The warning in the dark: Nightmares may foreshadow suicidal crisis by months
A new study from Paris suggests that nightmares, especially those with suicidal themes, can emerge weeks to months before a person reaches psychiatric emergency care, offering clinicians a potential early warning signal for imminent suicide risk.
Published in Sleep Medicine, the prospective study of 42 patients arriving at the psychiatric emergency department of GHU Paris Psychiatrie et Neurosciences with severe suicidal ideation or after a suicide attempt found that more than 8 in 10 reported some form of dream alteration before their crisis. The findings build on evidence that what happens in the sleeping mind may reveal what the waking mind cannot yet articulate.
What they found
Researchers interviewed patients at the Centre Psychiatrique d’Orientation et d’Accueil (CPOA) at Sainte-Anne Hospital in Paris between March and April 2025. Using a self-administered questionnaire alongside validated instruments for sleep disturbances, anxiety, and depressive symptoms, 83.0% of participants reported at least one alteration in their dream content before the crisis.
Bad dreams, frightening or upsetting dream experiences that do not wake the sleeper, were reported by 61.9% of participants. Nightmares, defined as disturbing dreams that do cause awakening, were reported by 52.4%. Most strikingly, 35.7% reported nightmares with suicidal content: dreams in which they saw themselves attempting or dying by suicide.
The temporal pattern was particularly telling. Nightmares first appeared an average of 128 days, roughly four months, before the suicidal crisis. Suicidal-themed nightmares emerged much closer to the emergency visit, approximately 16 days before. Over half of patients (54.8%) reported an evolutionary pattern, with dream content progressing through at least two successive stages: bad dreams, then nightmares, and in some cases suicidal nightmares.
Nightmare severity also correlated significantly with insomnia severity and anxiety symptoms, suggesting that the relationship between disturbed sleep and suicide risk may involve a mutually reinforcing cycle.
Why it matters
Suicide remains a major public health challenge worldwide. Identifying individuals at imminent risk before they act is one of the most difficult tasks in clinical psychiatry. Current screening tools in emergency departments rely heavily on direct questioning about suicidal thoughts and plans, which patients may be reluctant to disclose or may not fully recognize themselves.
Nightmares offer a potential non-invasive, low-stigma screening entry point. Unlike directly asking about suicidal intent, asking about dream content may feel less threatening and may capture individuals who are in a prodromal phase of crisis but have not yet formed explicit suicidal plans. The fact that nightmares preceded the acute crisis by an average of four months creates a theoretical window for early intervention.
The evolutionary pattern identified in this study is especially useful clinically. A patient whose dreams have been getting progressively worse, from mildly unpleasant to terrifying and eventually to sleep-interrupting nightmares, may be signaling escalating distress long before a suicide attempt. Suicidal content in dreams appears to be a particularly acute marker, appearing just over two weeks before the emergency visit.
These findings align with earlier work by the same research group. A 2022 study published in the Journal of Clinical Psychiatry, also led by researchers at GHU Paris, found that 80% of patients hospitalized after a suicidal crisis reported altered dreams, with bad dreams appearing about 111 days before the crisis, nightmares at 87 days, and suicidal scenarios at 45 days. The new study extends these observations to the emergency department setting, where rapid triage decisions must be made.
The authors recommend that routine psychiatric assessment should include systematic screening for nightmares and dream alterations, a practice that is not currently standard in most emergency psychiatric services. Simple questions about dream frequency, content, and whether dream patterns have changed recently could be incorporated into existing suicide risk assessment protocols at minimal cost.
Limits
The study has important limitations. The sample is small, just 42 patients, all from a single psychiatric emergency center in Paris, which limits generalizability. The data are self-reported and retrospective, relying on patients’ recall of dream content over the preceding months, which may be subject to memory bias. The study also lacks a control group of patients without suicidal crisis, making it impossible to determine how specific these dream alterations are to suicide risk versus general psychological distress.
Additionally, the study period (March-April 2025) was brief, and the researchers did not follow patients prospectively to see whether dream alterations predicted future suicidal behavior. Without longitudinal follow-up, it remains unclear whether nightmare screening in emergency settings would actually lead to improved outcomes.
Cultural factors may also play a role. Dream recall and reporting are influenced by cultural attitudes toward dreams, and findings from a single French urban center may not translate directly to other populations or healthcare settings.
Despite these caveats, the consistency of findings across this study and prior work by the same group strengthens the case that dream alterations deserve serious attention as a clinical signal.
Bottom line
This study adds to mounting evidence that nightmares are not merely a distressing symptom of mental illness but may serve as a temporally informative marker of escalating suicide risk. For clinicians, the message is straightforward: ask about nightmares.
A brief screening question about whether dream content has changed recently, whether nightmares have become more frequent, and especially whether suicidal themes have appeared, could help identify patients approaching a crisis. When combined with assessments of insomnia and anxiety, both correlated with nightmare severity in this study, dream content may help emergency psychiatrists make more informed triage decisions.
For the general public, the takeaway is equally important. Persistent nightmares, particularly worsening dream content over time, warrant attention. They are not just bad dreams. They may be the mind’s way of sounding an alarm.
Source
Gagneron E, Rezaei K, Perot R, Geoffroy PA, Ambar Akkaoui M. Dream and nightmare alterations in patients presenting with suicidal crisis: A prospective study in psychiatric emergency department. Sleep Medicine. 2026;146:109110. PMID: 42400990. DOI: 10.1016/j.sleep.2026.109110.