Homelessness shortens lives 
Homelessness is both a housing crisis and a major public health crisis. People experiencing homelessness face more illness, injury, and early death. A recent meta-analysis found excess mortality across many causes. People also face major barriers to timely and respectful care.
A new Medical Care study examines these risks among veterans. Both authors are affiliated with the VA’s National Center on Homelessness Among Veterans. The Center develops research, programs, and training to address veteran homelessness. Homelessness is both a housing crisis and a major public health crisis.
The missing timeline
We know homelessness harms health, but not how those harms build over decades. Most studies capture one night, one program, or one short follow-up period. These snapshots cannot show a complete housing history.
Researchers rarely know when homelessness started, ended, or returned. We often lack its frequency, duration, and spacing across a lifetime. Federal reviews note serious limits in national homelessness measurement. We know homelessness harms health, but not how those harms build over decades.
A rare national view
VA data create a rare view of long-term outcomes. The 2004 cohort included 85,533 veterans. The 2014 cohort included 222,974 veterans. National administrative records followed both groups through 2024.
This design draws strength from the VA setting. Lead author Jack Tsai, PhD, MS, said the VA’s national reach made the long follow-up possible.
The VA is the largest integrated health care system in the United States, with more than 140 medical centers nationwide. That allowed us to follow veterans who moved between medical centers or traveled to different locations. We could examine long-term outcomes, including mortality and VA homeless program use, across many years.
The VA also connects health care, benefits, and homelessness services. So, VA data create a rare view of long-term outcomes.
What continued program use means
Continued program use can show both need and connection to care. In both cohorts, one-quarter of survivors used VA homelessness programs in 2024. About one-third used emergency or urgent care that year. Dr. Tsai emphasized that the study measured long-term service use, not each separate episode of homelessness.
“We did not focus on counting every episode of VA homeless program use. We wanted to understand who continued using these programs one or two decades later.”
Those visits may reflect persistent homelessness, a later return, or helpful reentry into services. Administrative records cannot fully separate these paths. That distinction remains an important next step for research, Dr. Tsai said:
“There are important clinical differences between veterans who experienced homelessness throughout the period and those who later returned to homelessness. Future research needs to evaluate those different trajectories.”
Other VA research shows that care location can affect hospital outcomes. Continued program use can show both need and connection to care.
A shocking mortality burden
Access to comprehensive care did not eliminate early death. Half of the 2004 cohort died within 20 years. About one-quarter of the 2014 cohort died within 10 years.
The mean age at death for homeless veterans was only 63.9 and 65.3 years. Dr. Tsai said many health risks may have accumulated long before the study identified veterans through homeless programs.
“I suspect much of the risk for premature mortality accumulated over many years. There is no single solution. These findings leave a great deal of room for earlier prevention, but an important question is when that prevention should begin.”
He added that prevention may need to occur at several stages of life.
“It could begin with adverse childhood experiences, continue into early adulthood, or take place later. We should study prevention strategies that operate at multiple points across the life course.”
Earlier research and a 2019 presentation[pdf] showed elevated deaths among younger veterans. VA research also found major years of potential life lost. In those studies, access to comprehensive care did not eliminate early, or excess deaths.
Health care cannot carry this alone
Health systems cannot replace stable housing and sustained support. Older age, medical illness, and alcohol use disorder predicted death. Different factors predicted continued homelessness program use.
The racial findings also require care. White veterans had higher mortality risk, while Black veterans had higher continued program use. Continued housing services can lower mortality risk for veterans facing housing instability. Health systems cannot replace stable housing and sustained support.
Treatment for homeless veterans must last
Treatment access matters, but continuity matters just as much. Most veterans with substance use disorders received some treatment. However, treatment generally lasted only a short time.
Short care may not match a long-term or recurring disorder. Stable housing can support health and continued treatment. VA has reduced veteran homelessness by 56% since 2010, but this study shows unfinished work. Treatment access matters, but continuity matters just as much.
Ben King