How Long Do People Live with Dementia?
Everybody wants to know and almost nobody asks out loud. Here are the numbers, and what they are actually useful for.
Key takeaways
- A 2025 analysis of 261 studies and more than five million people found roughly half of those newly diagnosed live beyond five years.
- Age at diagnosis moves it enormously. Diagnosed at 65, women averaged about 8 more years and men about 5.7. Diagnosed at 85: about 4.5 and 2.2.
- About 90% are alive one year after diagnosis. About 21% reach ten years.
- Women outlived men at every age. Alzheimer's ran longer than vascular, frontotemporal or Lewy body dementia.
- The average move into a nursing home came 3.3 years after diagnosis — roughly a third of remaining life.
- Once dementia is advanced the picture changes sharply. Median survival in that stage is measured in months, not years.
- In advanced dementia, pneumonia, fever and eating problems are the events that signal the end is close, and each carries high six-month mortality.
How long do people live after a dementia diagnosis?
Roughly half live beyond five years. That is the headline from the largest analysis ever done, and it hides more than it tells.
A 2025 review in The BMJ pooled 261 studies covering more than five million people diagnosed between 1984 and 2024. It found:
- About 90% alive one year after diagnosis.
- About half alive beyond five years.
- About 21% reaching ten years.
Which means the answer people are most afraid of — that a diagnosis is a short sentence — is wrong for most families. One in five is still here a decade later.
Age at diagnosis changes everything
The same disease, named at different ages, produces very different numbers.
- Diagnosed at 65: women averaged about 8 more years, men about 5.7.
- Diagnosed at 85: women about 4.5, men about 2.2.
Women outlived men at every age examined. Alzheimer's disease ran longer than vascular dementia, frontotemporal dementia or dementia with Lewy bodies.
And for someone diagnosed at 65, the analysis put the reduction in life expectancy at up to 13 years. That figure is the one that belongs in any conversation about young-onset dementia. The tragedy is not only how long the illness runs. It is how much did not happen.
Advanced dementia is a different question
Once someone reaches the advanced stage, prognosis is measured in months.
The landmark work here followed nursing home residents with advanced dementia and found a median survival of a little over a year from that point. More useful than the median were the events along the way.
Pneumonia, febrile episodes and eating problems were the three complications that dominated the final phase, and each one carried a high risk of death within six months — roughly four to five in ten after pneumonia or a fever, and around four in ten once eating problems began.
Those are not background details. They are the signal. When someone with advanced dementia develops pneumonia, or stops eating reliably, the question that should follow is not only how to treat it. It is whether the care being given still matches where the illness is.
The same body of work found that most family decision-makers did not know dementia was a terminal illness. Those who did understand the prognosis were far less likely to have their person put through burdensome interventions in the last months — hospital transfers, tube feeding, intravenous therapy — that changed nothing about the outcome.
What the late stage looks like, in order
Function usually goes in a recognizable sequence, and the sequence is more informative than any average.
- Speech reduces to a handful of words, then to none.
- Walking goes, then the ability to sit up unsupported, then to hold up the head.
- Swallowing becomes unreliable — coughing at meals, pocketing food, weight loss.
- Infections become frequent, particularly chest and urinary.
- Sleep takes over most of the day.
Any one of those is a reason to ask about hospice rather than a reason to wait for something more definite. Dementia patients are referred to hospice late and often, and families almost universally say afterward that they wish it had come sooner.
What to do with a number like this
Use it for planning. It is a poor instrument for anything else.
- Do not apply the average to your person. These are distributions. Somebody is at ten years and somebody is at two, and no clinician can tell you which one you are looking at.
- Use the 3.3-year figure as a planning horizon, not a prediction. Get the legal documents done, get the money mapped, and have the conversations while she is still in them.
- Ask the doctor the surprise question: “Would you be surprised if she died in the next twelve months?” A clinician's honest answer to that is worth more than any published median.
- Watch function, not the calendar. Walking, sitting, swallowing, infections. Those tell you where you are.
- Ask about hospice earlier than feels right. The most common regret is waiting.
Knowing roughly how much time there is does not make it shorter. It makes it possible to spend it on purpose.
Frequently asked questions
Why do the numbers I read vary so much?
Does an earlier diagnosis mean a longer illness?
Is young-onset dementia faster?
How do I know when we are close to the end?
Three years goes faster than anyone expects.
Day to Day Dementia is built by people who learned this the hard way and wrote it down so you would not have to. Join the waitlist for the guides as they are released.
Become a founding memberSources
- Survival after dementia diagnosis: systematic review and meta-analysis. The BMJ, 2025. 261 studies, more than 5 million people diagnosed 1984–2024.
- Mitchell SL, Teno JM, Kiely DK, et al. The Clinical Course of Advanced Dementia. New England Journal of Medicine, 2009; 361:1529–38.
- Alzheimer's Association. Stages of Alzheimer's.