Dermatologically Tested and Other Phrases That Mean Less Than You Think

Pick up almost any moisturiser and the front of the box will tell you it was dermatologically tested. It is a reassuring phrase. It is also one of the least informative sentences in the industry, because it tells you that something happened without telling you what, to how many people, or how it turned out. None of the common label phrases are lies. Most of them simply commit the manufacturer to far less than a reader assumes.

What dermatologically tested actually commits a brand to

The phrase has no fixed definition in European or British cosmetics law. In practice it usually means a dermatologist supervised some form of skin-contact test, most often a patch test: the product is held against a small area of forearm or back skin under a dressing for a few days, and the site is graded for redness and irritation.

What that establishes is narrow. This material, on this patch of skin, on a panel of volunteers, did not provoke an obvious reaction in most of them. It says nothing at all about whether the product does anything. Petroleum jelly would pass. So would water.

The part worth knowing is the floor underneath the phrase. Every cosmetic legally sold in the EU or the UK already carries a safety assessment signed by a qualified assessor before it reaches a shelf. That is a legal precondition of being on sale, not an achievement to advertise. Dermatologically tested sits somewhere above that floor without saying how far above, which is exactly what makes it such a comfortable thing to print.

None of which means the products are bad. Plenty of brands that print it run genuinely careful tolerance work and could show you the file. The point is only that the phrase is not the evidence. The file is, and you are not shown the file.

Hypoallergenic is a comparison with nothing on the other side

Hypo means less. Less than what? There is no reference product, no threshold, no required panel, no number that a formula has to come in under. It is a comparative with the comparison missing.

European guidance on cosmetic claims does address the word, and sensibly: a brand using it should have designed the formula to reduce allergenic potential, and must not imply that no risk remains. That is reasonable as far as it goes. But designed to reduce describes an intention, and none of the decisions behind the intention appear on the label.

The deeper problem is that allergy is individual. If you are sensitised to a particular preservative, or to one constituent of a fragrance blend, a hypoallergenic product containing it will make you react precisely as anything else would. The word on the front cannot know your history. The ingredient list on the back can be checked against it in about twenty seconds.

For reactive skin, the declared allergen names are worth more than the claim, and trying a new product on a small area for several days before it goes anywhere near your face is worth more than both. If a reaction is persistent or severe, that belongs with a doctor rather than with a label.

Non-comedogenic and the rabbit's ear

There is no agreed standard for this one either. No required method, no threshold, no body that certifies it.

Most of the ingredient tables that circulate online, the ones that score oils from zero to five, descend from an assay developed on the inside of an albino rabbit's ear. Rabbit ear skin is unusually rich in follicles and unusually quick to react, which is what made it convenient. Dermatology has questioned for decades how well it predicts what happens on a human face, and the honest position is that it predicts it loosely.

There are two structural problems beyond the animal. Ingredients were often tested neat, at full strength, when in a real product they appear at a small fraction of that and behave differently once emulsified. And congestion is not a property of an ingredient in isolation. It is a property of that ingredient, at that level, in that base, on that person's skin, in that climate.

So the defensible version of the claim is modest: this formula did not visibly congest the skin of the people who tried it. Whether it does the same on you is a question only your own face can answer, over several weeks. Breakouts that do not settle are a matter for a doctor, not for the front of a box.

Clinically tested says a test happened

Tested is not shown. The word describes an activity, not a finding, and a great deal hides in that gap.

At one end of the range sits instrumental measurement: hydration by corneometry, barrier function by transepidermal water loss, standardised photography, a control group, assessors who do not know which arm is which. At the other end sits a questionnaire, handed to a panel of volunteers who knew they were trying a new product and were asked whether their skin felt smoother. Both are clinical tests. Both print identically on a box.

Under European rules a claim has to be supported by adequate evidence, and a perception study is legitimate evidence for a claim about how skin looks or feels. It was never meant to be evidence of anything biological. The gap opens when the phrase on the front implies the second while the file behind it holds the first.

There is a cost asymmetry underneath all this. A properly controlled instrumental study is a serious line in a budget. A perception panel is not. Both compress into the same two words, and the difference never reaches the shelf, though it does show up in where the money in a jar actually goes.

Natural, clean and chemical-free

None of these has a legal definition in cosmetics. Natural comes closest: a voluntary international standard sets out how to calculate a natural-origin index for an ingredient. It also counts heavily processed derivatives as natural in origin, so a material that has been through several chemical steps can still score well. A company can use the word accurately under that standard and a shopper can still finish with the wrong picture, because the word carries more romance than the calculation does.

Clean has no definition at all. It is a position rather than a property, and it works by implication: if this is clean, something else must be dirty. That implication is doing the selling, and it is not a claim anyone has to support, because strictly speaking nothing has been claimed.

Chemical-free cannot be rescued. Water is a chemical, and every plant extract is a mixture of hundreds of them. But the phrase is reaching for something real. People who say it mean nothing in here that I have read something frightening about, and that concern deserves an answer rather than a smirk. The answer is that risk comes from dose and context, not from origin. Several of the most reliable sensitisers in cosmetics are natural constituents of essential oils, which is why European rules require a defined set of fragrance allergens to be named on the label instead of disappearing inside the word parfum.

The origin question becomes interesting the moment it is asked precisely. Whether a peptide is extracted from a plant or built in a reactor has real consequences for purity, batch-to-batch consistency and how exactly you can specify the sequence. The single word natural on a carton flattens all of that into a mood.

What a label can actually tell you

The regulated parts of a label are the useful parts, and they are all on the back in small type. The ingredient list runs in descending order by weight down to one per cent, below which the order stops meaning anything. That single rule does more work than every phrase on the front combined, and the mechanics of it are worth learning properly once.

The open-jar symbol with a number of months in it is the period after opening. A printed date, where one is required, is the durability date. Both are regulated, both are honest, and both are mostly ignored, which is a shame, because an antioxidant that has been oxidising in a warm bathroom for two years is no longer doing the job you bought it for.

Packaging is evidence too, of a quieter kind. An opaque airless pump around something light-sensitive tells you the formulator was still thinking about the molecule after the box was opened. A wide-mouthed clear jar around the same molecule tells you something else. It is inference rather than a claim, which is precisely what makes it difficult to game.

Then there is the name and address of the responsible person, a real legal entity in the EU or UK who holds the safety file, and the batch code, which is how a product is traced or withdrawn. Neither is glamorous. Both mean that somebody can be found.

Why we are careful about this

We have an interest in this argument and it is only fair to say so. Zero Lines has six formulations and nothing on sale. Most of the phrases in this article are ones we cannot honestly print, and being unable to print them has turned out to be a useful discipline.

The bench work is finished. The stages that involve real skin over real weeks are not, so there are no results to describe and we will not imply otherwise. The programme each formulation is built to pass is published with the method attached, deliberately, so that when there is finally something to report you can hold the report against what we said we would do. It is a slow way to earn a sentence on a box. It is the only one we would be able to defend.

So here is a test you can apply to us as readily as to anyone else. When a phrase on a package moves you towards buying, ask what would have to be true for that phrase to be false. If nothing plausible would make it false, it is not telling you anything. If something specific and checkable would, it is. That is more or less the whole method.