Four questions
These are in order of how often they settle the matter.
One: is there a mechanism, and does it exist? A claim that something repairs your barrier requires a route by which an applied material joins a production line that runs inside living tissue. There is not one. A claim that something slows water loss requires only that it forms a film, which is entirely plausible. The first is a mechanism that does not exist; the second is one that does.
Two: where did the number come from? Percentages of improvement, hours of hydration and proportions of users who agreed are the commonest form of unsupported specificity. Ask what was measured, in how many people, over what period, compared with what.
Three: does the study support the claim being made? This is the most frequent failure in writing that cites research at all. A laboratory measurement on isolated tissue does not establish an effect in a person. An effect in a person does not establish an effect at the concentration in a product. And an effect at one site does not transfer to another.
Four: what would count as this being wrong? A claim that cannot fail is not a claim. If the answer to every objection is that it works differently for everyone or that results take longer than you gave it, there is nothing there to assess.
The four questions applied to common claims
Worked examples, because the questions are easier to use once you have seen them used.
This cream repairs your skin barrier. Fails test one. It slows water loss, which lets the skin repair itself faster, which is a different and defensible claim.
Clinically proven to hydrate for seventy two hours. Fails test two until you know what was measured and how. Hydration of what layer, measured by what method, in how many people, in what conditions, and compared with what.
Contains an ingredient shown in studies to increase ceramide production. Usually fails test three. A demonstration in a laboratory system is not a demonstration in a product on a person's arm at whatever concentration is in the tub.
Your skin barrier is damaged and needs rebuilding. Fails test four as usually deployed, because no observation is offered that would show it was not.
Cold water causes dry skin. Passes test one, since immersion swells the cells and leaches the water attracting molecules inside them, and salt draws water at the surface afterwards. That is why it appears throughout this publication.
Applying it to this publication
It would be easy to write the article above and exempt ourselves. Here is where our own material sits.
What we assert as physiology is standard descriptive material about the structure of the stratum corneum, water loss, the effects of temperature and humidity, and the mechanics of friction and cracking. It is not novel, it is not ours, and it is available in dermatology and physiology sources including those we cite.
What we assert as a framework is the stressor and response pairing and the layer views. Those are ways of organising a decision written by this publication. Every one of them carries a note saying so, and none of them is a measurement or a study result. We say this on every single one.
What we deliberately do not assert is any number: no percentages, no hours, no thresholds, no quantities to apply, no temperatures, no durations expressed as figures. Where a reader wants a number, we say there is not a defensible one rather than inventing it. That is why our advice is expressed in directions rather than in values.
Where we could be wrong is in the relative weight we give things. We assert that indoor heated air over a season does more cumulative barrier damage than the outdoor sessions people blame. That follows from exposure duration and the physics of heated air, and it is a reasoning step rather than a measurement. If someone measured it and found otherwise, we would be wrong and we would say so.
| Test | The question | What it catches |
|---|---|---|
| Mechanism | What physically happens, step by step | Repairs, rebuilds, restores, detoxifies |
| Provenance of the number | Measured what, in how many, how long, against what | Percentages, hours, proportions who agreed |
| Fit of the evidence | Does the study support this specific claim | Laboratory findings presented as human outcomes |
| Falsifiability | What would count as this being wrong | Claims that survive every objection |
| Applied to us | Which of our statements are framework and which are physiology | We label every framework block as ours |
Ordering framework written by this publication from general skin physiology. It is not a measurement, it is not taken from any study, and no number in it is a reading.
Which sources to weight
For a British reader dealing with skin, a workable ordering.
NHS material for what to do and where to go. It is written to be acted on and it is conservative.
NICE and its Clinical Knowledge Summaries for what British clinical practice actually recommends.
The British Association of Dermatologists for patient information written by dermatologists.
The Primary Care Dermatology Society for material aimed at clinicians seeing skin in primary care.
Cochrane for whether an intervention has been assessed systematically and what was found.
The Health and Safety Executive for anything occupational.
PubMed if you want to read primary literature, with the caveat that reading a single paper well is harder than it looks and that abstracts routinely overstate.
Below those sit everything else, including us.
Corrections
If something on this site is wrong, tell us and we will correct it with a visible note. We do not remove accurate material at anyone's request, and we do not have advertisers who could ask, which is a structural property of the funding model rather than a virtue.
The one thing we will not do is soften a referral. Everywhere this publication says to get something looked at, it will keep saying it.