A 16-year-old boy, a high-school athlete with good grades, told his therapist that he was thinking about taking his own life. That therapist, Dennis Kolsch, got him admitted to an inpatient ward. “He didn’t have a great experience in there, but he was safe,” says Kolsch, a licensed mental-health counselor in Cocoa Beach, Florida. “The family felt comforted knowing that.”
Teens leaving an inpatient program like this one will have discharge instructions on how to continue care, which usually include medications and psychotherapy. The boy was discharged to Kolsch’s care, but Kolsch knew that weekly or biweekly therapy sessions were not enough. So he worked on getting the boy into an intensive outpatient program.
In the meantime, his parents were frantic. They didn’t want to let their son out of their sight, and felt they had to re-create the hyper-controlled structure of the hospital setting. It was all-consuming and exhausting. Further, the constant supervision was not helpful for the parent-child dynamic, which had been bumpy before the hospitalization and was now ramping up again. “The mom’s becoming overbearing and the son is withdrawing,” Kolsch says. “And then the mom gets worried because the son is withdrawing.”
Teens who have been hospitalized for a suicide attempt or suicidal ideation are at heightened risk of dying by suicide. A 2007 study, for instance, followed nearly 5,000 young people, from 15 to 24 years old, who sought care at a single hospital for “deliberate self-harm” over a 20-year period. Nearly 3 percent of the study’s subjects died, and more than half of those deaths were likely suicides—a rate 10 times higher than would be predicted for this age group. And the increased risk, research shows, can persist for years.
“We know that transition out of inpatient care is a particularly high-risk time period for suicide and subsequent suicide attempts,” says Michele Berk, a clinical researcher at Stanford University.
All of this suggests that where hospitalization provides effective crisis management in such situations, keeping young people safe back at home is a challenge that modern medicine has so far failed to solve. But a group of researchers at the University of Michigan has been working with a simple yet powerful tool that just might help: recruiting three or four familiar adults—not just the young person’s parents—who pledge ongoing support through recovery. The Michigan program trains both family and friends to become dedicated helpers and empathetic listeners—and to encourage their struggling charges to stick to the treatment plan.
The program is unique in both its approach and its results. The intervention is entirely focused on the adult volunteers, not on the child. (The teen’s only role is naming trusted adults.) And in a recent paper by the psychologist Cheryl King and her colleagues reporting a decade-long follow-up of teens in the program, those who received the attention of trained adults in their life were nearly seven times less likely to die than teens who received only standard care. The study was one of the largest suicide-intervention studies ever done, and it is the first clinical trial for suicide prevention in high-risk teens that found a change in death rates.
“There has never been a study that shows a reduction in actual deaths,” says Dennis Ougrin, a child and adolescent psychiatrist at Maudsley Hospital in London. That’s true regardless of the type of intervention, Ougrin says, whether it’s a medication or psychotherapy or purely social, as this one is. “It’s very exciting.”
This sort of “social prescribing” is too often the overlooked stepchild of mental-health treatment, experts say, even though most psychologists believe that social connectedness is vital to psychological health. And in the context of rising suicide rates among American teens—alongside the failure of most interventions to affect long-term outcomes in high-risk teens—efforts to boost social connectedness are now getting some new attention from mental-health professionals.
Of course, King’s results would need to be repeated in future studies to be fully corroborated, and there’s no call to adopt this type of intervention more widely—even by King and her colleagues. (The new paper is a secondary analysis of results measured 10 years after the study. That’s one reason King is cautious about her results.) But given the early signs and taken alongside other social-centric therapeutic approaches being used by researchers at Stanford and elsewhere, some experts say a potent tool in combatting teen suicide might have been hiding in plain sight.
“It’s kind of recognizing we’ve missed a trick,” says Martin Webber, a professor of social work at the University of York, “in terms of utilizing the assets that are available within people’s communities.”
Historically, it has been extremely difficult to show a change in suicide rates with enough statistical clarity to conclude a true change over chance or coincidence. That’s because actual rates of suicide are low—even in high-risk groups, such as teens with a history of self-harm and hospitalization. This means that researchers need huge sample sizes to detect a true change.
King and her University of Michigan team enrolled 448 people. About half—223—were prescribed coordinated support from friends and family on top of standard care, while 225 received only standard care. (Standard care consisted of psychotherapy and medication.) The combined cohort was large enough to detect a difference in overall death rates, though still insufficient to find a statistically significant difference in deaths attributable specifically to suicide: There was one known suicide among the intervention group, versus three in the control group.
But when King analyzed not just confirmed suicide deaths but also drug-overdose deaths that were not labeled “accidental,” a pattern emerged. There was just one of these in the intervention group, versus eight in the control group—a statistically significant difference. All deaths occurred in adulthood, when the subject was at least 18.
The study made a splash on Twitter with prominent experts in the field. Peter Kramer, who wrote the best-selling book Listening to Prozac, called the study a must-read. Allen Frances, who chaired the task force that authored the latest diagnostic manual for psychiatric disorders, wrote: “We’ve learned so very much, but still fail miserably in doing the simple things well.”
King started as a faculty psychologist in the adolescent psychiatric inpatient unit at the University of Michigan hospital system in 1989. “More than half of the adolescents hospitalized were there because of suicide risk,” she says. That clinical experience led her to develop the new intervention.
In King’s approach, teens nominate trusted adults—for example, parents, grandparents, aunts, uncles, family friends, teachers, or clergy—to serve as a support team. (Parents have veto power.) The adults then get an hour-long training session and weekly phone calls from King’s intervention specialists to talk about how things are going. They are cautioned to not feel responsible for the teen’s behavior—“We’re not asking them to be mental-health professionals,” King says—but they agree to check in with their teen once a week by phone, a face-to-face meeting, or an outing.
In the training session, which King calls “psychoeducation,” the adults learn about their teen’s situation—the specific diagnosis, the treatment plan, and the rationale behind them. They learn what to do in case of an emergency, and how to be a nonjudgmental shoulder for the teen to lean on. Training sessions are variable and flexible, to satisfy the needs of the people in the room. “It’s kind of an open discussion,” King says.
“A lot of it is answering their questions,” she adds.
The education and phone support arm the adults to act as informal caregivers, to stand up and support a child they know and who is at risk. Would this happen without training? It might, King and other experts suggest, but it’s easy to see why it might not. Suicide is scary and upsetting, and adolescents can be difficult to talk to. It’s daunting to take responsibility for something like that.
The goal of King’s program, she says, is to make taking on that role less daunting.
As initially conceived, the study did not include deaths as a main outcome measure. That’s because of the numbers problem—suicide rates are so low that it’s practically impossible to show intervention effects. King did what most suicide researchers do: She measured outcomes thought to be related to suicidal behavior. Does a particular intervention reduce thinking about suicide or self-harm? Does a treatment program help teens function day to day?
Even with these more malleable measures, there are only a handful of randomized controlled trials—considered the gold standard of study design—and very few of them show any impact on suicide-related behaviors, says Ougrin, who reviewed the scientific literature in 2015. Two other reviews from the same year pointed to the benefits of involving families and social support.
The family component is crucial, says Stanford’s Berk, who was not involved in King’s study. “We don’t know yet exactly what factors lead a person at risk to attempt suicide or die from suicide at any given time, and we don’t know yet exactly how to help people best through treatment,” Berk says. “Sometimes the most effective thing I think we have is the parents and their ability to create a safe environment around the teen, in terms of restricting access to lethal means, close monitoring of the teen, being the one who can call 911.”
Berk has worked with a psychotherapy intervention called dialectical behavior therapy, or DBT. In a 2018 study, she reported significant effects on suicide attempts and self-harm with the intervention, which also included a dedicated family component. The key thing, Berk says, is being nonjudgmental. “One of the assumptions [in DBT] is that the teen is doing the best they can … So if they’re screaming at you or if they have hurt themselves, we’re not going to judge that,” she says. “We’re going to say, ‘Okay, in that moment they were operating at the top of their skill set.’ And our job is to teach them more skills so they can engage in more effective behavior.”
In addition to adjusting parents’ mind-set, Berk’s intervention helps mitigate the ways that family can be harmful. “Family conflict is a risk factor for suicide in teens,” she says. “And family cohesion is a protective factor.”
Kolsch, the Florida therapist, agrees that King’s intervention seems promising. “I think it’s a pretty brilliant approach,” he says. In addition to supporting teens who face so much risk when they go home, he says, it helps reduce the anxiety and helplessness of families. (Kolsch’s patient—the boy he had admitted to inpatient care—is now in college and doing well.)
All the participants in King’s study received standard psychotherapy and medications, and these mainstays likely contributed to improvements observed in both the intervention and the control groups. “It is difficult to change youths’ trajectories,” she says.
“We were just trying to get an incremental benefit from a small add-on intervention,” she adds.
King says her intervention team long felt like they were having meaningful impacts on families’ lives, but they didn’t observe measurable changes in any potential predictors of suicidal behavior, such as ideation. So how to explain the big effect on death rates? “You know, small effects can have ripple effects,” King says. Perhaps the supporting adults facilitated teens sticking to treatment plans, she suggests, or maybe they helped teens make one or two better behavioral choices.
The University of York’s Webber sees two key elements working together. “There’s the direct effect of social support on that person’s mental health. We know that from existing studies that has an effect,” Webber says. “But there’s the indirect effect of engaging with treatment, which is obviously happening as well.”
Past research has shown that people with more connections, stronger social networks, and more social support will be better off in terms of mental health. And yet it’s one thing to say people who have more social connections are less likely to die by suicide—it’s quite another to create and foster those social connections. On that front, Webber notes that the teens in King’s program were able to choose who they wanted supporting them. “We know from relationships that where people are foisted upon them—and this often happens in professional relationships—it kind of nullifies that as a source of support,” he says.
To be sure, the scientific literature on social prescribing remains tiny compared with the research on medication and psychotherapy. Several thousand children at risk of suicide have participated in various studies of pharmacological and psychological therapies. “We don’t have the evidence base of the social-support-type interventions to put alongside those,” Webber says. Also, standardizing social interventions is tricky, he adds, especially when compared with the simple act of taking a pill.
“Sometimes the parameters are a bit broader,” Webber says.
At the same time, Webber points out that social components are always present. “The role of the family support, the role of friends, the role of people who are not necessarily professionally qualified to deliver any therapy, care, or support”—these things are active ingredients in any person’s treatment. And even in a research study, these social components can contribute to positive effects in both intervention and control groups, though they’ve traditionally been overlooked by researchers seeking to measure the benefits of drug treatment or psychotherapy alone.
According to Berk, suicide prevention has got to be comprehensive. Rather than just providing psychiatric treatment, she says, there are issues that need to be addressed at the community and social-relationship level.
“The more fronts of intervention we have,” Berk says, “the better.”