Conference Agenda

Background: Large-scale neuroimaging research has advanced our understanding of suicidality. A landmark ENIGMA study of nearly 19,000 participants (Campos et al., 2021) identified a specific subcortical signature of suicide attempt: reduced volume in the left thalamus (Thal-L), right thalamus (Thal-R), and right pallidum (Pall-R), with no comparable associations for suicidal symptoms or ideation (Rentería et al., 2017; van Velzen et al., 2022). Thus, while this thalamo-pallidal pattern is established for suicide attempts, the neural basis of suicidal ideation remains unclear. The thalamus and pallidum, key subcortical hubs, primarily support emotional reactivity, motor gating, and inhibitory control rather than deliberation or planning. This aligns with frameworks distinguishing “hot” (affectively driven, unplanned) from “cold” (intentional, deliberative) suicidal ideation (Allen et al., 2021; Reis et al., 2023; Bryant et al., 2023; Dobbertin et al., 2023). We therefore hypothesised that ENIGMA-defined subcortical alterations would be associated specifically with hot (non-planned) ideation.

Methods: We tested this hypothesis in 220 individuals hospitalised for a major depressive episode who underwent 3T MRI and completed the Beck Depression Inventory-II. The suicide item (item 9) was used to classify ideation subtype. A latent thalamo-pallidal index combining the three ENIGMA-defined volumes (Thal-L, Thal-R, Pall-R) was derived using z-scores, PCA, and factor analysis.

Results: ANCOVAs adjusting for age, sex, depression severity, history of suicide attempt, and intracranial volume revealed a clear dissociation: non-planned (“hot”) ideation was associated with reduced thalamo-pallidal volume compared with planned (“cold”) ideation.

Conclusion: These findings parallel ENIGMA’s suicide-attempt signature and suggest that differentiating hot from cold suicidal ideation may be critical for identifying meaningful neurobiological markers. However, the study relies on a small sample, which limits the generalisability of the effects. Replication in larger, multi-site datasets will be necessary to determine whether the thalamo-pallidal pattern represents a reliable neural correlate of hot suicidal ideation.

Background: Little attention is paid to physiological markers of restricted self-regulatory capacity in suicide risk assessment and clinical care. This study aimed to examine heart rate (HR, as arousal marker) and heart rate variability (HRV, as physiological marker of restricted self-regulatory capacity) as potential indicators of risk for suicidal thoughts and behaviour (STB) in a clinical population. We expected HR to be positively correlated and HRV to be negatively correlated with the intensity and frequency of suicidal thoughts and behaviour.

Methods: Data of N = 331 psychiatric inpatients admitted after suicide attempt or acute suicidal crises were collected from March 2022 until October 2025 in nine psychiatric hospitals in Germany as part of a larger prospective multicenter-trial. Preliminary data of N = 150 patients, who participated from March 2022 until February 2024, were already analyzed using t-tests and correlation analyses.

Results: Results showed a positive significant correlation between HR and intensity of suicidal ideation, while no correlation between HR or HRV and persistence of suicidal ideation was found. HR was significantly higher, while high frequency and low frequency HRV indices were significantly lower in suicide attempters than non-attempters. Primarily reported reasons for suicide attempt were, for example, stop intense feelings or to stop a dissociative state. Those could be interpreted as an expression of a failure of self-regulatory capacities. We will be able to present further analyses of all 331 participants of the study at the conference.

Conclusion: Results suggest that HR and HRV are related to STB in suicidal inpatients. As the present analyses were cross-sectional, future prospective studies are warranted. They could aim at clarifying whether unusual HR/HRV values should rather be deemed as risk factors for STB, as moderators of the relation between other risk factors and heightened STBs or as mere correlates of STBs.

Background: Implicit cognitive processes may contribute to suicidal behaviour in ways not fully captured by self-report. This study examined two versions of the Death-Implicit Association Test (D-IAT): the standard identity-based D-IAT (iD-IAT), assessing self-death associations, and a novel control-adapted D-IAT (cD-IAT), targeting internal control associations with death. The study investigated whether these implicit markers differentiate suicidal behaviour based on attempt recency and frequency.

Methods: N = 116 psychiatric inpatients (58.1% female; M age = 33.6, SD = 12.4) with a history of suicidal behaviour were classified as single versus multiple attempters and recent versus lifetime attempters. Participants completed the iD-IAT and cD-IAT alongside self-report measures of suicidal ideation, depression, and locus of control. Group comparisons, correlations, logistic regression, internal consistency analyses, and ROC curves were conducted.

Results: Recent attempters showed weaker self-life (t(114) = 2.18, p = .016) and weaker internal control-life (t(114) = 2.26, p = .013) associations than lifetime attempters. Multiple attempters showed weaker internal control-life associations than single attempters (t(114) = 2.25, p = .007). The iD-IAT correlated with suicidal ideation (r_s = .20, p = .032), depression (r_s = .20, p = .033), and external control (r_s = .21, p = .021). The cD-IAT correlated with suicidal ideation (r_s = .25, p = .006) and depression (r_s = .26, p = .006). The cD-IAT predicted multiple attempts (χ²(1) = 3.88, p = .049), showed higher internal consistency (r_sb = .31, p = .001), and achieved better discrimination of multiple (AUC = .64, p = .013) and recent (AUC = .62, p = .028) attempters than the iD-IAT.

Conclusion: The control-adapted D-IAT shows preliminary promise as an implicit marker distinguishing subgroups of suicidal behaviour based on recency and frequency. Findings support the relevance of implicit control-related processes in suicidal vulnerability, warranting further refinement and prospective validation.

Background: Psychological pain is considered a central construct in contemporary models of suicidal thoughts and behaviours (STB). This study examines group differences and psychometric properties of the Mee-Bunney Psychological Pain Assessment Scale (MBPPAS) in individuals with varying levels of suicidal ideation and behavior.

Methods: A cross-sectional design as employed to collect two independent samples at multiple centres (Sample 1: n = 409; Sample 2: n = 171), resulting in a pooled dataset of N = 580 participants (59% female; M_age = 34.6, SD = 13.2). Participants were assigned to four diagnostically defined groups: suicide attempters (SUAT), individuals with suicidal ideation without attempt (SUID), clinical controls without STB (CLIN), and healthy controls (HLTH). Psychological pain was assessed using the MBPPAS and compared across groups. Internal consistency, convergent, incremental, and criterion validity were evaluated.

Results: SUAT reported significantly higher psychological pain than SUID (p < .001, d = 0.42), CLIN (p < .001, d = 1.46), and HLTH (p < .001, d = 1.04). SUID also showed markedly higher scores than CLIN (p < .001, d = 3.01) and HLTH (p < .001, d = 2.57). The MBPPAS demonstrated strong internal reliability (α = .93, ω = .93) and strong convergent, incremental, and criterion validity.

Conclusion: Psychological pain reliably differentiates between groups with suicidal behaviour, suicidal ideation, clinical controls, and healthy controls. The findings support psychological pain as a quantifiable marker of suicide risk. The MBPPAS provides a robust methodological tool for assessing psychological pain in clinical research.