Beyond Suicide Risk: From Prediction to Precision Prevention

A Functional Psychiatry Perspective on the Dynamic Suicidal State

September is National Suicide Prevention Awareness Month, a time to acknowledge the tragic reality that far too many lives are lost to suicide. Despite decades spent identifying risk factors and refining our ability to predict who is at risk, suicide remains notoriously difficult to predict at the individual level. Perhaps part of the problem is that we have focused too heavily on who is at risk rather than when a particular person becomes at risk.

Emerging research increasingly supports viewing suicidality as a dynamic state—one that can fluctuate over hours or days as biological vulnerability interacts with psychological distress, interpersonal events, environment, sleep,diet, substance use, medical illness, pain, and other rapidly changing variables. This is where functional psychiatry has an important contribution to make. By asking not only What diagnosis does this patient have? but also What has changed in this patient’s biological, psychological, and environmental system?, a functional approach may help identify modifiable factors that intensify the suicidal state—and provide additional opportunities for intervention.

In 2025, nearly 49,000 Americans died by suicide—approximately one person every 11 minutes. Suicide is the 11th leading cause of death overall and the second among adolescents and young adults, while the highest rates occur among older men. After rising substantially during the first two decades of this century, U.S. suicide rates remain near historically high levels.

Beyond the Suicide Risk-Factor Checklist

When we hear the word suicide, our thoughts often turn immediately to risk factors—or to the urgent challenge of assessing and triaging an acutely suicidal patient. And even a cursory review of recent headlines continues to expand the list of vulnerabilities associated with suicide:

• Teens Who Don’t Get Enough Sleep May Face a Higher Risk of Suicide Attempts
• Knowing Someone Who Recently Attempted Suicide May Raise a Young Person’s Own Risk
• Loneliness in Older Adults May Be an Overlooked Risk Factor for Suicide
• A New Cancer Diagnosis May Create a Critical Window of Increased Suicide Risk
• Chronic Pain Carries a Hidden Mental Health Risk: Suicidal Thoughts Are Surprisingly Common

As psychiatrists, we know the drill: assess depression, hopelessness, previous attempts, substance use, psychiatric diagnosis, family history, recent losses, access to lethal means, and suicidal thoughts, intent, and planning. These observations are synthesized into an estimate of low, moderate, or high risk.

The model becomes: Assess → stratify risk → hospitalize the “high-risk” patient.

Yet suicide remains notoriously difficult to predict at the individual level. Contemporary systematic reviews suggest that advanced algorithms can categorize populations into risk tiers, but external validation remains limited, risk of bias persists, and positive predictive values for individual attempts remain discouragingly weak.^(1,2)

Psychiatry has spent decades trying to become better at predicting suicide risk. That work should continue. But perhaps prediction has occupied too much of our attention. The more clinically useful question may be whether we can recognize the transition into a suicidal state early enough to intervene—and whether we can rapidly alter that state.

From “Who Is at Risk?” to “When Is This Patient at Risk for Suicide?”

Suicidality may be better conceptualized as a dynamic state arising from the interaction of underlying vulnerability with rapidly changing biological, psychological, interpersonal, and environmental factors.The most useful clinical question may therefore not be:“How high is this patient’s suicide risk?”but rather:“What changes when this particular patient enters a suicidal state?”

Understanding suicidal ideation in real time has become a growing focus of suicide research, including increasing use of ecological momentary assessment (EMA). Although methodological limitations remain, EMA reflects an important conceptual shift toward measuring within-person fluctuations rather than relying solely on static risk characteristics.⁽³⁾

Functional psychiatry adds another dimension by asking which potentially reversible biological or environmental factors may be contributing to a change in state. These factors may not independently cause suicidality, but they may alter mood regulation, cognition, impulsivity, and resilience.Rather than searching for a single biological cause of suicide, the functional approach asks whether modifiable contributors are increasing vulnerability in this patient, at this particular time.

Turning the Dimmer Switch: Chronic Suicidality and Self-Harm

This dynamic-state model is particularly relevant for patients with chronic suicidal ideation or nonsuicidal self-injury (NSSI). For these patients, suicidal ideation can be thought of as a dimmer switch, with a range of intensity and risk. Helping patients recognize variation in their own state opens treatment possibilities that are more state- than trait-dependent.

Many patients are reluctant to disclose suicidal ideation because they fear that doing so will automatically result in hospitalization. A therapeutic partnership should allow patients to disclose changes in their suicidal state knowing that the response will be thoughtful and proportional—while recognizing that a higher level of care remains necessary when safety cannot otherwise be maintained.

The dimmer-switch analogy invites another question: What is turning up the intensity? A patient whose chronic suicidal ideation shifts abruptly from a “3” to an “8” may be responding to an interpersonal crisis, but a biological or behavioral change may also be contributing.These factors do not replace traditional psychiatric assessment. They enlarge it.

NSSI deserves similar nuance. Clinicians can easily panic over self-harm or, conversely, minimize it because it is not a suicide attempt. Yet NSSI is an important marker for future suicidal behavior.

Self-harm serves a function that varies among individuals.⁽⁴⁾ Affect regulation is the most commonly endorsed function, but NSSI may also interrupt dissociation, communicate distress, elicit support, or provide escape from overwhelming circumstances. Paradoxically, although NSSI is associated with increased future suicide risk, for some individuals it may simultaneously function as a short-term strategy for avoiding a suicide attempt. Understanding what the behavior accomplishes for the individual becomes essential to replacing it with something safer.

Treating Suicidality as a Target

Traditionally, suicidality has often been viewed as downstream from another disorder: treat the depression, bipolar disorder, psychosis, PTSD, or substance-use disorder, and suicide risk should improve.That remains essential—but it may not be sufficient.

If suicidal thinking can intensify rapidly, waiting for the underlying disorder to improve may leave a critical window of vulnerability untreated. The goal becomes not only to treat the diagnosis but also to identify and interrupt the suicidal state itself.

The emerging paradigm is: Identify vulnerability → detect state change → intervene rapidly → reduce access to lethal means → directly treat suicidality → maintain contact while the state changes.

Several existing treatments demonstrate that suicidality can be targeted more directly, although they affect different dimensions of suicide risk.

Lithium has one of the strongest and longest-standing evidence bases for suicide prevention in mood disorders. A landmark meta-analysis of 48 randomized trials found significantly fewer suicides with lithium, and observational studies have similarly demonstrated lower rates of suicide attempts and deaths.^(5,6) More recent meta-analyses continue to show fewer suicidal outcomes with lithium, although differences have not consistently reached statistical significance.^(7,8)

Ecological studies have also reported lower suicide rates in communities with higher naturally occurring lithium concentrations in drinking water, although findings are inconsistent.^(9–11) These observations are provocative and although do not establish that trace lithium prevents suicide, clinical experience with low dose nutritional lithium has been observed to be impactful particularly in high risk patients with family histories of suicide, substance abuse and significant mood disorder.

Many patients and providers remain reluctant to consider lithium because of concerns about renal and thyroid toxicity. Clinicians can help patients understand the risk-benefit ratio and individualize dosing and monitoring. Lower-dose strategies may reduce dose-related adverse effects and retain positive benefits.

Clozapine has the strongest evidence for reducing suicidal behavior in schizophrenia and schizoaffective disorder. In the landmark InterSePT trial of 980 high-risk patients, clozapine significantly reduced suicidal behavior compared with olanzapine.⁽¹²⁾

Ketamine, in contrast, is notable for its rapid reduction of suicidal ideation, sometimes within hours. Recent systematic reviews and meta-analyses support a rapid but generally time-limited anti-suicidal-ideation effect.^(13,14) Importantly, evidence is considerably stronger for reducing suicidal thoughts than for preventing suicide attempts or deaths.

These treatments are very different, but collectively they make an important point: suicidality is not simply an outcome we monitor while treating something else. It can itself be a target of treatment.

The pharmacology of suicide prevention contains an important paradox. Antidepressants carry an FDA boxed warning for increased suicidal thoughts and behaviors in children, adolescents, and young adults, and atomoxetine carries a boxed warning for suicidal ideation in pediatric patients. Antiepileptic medications, including several widely used as mood stabilizers, also carry an FDA warning regarding suicidal thoughts and behavior. These warnings do not imply that these medications increase suicide mortality overall, nor should they discourage appropriate treatment. Rather, they underscore the importance of monitoring for  changes in suicidal thinking particularly following medication initiation or dose changes.

The Biology of the Suicidal State: A Functional Psychiatry Perspective

If the suicidal state is psychologically dynamic, might it also be biologically dynamic?

A major 2026 Molecular Psychiatry review examining biological findings associated specifically with suicide—not simply with underlying psychiatric diagnoses—identified converging abnormalities involving stress-response biology, inflammation, neuroplasticity, and serotonergic function. These systems appear interconnected, with genetic vulnerability and environmental adversity further shaping their regulation.⁽¹⁵⁾

Importantly, there is no clinically validated biomarker for suicide. The emerging picture is one of systems biology, in which biological vulnerability interacts with environmental exposures and acute stress.This framework aligns naturally with functional psychiatry. Rather than searching for a single biomarker or “cause” of suicidality, we can ask which interacting—and potentially modifiable—systems may be contributing to vulnerability at a particular moment.

Several biological correlates are intriguing.

Lipid status is one example. A meta-analysis of 65 studies involving more than 500,000 participants found lower total cholesterol and LDL cholesterol among suicidal patients, with lower total cholesterol associated with increased risk of suicide attempts and deaths.⁽¹⁶⁾ These observational findings do not establish that low cholesterol causes suicide or that raising it prevents suicide, but they suggest that lipids are  an easily modifiable risk factor – a biological variable traditionally considered outside psychiatric assessment which may contribute to the terrain in which suicidal states emerge.

Vitamin D provides another example. A 2025 meta-analysis involving more than 500,000 participants found lower vitamin D levels among individuals with suicidal behaviors, although categorical vitamin D deficiency did not significantly predict suicide, suicide attempts, or suicidal ideation.⁽¹⁷⁾ Observational data have also associated vitamin D supplementation with lower rates of suicide attempts and intentional self-harm, but randomized trials are needed before supplementation can be considered a suicide-prevention intervention.⁽¹⁸⁾ Although vitamin D has been studied in RCTs for depressive symptoms and is generally seen as a modifiable risk factor for depression, it has not been established as a suicide-prevention treatment.

The functional psychiatry conclusion is not that vitamin D deficiency, low cholesterol, inflammation, or another isolated abnormality “causes” suicide. Rather, clinically relevant biological abnormalities may be part of the terrain in which vulnerability occurs—and some may be modifiable.

Intervening in the Suicidal State

Medication is only part of direct suicide treatment.

Dialectical Behavior Therapy (DBT) is among the best-established psychotherapies for recurrent self-harm and is particularly important for NSSI, chronic suicidality, emotion dysregulation, and borderline personality disorder. DBT seeks to understand the function of self-injury and replace it with safer strategies through distress tolerance, emotion regulation, mindfulness, and interpersonal effectiveness. Randomized trials and meta-analytic evidence demonstrate reductions in self-harm and suicidal behavior.^(19–21)

DBT therefore fits particularly well with a state-based model: patients learn to recognize escalating vulnerability and intervene before distress progresses to self-injury or suicidal action.

Suicide-focused cognitive behavioral therapy similarly treats suicidal behavior as a therapeutic target. It identifies the individual’s pathway into suicidal crisis and develops strategies to interrupt that sequence. A 2024 randomized trial of telehealth brief CBT for suicide prevention found an approximately 41% reduction in suicide attempts over 12 months compared with present-centered therapy.⁽²²⁾

Other interventions are deceptively simple but important. Safety planning, lethal-means counseling, rapid follow-up, and continued contact help patients recognize escalation, access support, and create time for a suicidal state to change. Evidence is strongest when these strategies are integrated rather than treated as a checklist or discharge document.^(23–25)

Reducing access to a highly lethal method is particularly important because it does not require us to predict perfectly whether someone will attempt suicide. It creates time—for a crisis to change, another person to intervene, or the suicidal state itself to diminish.

The objective is not simply to assess risk and send the patient away with instructions. It is to maintain enough connection and support to help the patient move through a fluctuating state safely.

Toward Precision Suicide Prevention

Our inability to reliably predict individual suicidal behavior suggests that risk stratification alone will never be enough. A more useful model may be to understand suicidality as a dynamic state arising from the interaction of enduring vulnerability with rapidly changing biological, psychological, interpersonal, and environmental influences.

Functional psychiatry has a role in this evolving model—not because we have discovered a laboratory test or nutritional protocol for suicide, but because it asks a fundamentally useful question:What has changed in this particular patient, and which of those changes might be modifiable?

Precision suicide prevention may ultimately depend less on perfecting our ability to classify patients as low, moderate, or high risk and more on recognizing within-person change, directly treating suicidality, and intervening on the factors driving the state.