Does starting levodopa early make Parkinson's worse?

There is a worry that has followed levodopa around for decades: that it is something you spend rather than something you take. Start it too early, the thinking goes, and you will have used up its good years by the time you need them most — or worse, hastened the underlying disease.

It is a reasonable thing to fear. It is also, unusually for a question of this kind, something that has been tested directly. Twice.

The first attempt raised more questions than it settled

ELLDOPA, published in 2004, gave 361 people with early Parkinson's either placebo or one of three daily doses of levodopa for 40 weeks, then withdrew the drug for two weeks before measuring everyone again. The withdrawal was the whole point: if levodopa were only masking symptoms, the treated groups should have snapped back to where the placebo group had got to.

They did not. Symptom scores had worsened by 7.8 points in the placebo group and by 1.9, 1.9 and −1.4 points in the three levodopa groups — the highest dose group finishing better than it started, two weeks after stopping the drug.

Then the imaging complicated it. In a substudy of 116 people, scans of the dopamine transporter declined more in the levodopa groups than in the placebo group. The clinical measure said one thing and the scan said the opposite. The authors did not resolve it, and said so plainly: the potential long-term effects "remain uncertain".

Two things can explain that split. Levodopa might genuinely slow the disease while the scan is misleading — the drug can alter the dopamine transporter it is being measured by. Or levodopa might have a benefit that outlasts two weeks of withdrawal, meaning the washout was simply too short. ELLDOPA could not tell these apart.

The second attempt was built to settle it

LEAP, published in 2019, used a design meant to close that gap. It randomised 445 people to either 80 weeks of levodopa, or 40 weeks of placebo followed by 40 weeks of levodopa. By the end, both groups had been on the same dose for 40 weeks. If starting earlier changed the disease itself, the early group should have stayed permanently ahead. If levodopa only treats symptoms, the two groups should have converged.

They converged. There was no disease-modifying effect over 80 weeks.

A five-year follow-up published in 2024 found the same thing further out. Scores did not differ significantly between the groups at three years or at five. At five years, dyskinesia had appeared in 46 of 160 people who started early and 62 of 161 who started late — if anything fewer in the early group, though not by a margin the study could call significant. Wearing off, and the total daily dose people had ended up on, did not differ either.

What this does and does not show

The specific fear — that early levodopa accelerates Parkinson's, or exhausts a finite supply of benefit — is not what these trials found. On the evidence here, starting later does not buy you a better position later. It buys you the months in between, spent with symptoms that were treatable.

Three limits are worth being clear about.

Dose is a separate question from timing. ELLDOPA found more dyskinesia, hypertonia, headache and nausea in the group on 600 mg daily than on placebo. That is a finding about how much, not about when, and the two get run together constantly.

Five years is not a lifetime. These are the longest randomised comparisons available, and Parkinson's is measured in decades. What they rule out is a large effect on the timescale they looked at.

Trial populations are not everyone. Both trials enrolled people with early disease who were, by design, well enough not to need treatment immediately. That is exactly the group the question is about — but it is not everyone the question gets asked about.

Why this one is worth knowing

Most questions in Parkinson's research resolve slowly, in fragments, across studies that disagree. This one is unusual: a specific, widely held, clinically consequential belief that was put to a direct test and did not survive it.

None of which is a reason to change anything you are doing. When to start levodopa is a decision with more in it than this single question — how much your symptoms cost you now, what else you have tried, what you want the next few years to look like. That conversation belongs with your neurologist or specialist nurse. What the trials can offer is one fewer thing to be afraid of while you have it.