Predicting Suicide

Simple models, better predictors • Lonely teens and their parents • Low self-esteem at any age • Treating the underlying cause

If you or someone you love is struggling with thoughts of suicide, help is available. View the suicide hotline for your country or call your local emergency number for immediate assistance.

Globally, over 720,000 deaths by suicide occur every year. In the United States, the rate of suicide has been trending upwards for the last 20 years—suicide was one of the leading causes of death in 2020.  

In 2021, the U.S. Surgeon General and National Action Alliance for Suicide Prevention (NAASP) jointly released a Call to Action that emphasized the need to improve risk identification. To address this, researchers are trying to home in on the traits that best predict suicidality and find ways to model risk more accurately and easily. Scientists worldwide are also trying to sort out distinctions between suicidal thoughts (including thinking about or considering suicide, also called suicidal ideation) and suicidal behaviors (including attempts—even aborted or interrupted ones).  

“While suicidal ideation is a strong predictor of suicide, most individuals who experience suicidal thoughts do not engage in suicidal behavior,” wrote Susanne Buecker (Witten/Herdecke University) and colleagues in a 2025 Clinical Psychological Science paper.  

For those who do move from having suicidal thoughts to acting upon them, the transition can be quick. Studies have shown that although people may have suicidal thoughts for years, the time between deciding to act and actually doing so is usually only a few hours and sometimes only minutes (Millner et al., 2017). Because the risk for suicidal behavior varies across such short time periods, identifying the window when someone is most vulnerable is critical.  

Teenagers and young adults are particularly vulnerable. Suicide is the third leading cause of death among 15- to 29-year-olds (World Health Organization, 2025), and there is a sharp rise in suicidal behavior in late childhood to early adolescence. Despite this, children and adolescents have been relatively understudied compared to adults. Psychological scientists are trying to fill that gap to identify moments when teens are most at risk and ways to prevent suicide outside of the traditional approaches. 

Simple models, better predictors 

To help identify models that could better predict youth suicide attempts or suicide, Lauren O’Reilly, a postdoctoral fellow at Indiana University School of Medicine, examined 209 predictors of suicide in over 34,000 Swedish 9- to 18-year-olds who had been in contact with a mental health specialist in the last year. The predictors went beyond those usually seen in medical reports to include nonpsychiatric risk factors such as familial income, chronic conditions, and living in an area marked by poverty and unemployment. 

Related Research Topic: Understanding and Preventing Suicide

In the resulting 2024 Clinical Psychological Science paper, O’Reilly reported that the simpler models performed similarly to the more complex models, and machine-learning methods did not substantially outperform traditional logistic regression models. The results suggest that prior suicidality, psychiatric disorders and medications, and demographics (particularly age and female sex) may be sufficient to develop models that perform well.  

“Our results specifically can help people at risk by highlighting relevant risk factors to consider when working with youth who may be at risk,” said O’Reilly in an interview with the Observer. “It is essential to screen for past and recent suicide attempts, prior self-harm, and suicidal ideation when initiating care and regularly throughout therapy… Even if we aren’t using these complex models to predict risk, we can still help providers ask more standardized questions to inform risk assessments and tailor treatment.” 

Lonely teens and their parents 

Even with a good understanding of who is at risk for self-harm or suicide, researchers haven’t actually identified “that exact moment when an intervention could save a life,” said Julie Janssens, a postdoctoral researcher at KU Leuven, in an interview with the Observer. She studies self-harm, which includes both nonsuicidal and suicidal self-injurious behaviors. The distinction between these behaviors is not always clear and people may switch between them, she said. Self-harm thoughts and behaviors peak in adolescence, and self-harm is strongly associated with suicide. 

Janssens hypothesized that loneliness might be a momentary factor, as individuals who report more loneliness are at greater risk for suicide. She wondered, “Is loneliness really important or a critical warning signal for self-harm thoughts in the next minutes to hours?” 

In an upcoming article in Clinical Psychological Science, Janssens studied 1,602 Flemish adolescents. Via smartphone prompts, participants were asked to rate loneliness and self-harm thoughts and behaviors multiple times a day for several days. Higher momentary levels of loneliness were significantly associated with more intense self-harm thoughts within the next 90 minutes. If participants engaged in self-harm behaviors, loneliness in that 90-minute interval decreased, suggesting self-harm behaviors may provide immediate, if not lasting, relief. 

Related content: The Emerging Science of Suicide Prevention

Understanding loneliness as a potential warning signal for self-harm thoughts or behaviors can inform interventions. For example, a just-in-time adaptive intervention (JITAI) approach could use a smart phone to provide a live intervention whenever someone feels lonely and help them cope with those feelings rather than engage in self-harm. But Janssens doesn’t want to make teenagers more dependent on their smart phones. Instead, she wants to stop the loneliness by helping them connect more directly with people. And her study provided evidence for fostering one particular connection: that between teenagers and their parents.  

Janssens found the associations between loneliness and self-harm thoughts and behaviors were moderated by the adolescents’ attachment to their parents. Teenagers who had lower-quality attachment bonds with their mothers had both more frequent and intense self-harm thoughts when lonely. Additionally, teenagers who didn’t have strong attachment to their fathers found more relief from loneliness through self-harm behaviors. 

“When they feel lonely and they have a good relationship with their mother, that helps them to not go over to those self-harm thoughts,” said Janssens. On the other hand, “How they regulate feelings of loneliness via self-harm is maybe something they learn through that other parent, which is most often the father.”  

Janssens cautioned that these roles are not necessarily gender-based, just that two parents may each play different roles. She also stressed that although clinicians often work with the mothers, fathers also have a unique and important role and should be included in therapeutic conversations. 

Janssens found these results promising because attachment is modifiable and interventions can repair trust between adolescents and their parents. She suggested being responsive and setting clear boundaries.  

“Be there for all the emotion that comes with that and be curious, attuned to what your teenager is interested in,” she said. 

The results also show that parents still have a lot of influence during adolescence, even as peers and smartphones seemingly pull them away.  

“There is just a lot of value in having at least one good relationship with one of your parents,” she said. “That really can buffer against negative influences such as abuse or bullying or other adverse experiences or loneliness.” 

Low self-esteem at any age 

Like loneliness, low self-esteem has also been linked to high levels of suicidal thoughts and behaviors, but whether self-esteem plays a different role in the ideation and execution of suicidal behaviors is poorly understood. Additionally, self-esteem changes over the lifetime, generally dropping during adolescence, peaking in middle age, and falling again in old age. The self-esteem slump in adolescence corresponds to a sharp rise in suicidal behavior, suggesting the relationship between self-esteem and suicidal thoughts and behaviors may be moderated by age.  

To examine this, Buecker and colleagues performed a meta-analysis of 114 studies from around the world with subjects ranging from children to older adults. The research revealed strong links between self-esteem and suicidal thoughts, indicating that people with low self-esteem are at a higher risk for contemplating suicide. There was also a strong negative association between self-esteem and suicidal behavior. However, few studies looked at suicidal behavior, so the authors warned those effects must be interpreted cautiously and called for more research.  

Buecker’s study did not find any significant age or gender differences—suggesting, as the authors wrote, that “low self-esteem seems to be a rather global associate of [suicidal thoughts and behaviors], regardless of how old persons are or what gender they have.” 

“Overall, the results of this meta-analysis have important clinical implications, as enhancing self-esteem may be a promising strategy to prevent [suicidal thoughts and behaviors],” the authors wrote. “Clinically, the … results underscore the need to ask for perceptions of low self-esteem and being a burden to others in risk assessments with suicidal patients.”  

The authors suggested clinicians challenge patients’ low self-esteem with evidence gathering and discussion of cognitive distortions. They also noted that self-esteem and related beliefs like perceived burdensomeness can be malleable and that cognitive behavioral therapy shows the largest effects on increasing self-esteem in adult populations.  

Treating the underlying cause 

Just as therapy can improve parental attachment or self-esteem might make a meaningful dent in suicidal thoughts and behaviors, treating chronic, recurrent diseases may be a better approach in some cases than focusing on suicide-specific treatments. For severe cases, the primary goal is to keep the patient alive, and a suicide-specific treatment makes sense. But when the issue causing the most distress and impairment is another clinical diagnosis, treating that first may be a more effective approach.  

“You really have to think about what’s driving the suicidality in the first place,” Morgan Robison, a graduate student in the lab of APS James McKeen Cattell Fellow Thomas Joiner at Florida State University, said in an interview with the Observer. “If you don’t actually treat the underlying cause, that person likely will have those thoughts coming back.”  

Rates of suicide attempts are higher, for instance, in people with eating disorders; at least one quarter to one third of individuals with eating disorders have experienced suicidal thoughts in their lifetime. For anorexia nervosa, suicide risk is increased by about ninefold. “Most folks are either dying from this condition of eating disorders or they’re dying by suicide,” said Robison. 

To investigate whether eating disorder–specific treatment could also reduce suicidal thoughts, Robison led an assessment of around 3,500 eating-disorder patients before and after treatment. The results, published in a 2024 study in Clinical Psychological Science, showed that suicidal thoughts decreased from 51% of the patients at the start of treatment to 38% afterward. Rates of severe suicidal thoughts (such as daily thoughts of death or self-harm) decreased by 44%, from 12.7% to 7.1% of patients.  

Robison thinks this approach holds promise for other disorders with increased risk for suicide, including post-traumatic stress disorder, insomnia, borderline personality disorder, depression, panic disorder, and chronic pain. Treating only suicidality would do nothing for the underlying issues, whereas Robison’s study shows targeting the disorder may also increase someone’s desire to live.  

“We’re moving the needle some by addressing just core symptoms,” said Robison. 

But this research is focused on suicidal thoughts, she noted. As the other studies show, the difference between thoughts and behaviors is crucial.  

“What does it mean for behavior? That’s really the big question that I think needs to be addressed,” she said. “I’m hopeful that we can continue to explore this further.” 

Back to top

Feedback on this article? Email [email protected] or login to comment.