions beyond the feasibility of individual clinicians.
Work in the past decade has also evaluated biomarkers, such as altered stress responses in some people’s bodies, along with abnormalities in the adrenals and a collection of neurons across the brain stem called the serotonergic system. While that research is in its infancy, private companies have already started investing in pharmaceutical solutions that could address similar potential indicators.
Campbell Neurosciences, which launched two years ago, is zeroing in on a particular protein in the blood that may be associated with brain inflammation and suicidality. The company is currently recruiting participants for a clinical trial. “If there’s a biological reason for it, then we
might be able to intervene and reduce the probability,” says Thomas Ichim, a chief science officer with Campbell Neurosciences who has a Ph.D. in immunology. “We’re trying to create a new tool.”
As interest grows, some experts say that proposed solutions might be getting ahead of more fundamental research, like the challenge of defining suicide as a medical condition. But this wouldn’t be the first time that an ambitious fix preceded our full understanding of the ailment it treats — for instance, experts still can’t agree on how exactly acetaminophen relieves pain.
Maria Oquendo, chair of psychiatry at the University of Pennsylvania’s Perelman School of Medicine, has worked in the past few years on several studies assessing the response to stress — sometimes using brain imaging — in the bodies of people experiencing depression and suicidal ideation. Her recent research proposes at least two subtypes of suicidal behavior: One primarily relates to the stress hormone, cortisol, and reactive ideation, while the other focuses on the serotonergic system, and chronic ideation.
“If we can refine our biomarkers so that we can reliably identify those at acute risk, that would be hugely important,” Oquendo says. But she adds an important caveat regarding this burgeoning field. “We are not yet at the point where biomarkers have the type of sensitivity and specificity needed for them to have clinical utility.” Should we reach this point, it could open the door to some sticky ethical terrain, according Galynker: “Have you seen the movie Minority Report?”
The Steven Spielberg film depicts a world in which law enforcement can glimpse future homicides and arrest would-be murderers before they act. Hollywood aside, it’s no easy choice to decide how to treat someone before they become an actual risk to themselves or others. “You identify someone who is high-risk,” Galynker says. “So, what are you going to do?”
While that conundrum remains a problem for the future to solve, Goldstein Grumet for now proposes a meaningful, broader shift in prevention care: “What is a life worth living? How do we talk more about hope and wellness and resiliency, as opposed to necessarily focusing on stories of people who died?”
She knows that for every person who attempts suicide, the statistics barely capture the countless others who are consumed by the thought, even if they never act on it.
Finding support for those lives, as well as curbing the number of suicide victims, demands an approach that expands concern and responsibility beyond the individual — and toward the entire community.
Tips From Goldstein Grumet: How to talk to a loved one about suicide
Be direct and ask if someone is considering self-harm. People hoping to offer support often ask her, “What if I plant the idea?” This is a myth, she says. “You don’t put the idea in somebody’s head. You don’t make them angry. But we do know that people often are not going to tell you unless they’re asked directly.”
How we frame the question can make a difference: “We need to ask in a way that promotes you really want to know the answer.” You can reduce an individual’s anxiety by establishing a safe space to talk about suicide; this can help them feel less alone. So it’s best to avoid phrasing the question with judgment, such as: “You’re not thinking of killing yourself, are you?” This tone doesn’t create an inviting environment.
“Our job, if we’re opening up this dialogue, is not to offer solutions,” Goldstein Grumet says. “You can offer to be there, to listen and support them, and even go to an appointment with them.” Or, help them book an appointment or get in touch with a professional.
Support and Resources
As of July 16 this year, anyone in the U.S. can simply dial 9-8-8 to reach the National Suicide Prevention Lifeline and receive free emotional support, offered 24/7. This new phone number — which some phone providers began using in 2021 — replaces the 10-digit suicide crisis number (800-273-8255) that has been in place since 2005 under the U.S. Substance Abuse and Mental Health Services Administration (SAMHSA). The suicide prevention lifeline connects people in need with a network of 180-plus local crisis centers across the nation.
The Trevor Project offers 24/7 crisis support tailored for LQBTQ+ youth. Call or text a trained counselor immediately via thetrevorproject.org/get-help.
The Veterans Crisis Line offers crisis support for veterans via text message, as well as phone services by dialing the 9-8-8 number, then pressing 1. Visit veteranscrisisline.net for more resources.
For general info on mental health and substance abuse, or to locate a treatment center in your area, contact SAMHSA’s treatment referral helpline: 877-726-4727.
Who Is Affected?
U.S. groups with elevated risk of suicide, per data from the CDC:
Men experience a suicide rate 3.7 times higher than women.
American Indian and Alaska Native populations face the highest suicide rates in the country — 60 percent greater than the general population.
Among veterans, the suicide rate is 52 percent higher than the general adult population.
People living in rural areas have a suicide rate 43 percent higher than urban populations.
LGBTQ youth and adults report significantly more suicide attempts than their heterosexual and cisgender peers, though the specific suicide rate is unknown because death records don’t include sexual orientation and gender identity.