From the Editor
The loss of her job. A couple of poorly timed moves, motivated by her need to help care for an ailing parent. Financial woes. These problems converged, and my patient found herself ill and homeless. What were the potential implications for mortality?
In a new paper published by The Lancet Public Health, Sandra Feodor Nilsson (of Copenhagen University) and her co-authors attempt to answer that question. Drawing on Danish data involving more than six million adults, they found that men and women live far fewer years when experiencing homelessness, 15.9 years and 15.3 years, respectively. Though past work has examined the topic, Nilsson et al. offer a more comprehensive look. “Life-years lost exceeded those observed in individuals with schizophrenia, alcohol use disorder (for males), and drug use disorder.” We consider the paper and its implications.
In the second selection, from JAMA, Drs. Kirk B. Fetters and Joshua A. Barocas (both of the University of Colorado) describe recent cuts to Housing First programs in the United States and weigh its impact. They note the evidence for this approach. “Because the health care system is already overburdened and the homelessness crisis is worsening, maintaining and expanding Housing First is not only compassionate but also imperative.”
And in the third selection, Sean Kinsella writes personally about homelessness and addiction. In an essay for the Irish Examiner, he discusses his deep problems – and his recovery. He advocates that we do more for those who are homeless. “I wasn’t seen. I wasn’t heard. I was a file. A risk to be managed.”
DG
Selection 1: “Life-years lost in people experiencing homelessness and other high-risk groups in Denmark: a population-based, register-based, cohort study”
Sandra Feodor Nilsson, Annette Erlangsen, Camilla Munch Nielsen, et al.
The Lancet Public Health, September 2025
People experiencing homelessness represent some of the most vulnerable population groups in high-income countries, with mortality rates 3-12 times higher than in the general population. In some countries, such as the USA, mortality rates among people experiencing homelessness have increased over the past decade. Particularly, high risks have been observed for deaths due to external causes, including suicide and unintentional injuries, but also all-cause mortality, including during the COVID-19 pandemic. Previous findings further indicate higher rates of sudden death among individuals experiencing homelessness compared with housed individuals.
Approximately one-third of deaths among people experiencing homelessness are considered preventable or amenable to timely health care. Contributing factors include high rates of psychiatric disorders and chronic medical conditions, although homelessness itself is also an independent predictor of mortality.
Previous estimates of life expectancy in this population have been limited by methodological challenges, particularly in capturing transient housing situations (ie, individuals who experience homelessness at different stages of life). Calculating life-years lost provides a more accurate approach, as it accounts for age at first shelter contact, and provides estimates that are both directly comparable across populations and easy to interpret.
Here’s what they did:
- They conducted a “nationwide, register-based cohort study” of people aged 17-100 years living in Denmark between 2002 and 2020.
- They drew on national databases, linking the Danish Civil Registration System to the Homeless Register and the Cause of Death Register.
- Primary outcome: overall and cause-specific mortality during follow-up.
- They did different statistical analyses to estimate mortality rates per 10 000 person-years at risk and adjusted mortality rate ratios (using Poisson regression analysis). They “calculated both average and age-specific life expectancy and life-years lost for people experiencing homelessness and compared these figures with those of the general population, as well as with individuals with [mental disorders].”populations and easy to interpret.”
Here’s what they did:
- Among 6 286 512 Danish residents, 0.9% had at least one homeless shelter contact during the study period.
- Mortality. Among people with a history of homelessness, the mortality rate was 240.3 per 10 000 person-years; in the general population, it was 117.8 per 10 000 person-years.
- Mean excess life-years. They were 15.9 for males and 15.3 for females compared with the general population.
- Comparison to mental disorders. The loss was “higher than in other high-risk groups, such as people with schizophrenia, alcohol use disorder, and drug use disorder (with and without homelessness history).”
- Causes. “Most of this excess mortality in people experiencing homelessness was attributed to external causes of death, psychiatric disorders, and diseases of the liver and digestive system.”
A few thoughts:
1. This is a good study published in a solid journal. Though they walk on familiar ground, the authors bring new and better data to the discussion.
2. The key findings: “individuals experiencing homelessness had an even higher excess life-years lost than individuals with schizophrenia or an alcohol use disorder or a drug use disorder.”
3. Ouch.
4. How does this compare with past work? They note a Canadian study (done during the pandemic) that found a 17-year average loss compared with the general population. “Using more precise methods, our study found even greater reductions – up to 19 years for males and 16 years for females at age 30 years, compared with earlier estimates.”
5. What accounts for the lost years? “Substance misuse, particularly untreated drug and alcohol use disorders, increases the risk of overdose, unintentional injuries, suicide, and liver disease. Delayed treatment and undetected psychiatric and somatic conditions are also likely to have a role, as individuals experiencing homelessness are less likely to receive timely care for serious disorders such as cardiovascular conditions. Frequent emergency department visits without adequate follow-up suggests a cycle of acute care rather than long-term management. Additionally, stigma might reduce health-care use, worsening both mental and physical health outcomes.” Thoughtful.
6. What should and shouldn’t be done from a policy perspective? See the next selection.
7. Like all studies, there are limitations. The authors note several, including possible underestimates of homelessness. And, of course, the data was from one country, Denmark.
The full Lancet Public Health paper can be found here:
https://www.thelancet.com/journals/lanpub/article/PIIS2468-2667(25)00171-9/fulltext
Selection 2: “Advocating for Maintaining Housing First Policies”
Kirk B. Fetters and Joshua A. Barocas
JAMA, 12 August 2025
Homelessness in the US is a humanitarian crisis, with more than 771 000 individuals experiencing homelessness in a single night in 2024, representing a 38% increase from 2023. Homelessness is also a health crisis. At least one-third of our nation’s homeless population are unsheltered, exposing them to extreme heat and cold; two-thirds have used illicit drugs regularly; and nearly 50% have a mental illness… Of the interventions that have been tested and implemented, helping people into stable housing without precondition of sobriety, otherwise known as the Housing First model, is a compassionate way to decrease homelessness, reduce health care use, and possibly improve health outcomes. Although the Housing First model previously received broad bipartisan support, the Trump administration appears ready to abandon this approach.
So begins a paper by Drs. Fetters and Barocas.
They review the history of the model. “The Housing First model was developed in the 1990s by Pathways to Housing in New York City and implemented more widely by the 2010s. It is a patient-centered approach to break the perpetual cycle of homelessness and chronic mental and physical health conditions by providing housing without preconditions of sobriety or treatment for serious mental health disorders (although treatment is offered to individuals in these programs). The Housing First model prioritizes safety, stability, and autonomy with the goal of engaging people experiencing homelessness in treatment for health conditions that may exacerbate the circumstances leading to homelessness.”
They note the shift in politics. “Housing First programs are among many housing- and health-related policies being targeted for funding cuts and program discontinuation, following through on unfinished policy priorities during the first Trump administration. In September 2019, President Trump’s Council of Economic Advisors published a report titled ‘The State of Homelessness in America’ that questioned the effectiveness of a Housing First model.” It forwarded other ideas, like increasing housing supply by deregulation of the housing sector and controlling illicit drug trafficking. They note recent cuts to Housing First programs.
Drs. Fetters and Barocas argue against that.
Effectiveness
“Housing First programs are among the most effective policies there are for addressing homelessness, especially among people with SUDs and behavioral health issues. Housing First programs have been studied for decades and have demonstrated improved outcomes related to health care use, recidivism to incarceration, and housing stability, particularly in the population with SUD and behavioral health issues. A systematic review showed that, compared with Treatment First and abstinence-based programs, Housing First programs decreased homelessness by 88%, increased housing stability by 41%, decreased emergency department use by 5%, decreased hospitalization by between 7% and 36%, improved quality of life, and improved community integration. Since that review, the Denver Supportive Housing Social Impact Bond Initiative, a randomized clinical trial involving more than 700 participants who were heavily involved in the criminal legal system, found that a Housing First program reduced shelter stays, reduced total jail days by 27%, reduced emergency department use, and led to a high rate of retention in housing of 77% at 3 years.”
Cost effectiveness
“Housing First approaches are cost-effective. The most recent rigorous review on this topic included 20 studies in the US and Canada and found the median intervention cost to be $16 479 (IQR, $13 120-$26 452) per person per year. The median total societal benefit of $18 247 (IQR, $7522-$35 418) per person per year includes cost savings in health care, the criminal justice system, and other supportive government services. Few other policy interventions are as cost-effective.”
Impact on health
“The effect on clinicians and the US health care system is likely to be significant. Without access to Housing First programs and services, hundreds of thousands of individuals would face heightened risk for unsheltered homelessness, exacerbating exposure to illness, injury, and premature death. The substantial gains made combating overdose deaths, HIV and hepatitis C virus infections, and climate-related injuries could be reversed, putting further strain on health care workers and hospitals. Housing First programs shift health care use from unscheduled care to preventive outpatient care. Although housing appears to be a societal problem, supporting improved housing policies, including the effective Housing First model, has a direct effect on emergency departments, hospitals, and clinics.”
A few thoughts:
1. This is a good and thoughtful paper in an excellent journal.
2. The third point – on the impact on health and healthcare – may be subtle but important.
3. Housing First has been considered in past Readings. In a recent selection, writing for Nature Mental Health, Kerman and Dr. Stergiopoulos, offered an agenda, building on Housing First. That paper can be found here:
The full JAMA Viewpoint can be found here:
https://jamanetwork.com/journals/jama/fullarticle/2836034
Selection 3: “I was a homeless heroin addict so I know anyone can turn their life around with the right support”
Sean Kinsella
Irish Examiner, 21 July 2025
I didn’t choose to be homeless at 14. I didn’t choose the violence, addiction, or instability that filled my childhood home in Ballymun.
But when the system stepped in, it didn’t save me. It moved me. From one short-term placement to another. From emergency hostels to secure care units. From fear into deeper fear.
One of those placements was meant to be a place of rehabilitation. Instead, it was full of bullying, trauma, and isolation. I developed anxiety so severe I could barely breathe.
So begins an essay by Kinsella.
“And then came the spiral: addiction, prison, self-hatred. I injected heroin in hospital bathrooms, alleyways, and squats. I nearly lost my left arm to an infection that could have killed me. The worst part? I didn’t care.”
He talks about his recovery. “What saved me wasn’t a judge or a policy or a placement. It was recovery, community-led, relationship-based, and brutally honest. I met people who saw beyond the labels. People who said: ‘You matter.’ And for the first time in years, I started to believe it.”
“Today, I’m a father. A graduate. A recovery advocate.”
“I help people remember what they’re capable of.” He notes that he has started a business offering coaching and more.
He outlines an agenda for our society.
- Early intervention. “Catch the pain before it becomes crime. Invest in community mental health, trauma-informed education, and peer-led supports.”
- Aftercare that works. “Leaving care shouldn’t mean sleeping in a doorway. We need stable housing, mentorship, and access to addiction and mental health supports.”
- Lived experience leadership. “Those who have lived it, like me, must be at the policy table. We know what works. We know what hurts.”
- Addiction reform. “Addiction is a public health issue. Nobody heals in handcuffs. Nobody gets clean in a cell.”
“I’m not asking for sympathy. I’m demanding change. Because behind every chaotic young person is a child who was failed. And behind every person in recovery is proof people can come back from the edge. What they need is someone to walk beside them, not in judgement, but in belief.”
A few thoughts:
1. This is a moving essay.
2. This line is worth repeating: “I wasn’t seen. I wasn’t heard. I was a file. A risk to be managed.”
3. The call for those with lived experience to have “a seat at the table” is interesting and important.
The full Irish Examiner op ed can be found here:
https://www.irishexaminer.com/opinion/commentanalysis/arid-41672255.html
Reading of the Week. Every week I pick articles and papers from the world of Psychiatry.
Recent Comments