How long do people live with Parkinson's?

Ask the internet how long people live with Parkinson's and you get answers. Ten to twenty years. Over fourteen and a half years on average. Fifteen to twenty-five. Each stated plainly, each from a reputable-looking source, none agreeing with the others.

The disagreement is not sloppiness. It is the finding.

What the largest review found

A 2014 systematic review and meta-analysis gathered 88 studies of mortality in Parkinson's. Across them, mortality ratios — how much more likely someone with Parkinson's was to die in a given period than someone without — "ranged from 0.9 to 3.8, with major between-study heterogeneity".

Read that range again. At the bottom end, 0.9, is a study finding people with Parkinson's slightly less likely to die than controls. At the top, nearly four times as likely. These are not small differences in emphasis. They are different answers.

The reviewers were direct about the cause: Parkinson's "is associated with increased mortality, but major heterogeneity is seen in estimates of mortality, which is probably explained by variable methodology and patient selection".

In other words, the number you get depends enormously on who you counted and how you found them. A study recruiting from a specialist clinic gets sicker people than one recruiting from a population register. A study that starts counting at diagnosis gets a different answer from one that starts counting whenever it happened to find someone.

The most trustworthy subset gives a narrower answer

Not all 88 studies are equally informative. Inception cohorts — studies that enrol people at diagnosis and follow them forward, rather than picking up people at various unknown points in their illness — "were more consistent with a pooled mortality ratio of approximately 1.5".

That is the closest thing to a headline figure in the whole literature: among the best-designed studies, people with Parkinson's died at roughly one and a half times the rate of comparable people without it.

Note what that is not. It is not a life expectancy. It is a ratio, and it says nothing on its own about how many years anyone has. A ratio of 1.5 means something very different at 45 than at 80, because the baseline it multiplies is utterly different.

For an absolute figure, the review offers only this, from post-mortem studies: mean duration until death "ranged from 6.9 to 14.3 years". A range that wide, across studies of people who had died, tells you how unresolved this is.

Two findings that get left out

Mortality ratios rose with time within studies. The gap between people with Parkinson's and people without widened the longer studies followed them — which fits a progressive disease, and means a figure quoted at five years and one quoted at fifteen are not measuring the same thing.

"No robust evidence was found that mortality has decreased after the introduction of levodopa." This one is genuinely uncomfortable, and it is almost never mentioned in articles on this question. Levodopa transformed how people with Parkinson's live. Across these 88 studies, it did not show a robust effect on how long. Those are compatible — a treatment can improve years without adding them — but the second half tends to go missing.

What did predict mortality

Consistently, two things: "Increasing age and presence of dementia were most commonly associated with increased mortality."

Age is doing a great deal of work in every figure you will read. Most people are diagnosed with Parkinson's in their sixties or later, so a large share of the mortality in any Parkinson's cohort is the mortality of being that age. Studies that adjust well for this get lower ratios than studies that do not — which is part of why the range is so wide.

So what is the honest answer?

That the question, asked in the abstract, does not have a number as an answer. What exists is:

• Among the best-designed studies, a mortality ratio of about 1.5.
• Ratios across all studies from 0.9 to 3.8, mostly reflecting who was studied rather than genuine differences between populations.
• Post-mortem series with mean duration to death from 6.9 to 14.3 years.
• Age and dementia as the factors most consistently associated with dying sooner.

Anyone quoting a single confident figure has chosen one study, or averaged studies that should not be averaged. That does not make them dishonest. It makes the number less informative than it looks.

The version of this question that has a real answer is a personal one — age at diagnosis, how the disease has behaved so far, what else is going on — and that is a conversation with a neurologist rather than a search result. What the literature can tell you is what the published figures actually rest on, which is usually the thing nobody says.