The research, led by Mengtian Li, Feng Si, Yani Zhang, Qingyuan Yu, Zhengjun Wang, Hui Meng, Haiyan Zhao, and Jianqin Cao at Harbin Medical University and collaborating institutions in China, was designed to answer a deceptively simple question: does the heart’s rhythm look different in people who are depressed compared with people who are depressed and also experiencing suicidal ideation? The distinction matters clinically. Suicidal ideation is one of the most urgent warning signs in mental health care, yet identifying who is at risk remains notoriously difficult, relying largely on self-report questionnaires and clinical interviews. If a physiological marker could help flag elevated risk, it could eventually complement existing screening tools, particularly in settings where structured psychiatric assessment is impractical.

To explore this, the team ran two parallel studies with different populations. The first involved 86 university students, a group often studied in HRV research because depressive symptoms are common in young adulthood and recruitment is relatively straightforward. The second involved 150 clinical participants, recruited through a hospital setting, whose depression had been formally diagnosed. In both studies, participants were divided into three groups: healthy controls, individuals with depressive symptoms or diagnosed depression without significant suicidal thoughts, and individuals with depression accompanied by suicidal ideation. The researchers used study-specific operational criteria for these classifications, and in the university sample the group without suicidal ideation was defined by scores below a prespecified threshold on a standard suicide assessment rather than by a complete absence of such thoughts, a nuance that reflects the messy reality of measuring suicidal thinking in nonclinical populations.

The physiological measurement itself was refreshingly simple. Each participant underwent a five-minute electrocardiogram recording at rest, the standard window used in HRV research. From those recordings the team extracted two of the most widely used time-domain indices. The first, SDNN, the standard deviation of normal-to-normal heartbeat intervals, captures overall variability across the recording period and reflects both sympathetic and parasympathetic influences on the heart. The second, RMSSD, the root mean square of successive differences between adjacent heartbeats, is more specifically sensitive to parasympathetic activity, the vagus nerve’s calming brake on the cardiovascular system. Both measures were natural-log transformed before analysis, a common step that helps stabilize their statistical distribution. Lower values on either index indicate a heart rhythm that is more rigid and less adaptable, which in the HRV literature is often interpreted as a sign of autonomic dysregulation.

The statistical strategy was layered to guard against overinterpretation. Age- and gender-adjusted general linear models were used to test three kinds of effects: an omnibus test of whether the three groups differed at all, a test of a linear trend across the groups, and planned contrasts between adjacent groups. In the university sample, the omnibus effect for lnSDNN was significant, and both lnSDNN and lnRMSSD showed significant linear trends, meaning that HRV values declined in step with the severity of the depressive presentation, from healthy controls to depressed students to depressed students with suicidal ideation. However, the planned contrasts between adjacent groups did not survive statistical correction, suggesting that in this relatively small sample the individual steps along that gradient could not be confidently separated from one another.

The clinical sample told a sharper story. Among the 150 patients, both primary HRV indices showed significant omnibus group effects, significant linear trends, and, crucially, significant planned contrasts between adjacent groups. In other words, the differences were robust enough not just to detect a general gradient but to pinpoint where the drops occurred. Patients with depression and suicidal ideation had measurably lower heart rate variability than patients with depression alone, who in turn had lower variability than healthy controls. This pattern held for both the broad SDNN measure and the more parasympathetically sensitive RMSSD measure, pointing to a disruption in the vagal regulation of the heart that deepens as suicidal thinking enters the clinical picture.

One of the study’s most important analyses addressed an obvious confound. Depression severity itself is associated with lower HRV, and people with suicidal ideation tend to be more severely depressed, so the apparent link between suicidal thoughts and HRV could simply be a proxy for symptom burden. To test this, the researchers repeated their analyses while statistically adjusting for depressive symptom severity, and they excluded the suicide item from the Hamilton Depression Rating Scale when computing that severity score, ensuring the adjustment was not contaminated by the very construct under investigation. Even after this adjustment, the group with suicidal ideation remained associated with lower lnSDNN in both samples, and with lower lnRMSSD in the clinical sample. That suggests the association between suicidal ideation and autonomic dysregulation is not entirely reducible to how depressed a person is, a finding with real theoretical weight for models that treat suicidality as a distinct physiological state rather than merely an extreme of depression.

The researchers also conducted exploratory cross-study analyses, pooling the two samples after standardizing the HRV measures within each study. These pooled models showed significant group-related patterns and, notably, significant group-by-sample interactions, indicating that the relationship between diagnostic group and heart rhythm differed between the university and clinical contexts. The authors were careful to interpret these cross-study findings cautiously, and for good reason: the two samples differed substantially in how they were recruited, how thoroughly they were clinically characterized, the range of symptoms they displayed, and the assessment context itself. A student sitting in a university laboratory and a patient in a hospital psychiatric setting bring very different baseline states to a five-minute ECG recording, and the interaction effects likely reflect those contextual differences as much as any true population difference.

What emerges from the study is a coherent but nuanced picture. Lower heart rate variability was associated with depressive symptom burden and with suicidal ideation in both the university and clinical samples, supporting the broader idea that autonomic flexibility is a transdiagnostic marker of psychological distress. But the findings were statistically more robust in the clinical study, where the diagnostic boundaries were clearer and the symptom range wider, than in the student sample, where adjacent group contrasts fell short after correction. The authors themselves emphasize that differences between the two studies should be interpreted cautiously, a refreshing dose of restraint in a field where physiological markers of mental states are often oversold. The work was supported by the Fundamental Research Funds for the Provincial Universities, the Natural Science Foundation of Heilongjiang Province, and a Humanities and Social Sciences Project from China’s Ministry of Education, and it was approved by the Ethics Committee of the Daqing Campus of Harbin Medical University with all participants providing informed consent.

The implications reach in two directions. Scientifically, the study strengthens the case that suicidal ideation has a detectable autonomic correlate, consistent with theories linking reduced vagal tone to impaired emotion regulation, heightened stress reactivity, and difficulty disengaging from negative thought loops. Practically, the road from a five-minute ECG to a suicide risk screening tool is long, and this study does not claim to have traveled it. HRV is influenced by age, fitness, medication, caffeine, and countless other factors, and the cross-study interactions here show how fragile such measures can be across contexts. But as wearable devices make continuous heart rhythm monitoring ubiquitous, research like this helps map which features of that rhythm carry genuine psychological signal. A heart that beats with less flexibility may one day be one piece of a larger puzzle in identifying who needs help most urgently, provided the science continues to be pursued with the same caution the authors bring to their own results.

Subject of Research: The relationship between resting-state heart rate variability, depression, and suicidal ideation in university and clinical samples

Article Title: Heart rate variability across depressive presentations with and without suicidal ideation in university and clinical samples

Article References: Li, M., Si, F., Zhang, Y., Yu, Q., Wang, Z., Meng, H., Zhao, H., & Cao, J. (2026). Heart rate variability across depressive presentations with and without suicidal ideation in university and clinical samples. BMC Psychiatry. https://doi.org/10.1186/s12888-026-08773-9

Image Credits: AI Generated

DOI: 10.1186/s12888-026-08773-9

Keywords: heart rate variability, depression, suicidal ideation, autonomic nervous system, SDNN, RMSSD, vagal tone, psychiatry, ECG, mental health screening, clinical sample, university students