Suicide prevention research has historically focused on individual-level risk factors, including mental health conditions, previous self-harm and/or suicidal behaviour, and substance use. While these remain important, an exclusive focus on them can miss a more complex reality; that some of the strongest indicators of suicide risk are embedded in broader social and structural conditions. A recent population-based cohort study by Nilsson and colleagues (2025), published in The Lancet Public Health, highlights this by exploring the connections between homelessness, psychiatric disorders, and the risks of suicide and self-harm.
Although elevated suicide risk among people experiencing homelessness has been well documented (Ayano, Tsegay & Abraha et al., 2019; see also earlier discussion on The Mental Elf blog), this recent study is notable for both its scale and its theoretical framing. Rather than treating homelessness as a secondary or mere contextual factor, the authors position it as a primary exposure; one that intersects with mental illness, but is not reducible to it. For those working on issues of inequality, trauma, or marginalisation, the findings prompt a rethinking of how suicide risk is conceptualised.
Focusing on individual-level risk factors for suicide risks ignoring a more complex reality; that some of the strongest indicators of suicide risk are embedded in broader social and structural conditions.
Methods
The study is based on a large, population-level cohort using national registry data from Denmark, which allows the researchers to follow people over time and look at how homelessness is linked to later suicidal behaviour. Using this kind of dataset means the analysis isn’t limited to clinical samples and instead reflects patterns across the wider population.
The main elements of the design are:
- Exposure: Homelessness is treated as the key exposure, identified through administrative records.
- Outcomes: Both suicide and self-harm are included, giving a broader picture of self-harm and suicidal behaviour, rather than focusing only on deaths.
- Adjustments: The comprehensive analysis considers psychiatric disorders as well as a range of socio-demographic factors.
- Approach: The cohort design makes it possible to examine how homelessness is associated with varied outcomes over time.
Overall, the study is set up to look at homelessness as a central factor, while still recognising its relationship with mental health and wider social circumstances.
Results
The findings are clear and concerning. People who had experienced homelessness were at substantially higher risk of both suicide and self-harm compared with the general population.
- For suicide, rates were around two to four times higher depending on sex (males aMRR = 2.2, 95% CI 2.0 to 2.4; females aMRR = 3.8, 3.2 to 4.5).
- For self-harm, the risk was also markedly elevated (males aMRR = 3.5, 3.4 to 3.6; females aMRR = 3.3, 3.1 to 3.5).
- Importantly, this increased risk remained even after accounting for psychiatric disorders and other demographic factors.
- The risk was highest among those currently experiencing homelessness, particularly for women, where suicide rates were about eight times higher than those with no homelessness (aMRR = 8.0, 5.5 to 11.6).
This is a crucial point because it challenges a common assumption in the field; that the relationship between homelessness and suicide is mainly explained by higher rates of mental illness. The study shows the picture is much more complex. Although psychiatric disorders were strongly associated with increased risk, they did not fully account for the relationship. Even after adjusting for psychiatric diagnoses and other demographic factors, homelessness remained independently associated with elevated rates of suicide and self-harm. In some groups, the combined effect was particularly striking; for example, people experiencing homelessness who also had a psychiatric disorder had substantially higher suicide rates than those with neither exposure.
The authors also show that risk is particularly high when homelessness and psychiatric disorders occur together. Rather than operating independently, these factors appear to interact, leading to substantially higher levels of risk than either exposure alone.
In short, the study suggests that housing instability is not just a background characteristic; it is actively tied to suicidal behaviour and self-harm in its own right.
Key findings
- People with experience of homelessness had a significantly higher risk of suicide compared with the general population (approximately 2–4 times higher after adjustment; males aMRR = 2.2, 95% CI 2.0 to 2.4; females aMRR = 3.8, 3.2 to 4.5).
- The same pattern was seen for self-harm, with markedly elevated rates in the homelessness group (males aMRR = 3.5, 3.4 to 3.6; females aMRR = 3.3, 3.1 to 3.5).
- These associations remained even after adjusting for psychiatric disorders and socio-demographic factors.
- The combination of homelessness and psychiatric disorder was associated with especially high risk (aMRR ~19.9 in men).
- Risk was not fully explained by mental health diagnoses alone, suggesting an independent effect of homelessness.
Overall, the results point to a need to think beyond diagnosis-based explanations and consider the role of severe social disadvantage in shaping suicide risk.
This population cohort study found that people experiencing homelessness were two to four times more likely to experience suicide.
Conclusions
Rather than conceptualising homelessness solely as a consequence of mental illness, these findings situate it within broader patterns of social and economic disadvantage and systemic failure. In this case, homelessness is characterised by sustained exposure to instability, trauma, stigma, and restricted access to health and social care; reflecting breakdowns in housing, welfare, and support systems. These structural conditions often exacerbate pre-existing mental health difficulties and heighten vulnerability to self-harm and suicide (Fazel, Geddes & Kushel, 2014).
This reframes how risk is understood in this context. Risk is not solely an attribute of individuals but is also produced and shaped by the social and material conditions in which people live. While the identification and treatment of mental illness remains a necessary component of suicide prevention, it is unlikely to be sufficient in the absence of meaningful action on housing insecurity and the structural inequalities that underpin these patterns of instability and distress.
Overall, the study lends support to a more integrated approach to suicide prevention, combining timely and effective mental health care with interventions addressing homelessness, poverty, and wider social disadvantage.
These findings reframe suicide risk as being produced and shaped by the social and material conditions in which people live, not solely as an attribute of individuals.
Strengths and limitations
This study has several important strengths. Its population-level design and large sample size provide strong statistical power and a level of generalisability that is uncommon in this field. The longitudinal design also allows for temporal ordering of exposure and outcome, strengthening causal inference. In addition, the inclusion of both psychiatric and social variables enables a more comprehensive analysis than studies focused on a single domain.
However, a few limitations should be noted:
- Under-identification of homelessness:Â Administrative data are likely to capture only the most visible or acute cases, potentially excluding people in more precarious or hidden forms of housing insecurity. As a result, the full spectrum of homelessness may not be represented.
- Reliance on registry-based diagnoses:Â Psychiatric diagnoses depend on contact with services, meaning the most marginalised individuals may be under-recorded. This may introduce systematic bias in how both exposure and outcome are measured across groups.
- Lack of lived experience perspective:Â While quantitative data identify important population-level associations, they cannot fully explain the meanings, pathways, or contextual factors involved. Incorporating qualitative evidence would deepen understanding of how homelessness and suicidal behaviour are connected.
Whilst this study uses a large sample size to identify important population-level associations, the data cannot fully explain the meanings, pathways or contextual factors involved.
Implications for practice
The implications of this study are both clear and important.
Research implications
There remains a need to move beyond binary approaches that separate social and clinical risk factors. Future work should adopt intersectional frameworks that examine how homelessness interacts with other axes of inequality, such as ethnicity, sexual orientation, gender identity, and migration status. Greater integration of lived experience is also essential, particularly in understanding pathways into and out of risk.
Policy implications
Perhaps the most important message is that suicide prevention cannot be disentangled from social policy. If homelessness independently increases risk, then interventions that fail to address housing are, by definition, incomplete.
This points towards the importance of upstream strategies, including:
- Housing-first approaches.
- Income support and social protection.
- Policies aimed at reducing poverty and inequality.
Such interventions are often discussed in public health, but less frequently integrated into mainstream suicide prevention frameworks.
Practice implications
At the service level, there is a need for more integrated models of care that connect mental health and housing support. Clinicians should be attuned not only to individual symptoms, but to the structural conditions shaping their patients’ lives. Risk assessment that ignores housing instability is likely to miss a key part of the picture.
At the same time, caution is needed to avoid further medicalising distress that is rooted in social deprivation. Expanding access to mental health services is important, but it should not substitute for addressing the conditions that generate distress in the first place.
Suicide prevention cannot be disentangled from social policy. If homelessness independently increases risk, then interventions that fail to address housing are, by definition, incomplete.
Statement of interests
Dr Emma Rebecca Wallace has no conflicts of interest to disclose. This work was completed independently and is not connected to, or funded through, any current research projects or grants within the University of Birmingham. The views expressed are those of the author alone and do not reflect those of the University or any affiliated research groups.
Editor
Edited by Laura Hemming.
Links
Primary paper
Nilsson, S. F., Laursen, T. M., Erlangsen, A., Hawton, K., Nordentoft, M., & Fazel, S. (2025). Homelessness, psychiatric disorders, and risks of suicide and self-harm: a population-based cohort study. The Lancet Public Health, 10(7), e559-e567.
Other references
Ayano, G., Tsegay, L., Abraha, M. et al. Suicidal Ideation and Attempt among Homeless People: a Systematic Review and Meta-Analysis. Psychiatr Q 90, 829–842 (2019). https://doi.org/10.1007/s11126-019-09667-8.
Cummins, I. Are homeless people more likely to die by suicide? The Mental Elf, August 2021.
Fazel, S., Geddes, J. R., & Kushel, M. (2014). The health of homeless people in high-income countries: descriptive epidemiology, health consequences, and clinical and policy recommendations. The Lancet, 384(9953), 1529-1540.
