Understanding Suicidality in Youths: Insights from a Consultation-Liaison Psychiatry Service

Objectives. This study examines the clinical and psychosocial correlates of suicidality in adolescents and young adults, focusing on youths who received psychiatric consultation after a suicide attempt.

Methods. A retrospective study was conducted at the University Hospital of Perugia, involving 128 patients aged 14–35 who underwent psychiatric assessment during medical hospitalization. Data was collected for the first consultation performed after the hospital admission. Participants were divided in those referred after a suicide attempt (SA group) and those referred for other psychiatric reasons (non-SA group). Bivariate analyses were performed as appropriate to assess significant differences between the groups (p < 0.05).

Results. Most patients were females and aged over 18 years old. Youths in the SA group (n = 72; 56.3%) were significantly younger than those in the non-SA group and were less frequently employed. They also presented a higher rate of previous psychiatric hospitalizations, more frequent positive family psychiatric history, and a higher prevalence of prior suicide attempts. SA patients were more often under the care of community mental health services, while anxiety symptoms were more common in the non-SA group. Insomnia before admission was reported more frequently among SA patients.

Conclusions. These findings confirm the high prevalence of suicidality in youths. Comprehensive evaluation and integrated intervention strategies are essential to improve identification and prevention in this population.

INTRODUCTION

The transition from adolescence to early adulthood represents a period of increased vulnerability, characterised by rapid neurobiological changes, ongoing identity consolidation and considerable instability in education, work, residence, and personal relationships ¹. Within this developmental window, the emergence of psychiatric disorders or manifestations is frequent, including suicidal ideation and self-injurious behaviours, which may serve as early indicators of psychological distress ^2,3. Suicide represents a leading cause of mortality among young people, ranking fourth among individuals aged 15 to 29, with significant social and economic consequences ⁴. Furthermore, non-fatal suicide attempts are estimated to occur at least twenty times more frequently than completed suicides, underscoring the considerable burden of hidden psychological suffering in this age group ⁴.

Suicidal behaviours during adolescence and early adulthood frequently emerge against a backdrop of individual vulnerability, encompassing mood and anxiety disorders, psychosocial stressors, traumatic experiences, and relational or familial instability. Clinical studies focusing on adolescents and young adults presenting to emergency departments following suicide attempts consistently demonstrate high rates of psychiatric comorbidity, including depression, substance abuse, and personality disorders, alongside unstable social contexts, all of which may increase the likelihood of recurrence ^5,6. Notably, the majority of psychiatric disorders have their onset during this developmental period, with over 60% of lifetime diagnoses occurring before the age of 25, with a tenfold increase in suicide risk ^3,7.

In this context, the systematic investigation of data concerning psychiatric consultations provided to young individuals following a suicide attempt offers a unique opportunity to delineate potential post-attempt phenotypes, encompassing distinctive clinical and psychosocial profiles associated with severe self-injurious behaviour. Characterising these phenotypes facilitates the identification of subtle markers of vulnerability, which may enable earlier recognition of at-risk individuals and the shaping of tailored preventive strategies. Notably, a prior suicide attempt has been consistently identified as one of the strongest predictors of both future attempts and suicide completion ^8,9.

Clinical assessment conducted during medical hospitalisation constitutes a particularly valuable approach, allowing for comprehensive evaluation of psychiatric symptoms, psychological functioning, and family and social dynamics. Notably, for a substantial proportion of young individuals, a suicide attempt represents the first access to psychiatric care, making the hospital setting a crucial opportunity for timely recognition and treatment ¹⁰.

While consultation-liaison psychiatry primarily serves to integrate psychiatric care within medical contexts, it assumes a pivotal functional role in this framework, providing a structured opportunity to observe and document post-attempt phenotypes. Evidence from Italian and international settings indicates that approximately two-thirds of patients evaluated in medical wards present with at least one psychiatric disorder, most frequently depression, adjustment disorders, delirium, substance abuse, and acute stress reactions ^11,12,13. Integrating these clinical observations with detailed information on social and familial context permits the identification of vulnerabilities that may predispose to recurrent self-harm, thereby informing preventive interventions and structured follow-up.

Investigating post-attempt phenotypes offers concrete tools for early detection, reduction of mortality, and prevention of recurrent self-injurious behaviour, ultimately supporting the implementation of multidimensional and personalised interventions tailored to the unique vulnerabilities of this population ^4,5. Based on what stated above, this study aims to assess the prevalence and correlates of youth suicidality in a liaison-consultation setting. We hypothesized that suicide attempters in this population would present specific clinical, psychological, and socio-environmental characteristics, to be evaluated in a multidimensional perspective.

MATERIALS AND METHODS

This study employed a retrospective design and was conducted at the University Hospital of Perugia (Italy), within the framework of a consultation-liaison psychiatry service. Data were obtained from structured reports generated during psychiatric consultations performed in medical and surgical wards over the period July 2023-June 2025.

Eligible participants were patients aged 14 to 35 years who were admitted to non-psychiatric wards and underwent their first psychiatric assessment during the index hospitalisation. The age range was selected to capture a clinically relevant developmental continuum encompassing adolescence and young adulthood. This choice is consistent with international frameworks describing emerging adulthood as a prolonged transitional phase characterized by increased vulnerability to psychopathology and suicidal behavior ^1,3. Moreover, our choice reflects the real-world organization of the service. Indeed, in consultation-liaison psychiatry settings adolescents and young adults are often evaluated within the same clinical pathways; subsequently, our analysis supports a pragmatic and clinically grounded grouping. Based on the primary hospitalization reason, the cohort of 128 patients was divided into two groups: (i) the SA group, consisting of individuals referred following a suicide attempt (n = 72); and (ii) the non-SA group, comprising patients referred for other psychiatric concerns, e.g., affective symptoms or behavioural disturbances (n = 56). All the included patients signed their informed consent for data treatment at hospital referral, according to the local privacy policies and procedures.

To characterise clinical and psychosocial profiles, information was systematically extracted from consultation records using a case report form created ad-hoc for the study. Collected variables included sociodemographic data (age, biological sex, employment status), personal and family psychiatric history, current suicidal ideation, mood and anxiety symptoms (both somatic and psychological), sleep disturbances, major stressors occurring in the preceding six months, presence of comorbid personality disorders, chronic medical conditions, ongoing psychotropic treatments, and engagement with community mental health services (CMHS). This approach allowed detailed profiling of potential post-suicide attempt phenotypes and identification of subtle vulnerability markers. All the psychiatric diagnoses were based on the Diagnostic and Statistical Manual of Mental Disorders, 5th edition, text revision (DSM-5-TR) criteria ¹⁴. The assessment of suicidality was routinely performed by psychiatrists working in the service who were trained for the conduction of clinical studies. As for routine clinical practice, clinicians used the Italian version of the Columbia Suicide Severity Rating Scale (C-SSRS) ¹⁵, an instrument that evaluates suicidal ideation and suicidal behaviors through structured questions. According to the C-SSRS, SA were defined as self-injurious acts committed with at least some intent to die. The evaluation of affective symptoms was also performed by using validated instruments, namely the Hamilton Depression Rating Scale (HAM-D) ¹⁰, the Hamilton Anxiety Rating Scale (HAM-A) ¹⁶, and the Mania Rating Scale (MRS) ¹⁷. The presence of clinically significant affective symptoms was established using the following cut-offs: HAM-D ≥ 14 for depressive symptoms, HAM-A ≥ 18 for anxiety symptoms, and MRS ≥ 20 for manic symptoms ¹⁸.

Statistical analyses were conducted using IBM SPSS Statistics, version 26. Descriptive statistics were performed to examine the distributional properties of the variables of interest in the sample. Continuous variables were considered normally distributed according to the central limit theorem. Group comparisons between SA and non-SA patients were carried out using bivariate analyses: categorical variables were examined via Chi-square or Fisher’s exact test, while continuous variables were assessed using Student’s t-test. A significance threshold of p < 0.05 was applied throughout.

RESULTS

Sample description

Of the 128 patients included in the analysis, most were females (n = 89, 69.5%), with a mean age of 23.91 ± 6.88 and a higher prevalence of patients aged ≥ 18 (n = 97, 75.8%). The most represented reason for requesting the psychiatric consultation (n = 42, 32.8%) was a general psychopathological assessment, followed by the need of re-evaluating psychopharmacological treatment (n = 29, 22.7%). Suicide attempts in the study population displayed a prevalence of 56.3% (n = 72). The most frequent method among suicide attempters was medication overuse (n = 48, 64.9%), followed by ingestion of caustic substances (n = 11, 14.9%).

Correlates of suicidality in the sample

No sex differences were detected between SA and non-SA subgroups. When evaluating other sociodemographic characteristics, we found that 84.7% patients in the SA group were unemployed compared to 63.8% in the non-SA group (p = 0.024). Moreover, a higher proportion of participants in the non-SA group were parents (17.9%) compared to those in the SA group (2.9%) (p = 0.011). A positive psychiatric family history was more common in the SA group (47.5% vs. 21.7%; p = 0.022), and so were previous suicide attempts (36.1% vs 10.9%, p = 0.002). When analysing psychopathological correlates, a higher prevalence of insomnia prior to admission (47.1% vs. 11.1%; p = 0.020) was observed in the SA group whereas anxiety symptoms, particularly with both psychic and somatic features (44.4% vs. 8.7%; p = 0.023), were mainly observed in the non-SA group. In the SA group we also found a higher prevalence of personality disorders (55.7% vs. 29.3%; p = 0.015). As for treatment features, the use of mood stabilizers was more frequent in the SA group (54% vs. 24.1%; p = 0.019).

Only 54.2% of SA patients were being treated in CMH services, compared to the 32.1% of non-SA group (p = 0.021). Previous hospitalizations in a psychiatric setting were more frequent (34.4% vs 9.1%, p = 0.002). Life stressors in the preceding six months were not more prevalent in the SA group.

Patients in the SA group were more frequently assessed as having suicidal ideation (32.3% vs. 5.5%; p = 0.001). Remarkably, 39.3% of SA patients were assessed as having a euthymic or non-depressed mood state, defined as the absence of clinically significant depressive and manic symptoms as assessed by HAM-D and MRS (see Materials and Methods section). Recommendations for initiating or adjusting pharmacological therapy were more frequently observed in the SA group (63.8% vs 24.1%, p = 0.002). Patients with a chronic medical illness were less represented among those who had attempted suicide (p = 0.014). A graphical representation of the main findings of the bivariate analyses can be found in Figure 1.

DISCUSSION

This study reinforces the high burden of suicidality among adolescents and young adults in psychiatric consultation settings, with more than half of the sample referred to our service after a suicide attempt. These findings align with epidemiological research, showing that youth may represent a peak-risk developmental phase for self-injurious behaviours and suicide ^1,4,5,7. As previously hypothesized, the considerable rate of unemployment and the elevated prevalence of previous suicide attempts, psychiatric family history, and comorbid personality disorders observed among attempters is consistent with existing models describing multidimensional vulnerability contributing to suicidality ^5,6,7.

Our results support the existence of “post-attempt phenotypes”, an emerging framework aimed at identifying discrete clinical and psychosocial profiles with prognostic and preventive relevance ^8,9. In this context, the observation that about 30% of attempters were euthymic at evaluation underscores the limits of symptom-based screening and reinforces the need to incorporate behavioural markers, including circadian rhythm disturbances, and psychosocial instability into risk formulation. Insomnia, markedly more frequent among attempters, is increasingly recognised as an independent predictor of suicidal behaviour and should be routinely assessed in emergency and liaison contexts.

Notably, only around half of suicide attempters were engaged with community mental health services at the time of consultation. This confirms that the medical hospital setting frequently represents a first point of psychiatric contact, as reported in previous literature ^10–13. In this respect, consultation-liaison psychiatry is not merely a diagnostic interface but a strategic platform for early intervention. Recent work has emphasised that liaison services can actively shift from reactive to proactive models of care delivery, integrating systematic suicide risk assessment, safety planning, and facilitated referral into standard medical pathways ^20,21. Such innovations are essential to close the care gap between emergency presentation and sustained psychiatric follow-up.

Previous evidence consistently demonstrates that early psychiatric intervention following self-harm or emerging psychopathology can significantly reduce recurrence and mortality. Meta-analytic findings indicate that early intervention programmes are associated with lower rates of future suicidal behaviour, supporting the broader principle that early, assertive engagement matters regardless of diagnosis ²². Similarly, longitudinal data demonstrate that a first suicide attempt carries substantial long-term risk in terms of recurrence rates, persistent lethality, and non-negligible mortality, confirming the attempt as a sentinel event requiring intensive secondary prevention ²³. This evidence strengthens the rationale for early post-attempt care within medical settings, where patients are more easily accessible ²⁴. Consultation-liaison psychiatry is therefore ideally positioned to serve as an early detection and intervention node in the clinical pathway of suicidal youths. In line with this perspective, future models of liaison care should include systematic follow-up protocols, integrated crisis planning, and active coordination with community services, as these elements are associated with improved continuity of care and reduction of repeat self-harm ^6,8–10.

Some limitations must be acknowledged for the present study. The retrospective design and reliance on consultation reports may lead to collection bias, and the absence of longitudinal data precludes direct assessment of outcomes. The analyses are based on bivariate comparisons, which do not allow a deeper understanding of statistical associations and prevent to draw causality based on the observed findings. Nonetheless, the study benefits from a real-world clinical sample, which is also reflected by the choice of the statistical analyses, and provides novel insights into suicidality in a liaison-consultation setting.

Further research should adopt longitudinal and interventional designs to assess whether structured early intervention delivered at the point of medical contact reduces repetition and improves clinical trajectories. Integrating biomarkers, digital phenotyping, and psychosocial markers into assessment may refine suicide risk stratification in this population. Expanding consultation-liaison services to include proactive, standardised protocols – as recently proposed in paediatric and adolescent psychiatry – may represent a decisive step towards effective prevention.

CONCLUSIONS

Psychiatric consultation-liaison settings within the general hospital emerge as a critical point of engagement for youths showing high suicide risk, particularly in case of limited prior mental health contact. This window offers a unique opportunity for proactive assessment, initiation of treatment, and linkage to community services. Given the high prevalence of these features in youth populations, the need for a multidimensional risk assessment, as well as the long-term mortality and recurrence associated with a first suicide attempt, early structured intervention following hospitalisation should be considered an essential component of secondary prevention.

Ethical Considerations

Due to the retrospective nature of the study, ethical approval was not required for the present research. All the included patients signed informed consent for data treatment at hospital admission according to local privacy policies and procedures.

Funding

None.

Conflicts of interest

Authors declare no conflicts of interest concerning the present research.

Author contributions

CMC: writing (first draft), data collection and curation, statistical analysis; GC: writing (review & editing), data collection and curation, statistical analysis; AS, FDM, & ESG: data collection and curation; FDG & AT: writing (review & editing); GM: study conceptualization, supervision.

Figures and tables

FIGURE 1. Significant sociodemographic, clinical, and treatment-related differences between youths referred after a suicide attempt (SA group) and those referred for other psychiatric reasons (non-SA group). Only variables showing statistically significant differences in bivariate analyses (p < 0.05) are displayed.