Placebo effect examples show that an expectation of benefit can change how patients feel and, in some studies, what instruments record. The condition: an improvement only counts as treatment-driven when a placebo comparison group did not improve as much.
The cases below run from wartime saline injections to a modern heart-drug trial. For wider context on how to tell if a health study is reliable, start with the hub guide, then return here for the specific cases.
How expectation produces a placebo effect
The Harvard Health blog describes the placebo effect as a positive effect related to an expectation of benefit. A separate medical journal article frames it as a suggestion-based effect of a pharmacologically inert drug. In both descriptions, the active ingredient is absent and the response comes from what the person expects.
Expectation is not the only route. A review in a medical journal describes a conditioning effect, a learned response after personal experience. A person who has felt better after taking a pill before may respond to the next pill for that reason alone.
Placebos are not limited to sugar pills. BrainFacts notes that placebos can also be injections, devices and procedures. A sham surgery is the procedure version: patients undergo a simulated operation so researchers can ask whether the real procedure adds anything beyond the experience of being treated.
The same expectation mechanism works in reverse. The Harvard Health blog defines the nocebo effect as a negative side effect due to a placebo. The journal review lists drowsiness and headache as the most common nocebo effects. In trials, placebo-group participants can experience a side effect that was described to them in advance, which is why how a side effect is explained to participants matters.
Expectation also varies by person. BrainFacts reports that placebos are less effective at reducing physical pain complaints in people with mild to moderate Alzheimer’s disease, so the response is not universal.
Real placebo effect examples from research
BrainFacts reports that as many as one in three patients feel better after receiving a drug with no active ingredients. That is a ceiling for how common the response can be, not a typical rate for every condition. The cases below show where it has been observed and measured.
Saline treated as morphine
In a World War II case described by BrainFacts, soldiers were given saline as if it were morphine. About 40 percent of them reported that the treatment eased their pain. Pain is a self-reported experience, so a pain scale rating captures what the patient perceives. That is exactly why expectation can move it.
Heat pain and the brain
In a heat-pain study summarized by BrainFacts, imaging data showed that the placebo effect had suppressed incoming pain signals from the arm to the brain. This matters because it moves the placebo effect from “people say they feel better” toward a measurable change in how pain signals are processed.
A heart-drug trial and the placebo group
The REDUCE-IT trial, reported in 2019, found that cardiovascular events were 25% lower with the active drug than with placebo, according to the Harvard Health blog. A follow-up biomarker analysis found that biomarkers worsened in the placebo group with little change in the active group. That suggests part of the apparent benefit may have come from how the comparison group fared. It is a reminder that the placebo arm needs scrutiny as well as the drug arm.
Quick comparison of the cases
| Example | What was reported | Planning use |
|---|---|---|
| Wartime saline for pain | About 40 percent of soldiers reported pain relief | Shows how strongly a subjective symptom can respond to a belief about treatment |
| Heat-pain imaging study | Imaging showed suppressed incoming pain signals from the arm to the brain | Shows a measurable pathway rather than only a change in opinion |
| REDUCE-IT (2019) | 25% lower heart attack or stroke rate versus placebo; biomarkers worsened in the placebo group | Check how the placebo group behaved before crediting a drug with the full gap |
| Patients with mild to moderate Alzheimer’s | Placebos less effective for physical pain | Placebo response differs between patient groups |
| Any inactive drug | As many as one in three patients feel better | Treat this as an upper range, not a promise |
Figures are from BrainFacts and the Harvard Health blog.
Placebo without deception
One of the more surprising findings is that deception is not always needed. The Imperial CRS summary notes that in several open-label placebo studies, patients’ symptoms improved even though they knew they were taking a placebo. This suggests expectation and ritual can contribute without a hidden inert pill, though the evidence described is of several studies and does not say it works for every condition.
How trials control for the placebo effect
The Harvard Health blog describes the randomized, double-blind, placebo-controlled trial as the most powerful type of clinical study. The medical journal review calls randomized, ideally double-blinded, placebo-controlled trials the gold standard. The reason is that a placebo control separates the effects of an intervention from natural variation in the course of disease and from placebo effects.
What double-blind means
In a double-blind trial, neither participants nor researchers know who is receiving active treatment and who is receiving placebo. Treatment assignment is coded and kept secret until the end of the study, or decoded at planned intervals. Blinding the researchers matters as much as blinding participants, because staff expectations can leak into how treatment is delivered and how outcomes are judged.

Checking that blinding worked
Blinding can fail, for instance when a side effect gives the active drug away. According to Harvard Health, studies can assess this by asking participants during or after the trial whether they thought they had active treatment or placebo. Random guesses or “I don’t know” answers indicate that blinding held up.
Where this started
Imperial CRS credits U.S. physician Austin Flint with the first clinical study comparing placebo to active treatment, for rheumatism, in 1863. It adds that placebos remain at the core of clinical trials and help establish efficacy. To see where placebo-controlled testing fits in the development process, read The 4 Phases Of Clinical Trials: How New Treatments Move From Lab To Clinic (What To Expect In 2026) – Testpledge.
Why placebo effects make supplement testimonials misleading
A testimonial usually reports one person who took a product and felt better. That story has no placebo comparison and no blinding. Three things could explain the improvement: the product, the person’s expectation, or the natural course of the condition. The journal review notes that placebo controls exist precisely to discriminate an intervention’s specific effects from natural variation in disease and placebo effects. A testimonial cannot do that.
The nocebo effect is a further reason for caution in the other direction: a person who expects side effects may notice and report them with an inert product too.

- Look for a control group. Claims backed by a randomized, placebo-controlled comparison carry more weight than before-and-after stories.
- Look for blinding. If participants knew what they were taking, expectation can account for part of any reported change.
- Treat self-reported symptoms carefully. Pain, fatigue and mood are the kinds of outcomes where expectation has been shown to matter.
- Separate feeling better from causing it. Whether a product caused an improvement is a causal question, covered in Causation vs Correlation: How To Tell What’s Really Driving Results (2026 Guide) – Testpledge.
None of this says an improvement is imaginary. The imaging evidence above suggests a placebo response can involve real changes in pain processing. The point is that a sincere testimonial cannot tell you which cause was responsible.
Frequently asked questions
Can a placebo help if the person knows it is a placebo?
Imperial CRS reports that symptoms did improve in several open-label placebo studies, where participants were openly told they were receiving a placebo.
What is the difference between the placebo and nocebo effects?
The placebo effect is a positive effect tied to expectation of benefit. The nocebo effect is a negative side effect caused by a placebo, with drowsiness and headache the most common examples.
Why do researchers keep treatment assignments secret?
Coding assignments until the end of the study, or decoding at planned intervals, keeps participants and researchers from knowing who has active treatment, which protects results from expectation-driven bias.
Do placebo effects only apply to pills?
No. Sugar pills are the most common form, but placebos can also be injections, devices and procedures, including sham surgery.
Does a good testimonial prove a supplement works?
No. Without a placebo-controlled comparison, product effects cannot be separated from expectation and natural changes in the condition.



