Overview
Veterans living with cancer face a meaningful and persistent risk of suicidal self-directed violence (SSDV), a term that includes both fatal suicide and nonfatal suicide attempts. A large national study of more than 292,000 veterans with invasive solid tumors or hematologic cancers found that this risk remains elevated not only near the time of diagnosis, but also years later during survivorship.
This study is important because it goes beyond fatal suicide alone. Nonfatal attempts are a major warning sign, a source of serious injury, and often the strongest predictor of future suicide-related harm. By tracking both fatal and nonfatal events, the study provides a fuller picture of suicide risk in cancer care.
Why this study matters
Cancer can affect mental health through many pathways: pain, fatigue, treatment burden, loss of independence, financial stress, fear of recurrence, and changes in family or social roles. For some patients, these pressures may be compounded by preexisting depression, anxiety, trauma, substance use, or social isolation.
Veterans may face additional challenges, including higher baseline rates of mental health conditions, prior trauma exposure, and barriers to timely behavioral health care. Understanding which veterans are at greatest risk can help clinicians screen earlier, intervene sooner, and tailor prevention efforts more effectively.
Study design
This national cohort study followed veterans diagnosed with cancer between January 2014 and December 2023. Investigators used data from oncology and suicide registries as well as the Veterans Health Administration. They analyzed outcomes from January 2025 to February 2026.
The exposure of interest was a diagnosis of invasive solid cancer or hematologic cancer. The primary outcome was SSDV, measured as events per 100,000 person-years. The research team also used multivariable Cox proportional hazards models to estimate adjusted hazard ratios for different risk factors.
Main findings
Among 292,271 veterans studied, there were 2,400 SSDV events, affecting about 1% of the cohort. The overall rate was 203 events per 100,000 person-years.
The most common method of self-harm was poisoning, often involving medications such as opioids, accounting for 617 attempts, or 26% of events. This finding highlights the importance of careful medication review, safe storage, and monitoring of high-risk prescriptions.
Certain cancer types were associated with higher estimated SSDV probability, especially cancers of the central nervous system, pancreas, head and neck, liver and biliary system, and thyroid. These cancer sites may carry a heavier symptom burden, greater functional impairment, or more distressing prognoses, all of which can influence suicide risk.
Groups with especially high rates
Several clinical and social factors were linked to higher SSDV rates than the overall cohort:
Veterans with severe frailty had a rate of 544 events per 100,000 person-years.
Veterans with advanced cancer had a rate of 261 events per 100,000 person-years.
Those with chronic mental illness had a rate of 419 events per 100,000 person-years.
Veterans reporting high pain scores had a rate of 236 events per 100,000 person-years.
These findings reinforce a familiar but crucial point in oncology: physical symptoms and mental health are closely connected. Pain, weakness, disability, and loss of function can worsen hopelessness, especially when combined with depression or social isolation.
Who was at highest risk for nonfatal attempts
The study found especially high rates of nonfatal suicide attempts among:
Younger veterans aged 45 years or younger, with a rate of 643 events per 100,000 person-years.
Female veterans, with a rate of 369 events per 100,000 person-years.
American Indian or Alaska Native veterans, with a rate of 286 events per 100,000 person-years.
Veterans with central nervous system cancer, with a rate of 394 events per 100,000 person-years.
Veterans with thyroid cancer, with a rate of 359 events per 100,000 person-years.
These patterns are clinically important because some of these groups are often underrecognized in cancer survivorship planning. Younger adults with cancer may experience sharper disruption to work, parenting, finances, and identity. Women may face different patterns of distress and help-seeking. Indigenous veterans may also encounter structural barriers to care, historical trauma, and inequities in access to behavioral health services.
Risk in the early months after diagnosis
The first six months after diagnosis appeared to be a particularly vulnerable period for some veterans. Increased SSDV hazard was observed among:
Asian veterans compared with White veterans, with an adjusted hazard ratio of 2.55.
Unmarried veterans, with an adjusted hazard ratio of 1.83.
Veterans with central nervous system cancer compared with lung cancer, with an adjusted hazard ratio of 2.07.
Veterans with head and neck cancer compared with lung cancer, with an adjusted hazard ratio of 1.67.
Veterans with advanced cancer, with an adjusted hazard ratio of 1.30.
The early postdiagnosis period often includes intense uncertainty, treatment decisions, new symptoms, and emotional shock. These data suggest that screening for suicide risk should not be delayed until late-stage disease or obvious psychiatric crisis. It should begin early, ideally at diagnosis and during active treatment planning.
Risk over time
For most veterans, suicide-related risk declined over time after diagnosis. However, the risk did not disappear. Five years after diagnosis, elevated hazard remained for several subgroups, including:
Younger veterans aged 45 years or younger.
Unmarried veterans.
Veterans with central nervous system cancer.
Veterans with advanced cancer.
This long tail of risk is an important survivorship issue. Cancer follow-up care often focuses on recurrence, treatment toxicity, and physical recovery, but mental health surveillance may fade as time passes. The study suggests that behavioral health support should remain part of long-term cancer care, not just the initial treatment phase.
What may explain these findings
The study was designed to identify associations, not prove causes. Still, several plausible explanations exist.
Cancers affecting the brain or nervous system may directly influence mood, cognition, impulse control, and personality through tumor effects or treatment effects. Head and neck cancers can cause pain, difficulty speaking or swallowing, and visible disfigurement, all of which can worsen distress and social withdrawal. Advanced disease and severe frailty are markers of high symptom burden and declining function. Pain itself is a known risk factor for suicidal thoughts and behaviors.
Mental illness, relationship status, and race or ethnicity also matter. Being unmarried may reflect less day-to-day social support. Chronic mental illness can amplify distress during cancer treatment. For some racial and ethnic groups, differences in cultural context, access to care, and treatment experience may affect how risk appears in administrative data.
Implications for cancer care
The practical message from this study is clear: suicide prevention should be part of routine oncology care.
Clinicians may consider:
Screening for depression, suicidal thoughts, and self-harm history at diagnosis and during follow-up.
Paying special attention to younger patients, veterans who are unmarried, and those with brain, thyroid, or head and neck cancers.
Assessing pain, frailty, and psychiatric comorbidity as part of every major oncology visit.
Reviewing medication access, especially opioids and other potentially toxic drugs.
Coordinating closely with mental health specialists, social workers, palliative care teams, and primary care.
Because many attempts involved poisoning, safer prescribing and dispensing practices may be especially valuable. This can include limiting medication quantities when appropriate, involving family or caregivers in medication management, and ensuring rapid access to crisis support.
Clinical and public health significance
The findings support a broader shift in survivorship care. Cancer outcomes are not defined only by tumor control or survival time. They also include psychological safety, quality of life, and the ability to live through and beyond treatment without preventable harm.
For health systems, the study suggests that suicide prevention in oncology should be risk-stratified and longitudinal. High-risk patients may benefit from repeated screening, care navigation, pain management, integrated behavioral health, and clear pathways for emergency intervention.
For veterans, these results are especially relevant because the Veterans Health Administration has the infrastructure to combine oncology, primary care, mental health, and social support. With systematic tracking and proactive outreach, some suicidal behaviors may be preventable.
Limitations to keep in mind
As with any observational study, there are important limitations. Administrative data may miss some risk factors, such as relationship quality, detailed psychosocial stress, or cancer-specific symptom severity. The study also reflects veterans receiving care within the VA system, so results may not generalize perfectly to all populations with cancer.
In addition, associations may be influenced by unmeasured confounding. Even so, the large sample size and national registry linkage make the findings highly informative and clinically useful.
Bottom line
Veterans with cancer face a sustained risk of suicidal self-directed violence, with the highest danger often appearing in the early months after diagnosis but persisting for years in some groups. Younger veterans, unmarried veterans, those with central nervous system or thyroid cancer, and patients with advanced disease, frailty, pain, or chronic mental illness deserve especially careful attention.
The study argues for a simple but important change in cancer care: do not screen only for fatal suicide risk. Track all suicidal behaviors, including nonfatal attempts, and integrate suicide prevention into routine oncology practice from diagnosis through long-term survivorship.
