Factors associated with lifetime suicidal ideation at baseline

Students identifying as a sexual minority or gender-diverse, male students, and those of Chinese ethnicity all showed elevated odds of lifetime suicidal ideation. These findings are consistent with research documenting higher suicide risk among minoritised populations, ^{Reference Zhao, Xin, Bai, Zhang, Liu and Xu6} with minority stress theory suggesting that experiences of marginalisation may contribute to poorer mental health outcomes in these groups. ^{Reference Frost and Meyer35} International students demonstrated reduced odds of risk compared with domestic students, although this warrants cautious interpretation given potential cultural differences in disclosure, ^{Reference Ahad, Sanchez-Gonzalez and Junquera36} as well as the possibility that students with a prior history of mental health difficulties may be less likely to pursue study abroad. Elevated risk among students whose parents held professional degrees may reflect greater perceived pressure to achieve, willingness to disclose or selective participation, warranting future investigation. Lower risk at site 5 may reflect recruitment of all first-year students, reducing self-selection toward elevated risk or interest in mental health research.

Childhood adversity exhibited associations with baseline suicidal ideation, with experiences of sexual abuse demonstrating the strongest association, consistent with dose–response models linking adverse experiences to suicidality. ^{Reference Tan and Mao37} Within the integrated motivational–volitional framework, early adversities may be conceptualised as diathesis factors increasing vulnerability to suicidal behaviour through effects on mood, interpersonal relationships and stress regulation. ^{Reference O’Connor and Kirtley12} History of mental health treatment was associated with suicidal ideation, likely reflecting treatment history as a marker of severity and persistence of symptoms and a potential protective factor through access to support. ^{Reference Franklin, Ribeiro, Fox, Bentley, Kleiman and Huang16}

Baseline psychosocial functioning was strongly associated with suicidal ideation. Depressive symptoms showed the most robust association, consistent with models suggesting depressive symptoms often contribute to suicidal thinking. ^{Reference Franklin, Ribeiro, Fox, Bentley, Kleiman and Huang16} Higher loneliness and alcohol use were also associated with increased odds, consistent with previous literature. ^{Reference Campbell, Blank, Cantrell, Baxter, Blackmore and Dixon38} However, the positive association with wellbeing was unexpected, potentially reflecting statistical suppression, given the strong inverse correlation between well-being and depression, or collider bias from selective participation: students with a history of mental health problems may have been more motivated to participate, while having sufficient current well-being to complete the survey. ^{Reference Tönnies, Kahl and Kuss39} Conversely, higher resilience and higher anxiety symptoms were independently associated with reduced risk. The latter contrasts with research reporting positive associations between anxiety and suicide risk, ^{Reference Kabir, Syed, Vinnakota, Okello, Isigi and Abdul Kareem9} and may reflect a distinct relationship in this population, artefact of recruitment or sampling, or the predominantly subclinical anxiety severity in this sample, warranting future study.

Factors associated with lifetime suicide attempts at baseline

Previously receiving mental health treatment was most strongly associated with past suicide attempts at baseline, with over three times greater risk. Students identifying as non-Chinese Asian ethnic background showed elevated odds relative to White students, and those identifying as homosexual/gay/lesbian demonstrated increased risk compared with heterosexual students. As with ideation, international students showed reduced odds of reporting a past attempt, again warranting cautious interpretation. ^{Reference Ahad, Sanchez-Gonzalez and Junquera36}

Childhood adversity was also associated with past suicide attempts. Being hit by a caregiver demonstrated the strongest association, followed by sexual abuse, bullying and harsh parenting. The pattern mirrors that observed for suicidal ideation, consistent with dose–response models linking adverse childhood experiences to later suicidal behaviour. ^{Reference O’Connor and Kirtley12}

Baseline psychosocial functioning showed several associations with past suicide attempts. Depressive symptoms and rumination were both positively associated with lifetime attempt history, consistent with research highlighting repetitive negative thinking as a cognitive mechanism that may contribute to suicidal risk beyond ideation. ^{Reference O’Connor and Kirtley12,Reference Watkins40} Higher alcohol use was also positively associated, whereas better sleep quality and greater self-care were protective, consistent with previous literature. ^{Reference Zhao, Xin, Bai, Zhang, Liu and Xu6,Reference Kabir, Syed, Vinnakota, Okello, Isigi and Abdul Kareem9} Similar to ideation, higher well-being was also associated with past attempts, which may also reflect collider bias from selective participation. ^{Reference Tönnies, Kahl and Kuss39} Likewise, higher resilience was also positively associated with past attempts, potentially reflecting adaptation or response bias among those with attempt histories. Students with no previous mental health diagnosis were less likely to report a past attempt, reinforcing the strong link between clinical history and suicidal behaviour. ^{Reference Zhao, Xin, Bai, Zhang, Liu and Xu6}

Reduction of suicidal ideation

Among students beginning the academic year with past suicidal ideation, 61% presented no symptoms at follow-up. Older students were slightly more likely to experience symptom reduction, potentially reflecting more developed coping strategies. Students reporting physical disabilities also showed higher odds of improvement, although this should be interpreted cautiously, given the heterogeneity of physical disabilities and varied support needs.

Conversely, students reporting prior mental health treatment were less likely to report no symptoms at follow-up, potentially reflecting treatment history as an indicator of severity. ^{Reference Franklin, Ribeiro, Fox, Bentley, Kleiman and Huang16} Similarly, increases in rumination over the academic year were associated with lower risk of reduction, consistent with research linking repetitive negative thought to maintenance of depressive symptoms. ^{Reference Watkins40} Notably, concurrent change in rumination and suicidal ideation does not suggest directionality, but highlights an association between the two across the academic year.

New onset of suicidal ideation

New onset of suicidal ideation occurred in 3.5% of students without baseline suicidal ideation. When matched with students on baseline risk and demographic characteristics, the transition to new ideation was associated with declines in mental health, particularly anxiety, depression and well-being, similar to associations with baseline symptoms. This aligns with ideation-to-action frameworks suggesting depression primarily functions as a risk factor for ideation rather than attempts. ^{Reference Klonsky, Dixon-Luinenburg and May18}

The rarity of new-onset suicidal ideation has potentially meaningful implications for university mental health screening practises. Findings suggest the majority of students who experience suicidal ideation during university already demonstrate lifetime risk factors or symptoms, indicating that early identification and intervention during the transition to university may be particularly important. ^{Reference Franklin, Ribeiro, Fox, Bentley, Kleiman and Huang16} Additionally, findings suggest that among students without baseline risk, monitoring for symptoms of mental health deterioration may be useful in identifying students most at risk, consistent with research demonstrating that current psychiatric symptoms are better predictors of suicidal ideation than demographic risk factors. ^{Reference Macalli, Navarro, Orri, Tournier, Thiébaut and Côté17}

Relationship between suicidal ideation and attempts

Approximately 50% of students reported no suicidal ideation or attempts at either time point, with over 70% presenting without risk indicators at follow-up. Additionally, no risk at follow-up was the most common outcome for students with a history of risk, suggesting the university environment likely did not exacerbate existing vulnerability. Among students with suicidal thoughts but no prior attempts, 3.2% made a new attempt during the academic year, whereas 10.5% of those with a prior attempt made a new attempt, consistent with research demonstrating that the majority of individuals with suicidal ideation do not progress to suicide attempts. ^{Reference Klonsky, Dixon-Luinenburg and May18} The ideation-to-action framework posits that development of suicidal ideation and progression to attempts are distinct processes with different predictors, with capability and thoughts operating as distinct mechanisms. ^{Reference O’Connor and Kirtley12}

These findings may have implications for how university systems approach suicide risk assessment. Although any suicidal ideation warrants clinical attention, presence of suicidal thoughts may have limited predictive potential in identifying those at risk of a future suicide attempt without considering additional risk factors. This may suggest that universal screening approaches focused primarily on detecting suicidal ideation potentially identify large numbers of students requiring support while having limited precision for identifying those at highest risk of a future suicide attempt. ^{Reference Klonsky, Dixon-Luinenburg and May18} Alternative approaches might include screening that considers both ideation and broader risk profiles, or enhanced support for students with ideation who demonstrate other high-risk factors, such as previous suicide attempts, previous intervention for mental health difficulties and disclosure of mental health disability at university entry.

Strengths and limitations

Key strengths of this study include the large, multi-institutional sample and longitudinal design capturing changes over the academic year, use of validated measures and robust statistical methods. Additionally, examination of both baseline risk factors and trajectories of change provides further understanding of suicide risk patterns among UK university students.

Several limitations should be noted. First, although the sample is representative of student populations, it may not be representative of all UK young adults; the majority of students identified as female, White and did not report a disability, whereas male students and those from minoritised backgrounds generally present with greater risk, ³ and may be underrepresented in the study. Second, suicide risk was assessed with self-report questions from the Columbia Suicide Severity Rating Scale, rather than measures from clinical interviews and case records, potentially limiting specificity of measurement. The ideation question in particular may have been subject to variable interpretation, capturing momentary thoughts as well as active planning, potentially elevating lifetime prevalence relative to meta-analytic estimates. ^{Reference Crispim, dos Santos, Frazão, Frazão, de Albuquerque and Perrelli7} Third, the sample was drawn from five UK universities and may not be representative of all UK higher education institutions. Moreover, although the sample included approximately 3000 participants, a relatively small number of students experienced new-onset suicidal ideation or made suicide attempts, limiting statistical power for some analyses; these subgroup findings should be considered exploratory. Finally, data were analysed over the course of a single academic year, making it difficult to draw conclusions about longer-term trajectories beyond the follow-up period. Additionally, data collection (2021–2025) overlapped with later stages of the COVID-19 pandemic, which disrupted social connectedness, academic experiences and support services, and may partly contribute to observed prevalence and within-year change; although academic year was adjusted for, generalisation to future cohorts warrants caution.

In conclusion, the present study provides important insights into patterns of suicidal ideation and attempts among UK university students. Findings suggest that minoritised and disadvantaged groups experience elevated baseline suicide risk, and that new onset of suicidal ideation and attempts during university are rare and difficult to identify beyond symptoms of mental health deterioration. Moreover, results suggest suicidal ideation in isolation may have limited predictive ability in identifying those most at risk of a future suicide attempt. These results underscore the importance of early identification and support for at-risk students during the transition to university, as well as ongoing monitoring of mental health symptoms and the need for screening approaches that consider both ideation and factors associated with progression to attempts. With further research to refine how to target and implement early interventions, universities may be better able to support student mental health within higher education settings.