Lead. A new prospective study from Japan adds strong evidence that sleep disturbances are an independent risk factor for suicidal thoughts, and that the link persists even after controlling for depression, a known confound. The findings suggest that treating sleep problems could be a distinct and modifiable pathway for suicide prevention.
Published in Archives of Suicide Research, the study followed 652 adults over three months and found that poor sleep quality and longer time to fall asleep predicted suicidal ideation at follow-up, regardless of depressive symptoms.
What they found. Researchers led by Akiyoshi Shimura at Tokyo Medical University administered an online survey to a stratified random sample drawn from a web panel and supplemented by local city government officers in Japan. Participants completed the Pittsburgh Sleep Quality Index (PSQI), the Patient Health Questionnaire-9 (PHQ-9) for depression, and the Suicidal Ideation Scale (SIS) at baseline and again three months later.
The sample averaged 49.7 years of age and was roughly balanced by sex. At baseline, 25.2% reported weak suicidal ideation (endorsing the item “wish to die” on the SIS), while 3.7% reported moderate-to-strong suicidal ideation. Clinically significant sleep disturbance (PSQI global score of 6 or higher) was present in 32.8% of participants, nearly one in three.
Using cross-lagged panel models, the team examined the temporal direction of the sleep-suicide relationship. They found that the PSQI global score and the sleep latency component (how long it takes to fall asleep) prospectively predicted suicidal ideation at three months. Critically, these associations remained statistically significant after adjusting for PHQ-9 scores, meaning the sleep-suicide link was not simply driven by depression, a long-standing explanation for the correlation.
The reverse direction also held: baseline suicidal ideation predicted worse sleep at follow-up, suggesting a bidirectional relationship. Sleep disturbance and suicidal thinking appear to reinforce each other over time, creating a cycle that may escalate risk.
In univariate analysis, the odds ratio for sleep problems predicting suicidal ideation was 3.97, meaning people with poor sleep were nearly four times as likely to report suicidal thoughts. After adjustment for depressive symptoms, the odds ratio dropped to 1.59 but remained statistically significant. This pattern signals that while depression accounts for much of the association, sleep problems contribute independent risk above and beyond what depression alone explains.
Why it matters. Suicide is a leading cause of death worldwide, and identifying modifiable risk factors is a public health priority. Depression is the strongest known predictor, but it is not the only one, and not all people who die by suicide are depressed at the time. Sleep disturbance offers a potential intervention target that has received comparatively little attention in clinical suicide prevention protocols.
Sleep disturbance is common, measurable, and treatable. Cognitive behavioral therapy for insomnia (CBT-I), chronotherapy, and even simple sleep hygiene interventions are relatively low-cost and scalable. If sleep problems contribute to suicidal ideation through pathways independent of depression, possibly through impaired executive function, emotional dysregulation, or increased impulsivity, then improving sleep could reduce suicide risk even in people whose depression is adequately treated.
The bidirectional finding also has clinical implications. Patients presenting with suicidal ideation may develop or worsen sleep problems as a consequence, setting up a self-reinforcing loop that clinicians should monitor and interrupt. Routine assessment of sleep quality in psychiatric and primary care settings could identify patients at elevated risk who might otherwise be missed by depression screening alone.
The study’s use of a cross-lagged panel design strengthens the case for causality compared with cross-sectional surveys, though the observational design still cannot prove causation. The structured temporal ordering of measurements, sleep at baseline predicting suicidal ideation at follow-up, independent of baseline depression, provides stronger evidence than earlier cross-sectional work.
Limits. The sample was drawn from a web panel and municipal officers in Japan, which limits generalizability to other populations. The three-month follow-up window is relatively short, and the study did not examine actual suicide attempts or deaths, only suicidal ideation. Self-report measures carry inherent bias. The researchers note that unmeasured confounders, such as anxiety disorders, substance use, or chronic pain, could influence both sleep and suicidal thinking. The PSQI assesses subjective sleep quality, not objective measures like polysomnography.
Bottom line. Sleep problems are a prospective predictor of suicidal ideation independent of depression. Clinicians should assess sleep quality in patients at risk for suicide, and sleep interventions should be investigated as a potential suicide prevention strategy.
Source. Shimura A, Iida M, Takada Y, Furui Y, Inoue Y. Sleep problems predict suicidal ideation: A three-month prospective study. Arch Suicide Res. 2026 Jul 19:1-21. doi: 10.1080/13811118.2026.2703972. PMID: 42472476.