How to read evidence

Pain Scores, WOMAC And What A Trial Result Actually Measures

Every claim in this category rests on a trial, and almost every trial rests on a questionnaire. Knowing what those questionnaires measure, and what a change in one does and does not mean, is the single most useful thing a buyer in this aisle can learn in ten minutes.

By Artivorin Team · Last updated October 29, 2026 · 9 min read · How to read evidence · Reviewed by Dr. Blake Peterson, MD

What a symptom score actually is

A symptom score in this field is a structured questionnaire. A patient is asked how much pain they have had doing particular things, how stiff the joint has been and at what times of day, and how much difficulty they have had with ordinary tasks such as stairs, getting out of a chair or putting on socks. Each answer becomes a number and the numbers are summed.

The instrument used in most osteoarthritis research asks about all three of those domains, which is why trial reports talk about pain, stiffness and function as separate figures that also roll up into one.

It is a subjective measure in the sense that the patient reports it, and it is a rigorous one in the sense that everybody uses the same questions in the same order, so results from different trials can be compared. Both halves of that are true and the second is why the field relies on it.

What a change in a score means

Researchers do not treat any change as meaningful. There are agreed thresholds for the smallest difference a patient would actually notice, and a result that clears statistical significance without clearing that threshold is a result that did not matter to anybody in the trial.

That distinction is the first thing lost when a trial becomes a bullet point. A page saying an ingredient significantly improved pain scores may be describing a change no participant would have been able to feel.

Why the field settled on questionnaires

Because the alternative measures are worse for this purpose, which is counter-intuitive until you look at them.

Imaging seems more objective and correlates poorly with how people feel. Two joints that look similar on a scan can produce very different amounts of pain, and a joint that looks worse over a year can hurt less. Measuring structure would be measuring something other than the thing a patient came in for.

Blood markers are objective and are a step further removed again. They describe a process rather than an experience, and a great many things move a marker without changing anything anybody would notice.

So the field measures what it is actually trying to change, which is how somebody feels and what they can do. That is the right decision and it comes with the honest limitation that patients know they are in a trial and expectation moves scores.

Supplement trials in this category overwhelmingly use the first two. That is worth knowing before reading the word proven anywhere.

The five kinds of measurement, and what each buys

Reading the table above, the useful skill is to look at any claim and ask which row it came from.

The 2014 curcuminoid trial and the 367-patient ibuprofen comparison both sit in the first two rows: symptom and function scoring in people with knee osteoarthritis. That is the right kind of evidence for a claim about how a joint feels.

a 2025 systematic review of inflammatory biomarkers in knee osteoarthritis sits in the third row. It is a legitimate piece of work and it answers a different question, and a claim about how you will feel cannot be built on it alone.

Nothing in this literature sits in the fourth row with a positive result on cartilage in humans. That is why a claim about rebuilding a joint is the one this desk quotes rather than repeats wherever it appears on a label.

How good are the trials themselves

A question almost nobody asks, and the answer is genuinely mixed.

A Bayesian network meta-analysis of curcumin dosing pools the field and models the dose relationship. a 2025 critical appraisal of the systematic reviews and meta-analyses in this area does something different and more uncomfortable: it examines how well the underlying reviews and trials were conducted rather than simply combining their results.

That kind of appraisal matters because pooling twenty small trials of uneven quality produces a confident-looking number built on shaky foundations. A meta-analysis is only as good as what went into it, and a critical appraisal is the field checking its own work.

The fair summary is that this is a real effect with real methodological argument around it, which is a more useful sentence than either proven or debunked.

• A single small trial is a reason to look further, not a reason to believe.
• A meta-analysis is stronger, and inherits every weakness of what it pooled.
• A critical appraisal of the meta-analyses is stronger again, and is almost never cited on a product page.
• A guideline is where all of that ends up once a professional body has weighed it.
• The 2019 rheumatology guideline is the top of that ladder for this condition.

Translating the phrases on a product page

The fourth row is the only one where the phrase and the evidence behind it are the same kind of thing. That is the phrase worth paying for.

The fifth row is the one to be most careful with, and it is common. A claim about rebuilding or repairing a joint is a structural claim, structural claims need imaging evidence, and imaging evidence in this literature is thin and does not support it.

The second row is the most common of all. Clinically studied ingredients is a phrase that can be entirely true while telling you nothing, because it does not say at what dose, in what population, measuring what, or whether the amount in the capsule resembles the amount in the trial.

Using this on your own sixty days

The last useful move is to borrow the method. A trial works because it measures the same things at the start and at the end, in the same way, and writes them down.

You can do a crude version of that in under a minute a week, and it converts a purchase into an experiment.

1. On day one, before the first capsule: minutes of stiffness in the morning, worst pain in the past week out of ten, and one activity you have been avoiding.
2. The same three, once a week, on the same day of the week.
3. A short note of anything else that changed: weather, activity, a holiday, new shoes.
4. At week six, read day one against week six rather than recalling it.
5. Decide inside the window, and use it if the answer is no.

Doing that puts you ahead of most people in this category, who take a capsule for six weeks and then try to remember how they felt in the first one. Memory is not a control group. The results timeline on this website sets out what buyers here describe in each week, and it is a useful thing to compare your own notes against once you have some.

References