Placebo and blinding in clinical trials and what it means for you

If you are thinking about joining a clinical trial, the word placebo can be the biggest worry. Nobody wants to spend a year taking a dummy pill while their condition gets worse. Yet placebo-controlled trials are often the only way to find out whether a treatment works, and sometimes they show that a popular idea does harm.

This guide explains what a placebo is, why trials use one, how blinding works and what it protects against, and what your rights are if you join a placebo-controlled study. The worked example is NOPARK, a Norwegian trial of the supplement nicotinamide riboside in early Parkinson's disease, published in JAMA on 8 October 2026. We covered the trial in our NOPARK report.

This guide is for general information and is not medical advice. Please discuss any treatment or trial with your own doctor.

What a placebo is

A placebo is a dummy treatment that looks, tastes and is taken exactly like the real one but has no active ingredient. In a pill trial, it is a capsule of the same size and colour. In an injection trial, it is usually a saltwater solution.

The point of a placebo is to give the comparison group the same experience as the treatment group in every way except the drug itself. They take the same number of pills, attend the same visits, and have the same tests. Any difference at the end can then be put down to the drug with more confidence.

Why the placebo effect matters

People often feel better when they believe they are being treated. They may also report symptoms differently, or a doctor may rate them more kindly, when everyone knows a treatment has been given. These effects are real and can be large, especially for outcomes such as pain, mood, tiredness, or movement scores that rely on judgement.

Diseases also change on their own. Some conditions fluctuate, and people often join trials when they feel at their worst, so many improve somewhat over time whatever they receive. Without a placebo group, all of that improvement would be credited to the drug.

What blinding means

Blinding means keeping the treatment assignment hidden. In a single-blind trial, participants do not know which group they are in. In a double-blind trial, neither the participants nor the doctors and staff running the trial know. Some trials go further and also blind the people who analyse the data or judge the outcomes.

Blinding stops expectations from leaking into the results. If a doctor knows a patient is on the new drug, they may unconsciously score a symptom test more favourably, or look harder for side effects in the placebo group. Double-blinding removes that risk.

Blinding is usually combined with randomisation, in which a computer decides by chance who gets what. Randomisation makes the groups similar at the start. Blinding keeps them treated the same afterwards.

The NOPARK example

Nicotinamide riboside (NR) is a form of vitamin B3 sold as a supplement. It raises levels of a molecule called NAD, which is involved in how cells make energy. Earlier small studies had suggested that NR might help brain cells in Parkinson's disease, and some people had started taking it on their own.

NOPARK, registered as NCT03568968, tested that idea properly. It randomised 410 people with early Parkinson's disease to NR 500 mg twice daily or a matching placebo for 52 weeks: 206 to NR and 204 to placebo. Results were analysed for 393 people.

The main outcome was the change in the MDS-UPDRS, a standard rating scale for Parkinson's in which higher scores mean worse symptoms.

The results were:

• the NR group's score rose by 2.45 points
• the placebo group's score fell by 0.27 points
• the difference was 2.72 points in favour of placebo, with a 95% confidence interval of 0.47 to 4.98 and a p-value of 0.02
• non-motor symptom scores were also worse on NR, by 4.22 points

What the placebo group showed

Without a placebo group, a trial of NR would have seen symptoms in the treated group creep up by 2.45 points over a year. Since Parkinson's is a progressive disease, that might have looked like the expected course. Nobody could have said whether the supplement helped, did nothing, or made things worse.

The placebo group gave the answer. People on the dummy capsule did slightly better over the year, so the supplement group's decline was larger than it would have been without the supplement. The result went the opposite way to what researchers had hoped.

How big is a difference of 2.72 points? The study reported clinically important change thresholds of 5.2 and 6.7 points, so the gap was below the level that people would usually notice in daily life. The confidence interval also runs from a very small difference up to nearly 5 points. Still, the direction was clear enough that the trial ruled out the benefit that had been hoped for, and raised a question about possible harm.

This is one of the strongest reasons for placebo-controlled trials. They can show that a widely used product may not help, or may make things worse, which no uncontrolled study could do.

Is it ethical to give a placebo?

Ethics committees ask this question about every placebo-controlled trial. The broad principles, set out in documents such as the Declaration of Helsinki, are:

• A placebo is usually acceptable when there is no proven treatment for the condition, or when the new treatment is being added on top of standard care.
• Using a placebo in place of a proven effective treatment is only acceptable when there are strong scientific reasons and the risk of serious or lasting harm from going without is very low.
• Participants must be told clearly that they may receive a placebo, and must agree.

In NOPARK, participants could carry on with their usual Parkinson's medicines. The placebo replaced only the supplement being tested, for which there was no proven benefit.

Many trials use an "add-on" design for this reason. Everyone gets standard treatment, and half also get the new drug while half also get a placebo.

Your rights in a placebo-controlled trial

If you join a trial with a placebo arm, you have the same rights as in any trial:

• You will be told the chance of getting a placebo before you agree. Common ratios are 1 to 1 or 2 to 1 in favour of the active drug.
• You can leave at any time, without giving a reason and without it affecting your normal care.
• Your doctors can break the blind in an emergency if they need to know what you are taking.
• You should be told what happens at the end of the trial, including whether you can access the treatment if it works.

Some trials offer everyone the active drug after the blinded period ends, in what is called an open-label extension. Ask about this before you sign up.

Questions to ask before joining

• What are my chances of getting the placebo?
• Will I keep taking my usual treatments?
• What happens if my condition gets worse during the trial?
• Will I be told which group I was in when the trial ends?
• Can I get the treatment afterwards if it works?

When trials do without a placebo

Not every trial uses a placebo. When a good treatment already exists, a new drug is usually compared with it head to head. Some treatments cannot be disguised, such as surgery or a procedure that removes a blood clot from the brain. In those cases, trials may be open-label, with everyone knowing who got what.

Our guide to open-label versus blinded trials explains how researchers reduce bias when blinding is impossible, using the LATE-MT stroke trial.

Spotting a well-blinded trial in a report

When you read a trial report, look for these points:

• the words "double-blind" or "placebo-controlled" in the title or methods
• a description of how the placebo matched the real treatment
• whether outcomes were judged by people who did not know the assignment
• whether the authors checked that blinding held, for example by asking participants which group they thought they were in

A treatment with obvious side effects, such as a distinctive taste or a skin reaction, can give the game away. Good trials think about this in advance.

Placebo results in other trials

Placebo groups appear in many of the studies behind recent approvals. In the trials behind Hibsago for chronic hepatitis B, nobody on placebo reached a functional cure, against about one in five on the drug. In the Icotyde psoriasis trials, 8.3% of people on placebo had clear or almost clear skin at week 16, which shows how much improvement can happen without the active drug.

Those placebo numbers are what make the drug results meaningful. A treatment can only be judged against what would have happened anyway.

Related reading

For more on this, see Nicotinamide riboside in Parkinson's, the NOPARK results, How to read clinical trial results and How to find a clinical trial that fits you.