Ceramides, cholesterol and free fatty acids: the barrier lipid question

Ceramide NP, AP and EOP at 0.01 to 1 percent, cholesterol at 0.5 to 2 percent, the 3 to 1 to 1 ratio evidence and the solubility problem in a wax base.

Ceramides are the rare cosmetic ingredient where the chemistry and the marketing are both real and still fail to meet. Your stratum corneum genuinely is built from ceramides, cholesterol and free fatty acids, and a balm genuinely can carry all three. This page covers what the letters after the word ceramide mean, which grades you can actually buy, what they cost per tin at real use levels, why they fall out of a cooling wax base, and how far the barrier repair evidence stretches before it stops.

• INCI: Ceramide NP
• Typical 0.01-1%
• Melts near 100 C
• Dispersible, not oil soluble

What a ceramide is, and what the letters mean

A ceramide is a sphingoid base with a fatty acid amide-bonded to it. Both halves vary, and the letter code names both halves in order. The first letter is the fatty acid: N for non-hydroxy, A for alpha-hydroxy, O for omega-hydroxy, E for an omega-hydroxy acid that carries a second fatty acid esterified onto its far end. The second letter is the base: S for sphingosine, P for phytosphingosine, H for 6-hydroxysphingosine, DS for dihydrosphingosine. So Ceramide NP is a non-hydroxy fatty acid on phytosphingosine, and Ceramide EOP is the long, double-tailed one that rivets adjacent lipid layers together.

At least fifteen subclasses have been separated from human stratum corneum, and mass spectrometry keeps finding more. Together they make up roughly half the lipid mass of the intercellular matrix, with cholesterol and free fatty acids supplying the rest, in proportions set out in stratum corneum and barrier lipids. The barrier contains essentially no triglycerides, which is the whole reason a shea and jojoba balm cannot be described as replacing what the skin has lost. That distinction is worked through in moisturisers and the skin barrier.

Older ingredient lists use numbers instead of letters, and both still appear on shelf. Ceramide 3 is Ceramide NP. Ceramide 2 is Ceramide NS, which spent a period on labels as Ceramide NG while the nomenclature was harmonised. Ceramide 1 is Ceramide EOS, the sphingosine version of EOP. A supplier who still quotes numbers is not necessarily selling anything worse, but ask which subclass they mean before you write the label, because the two systems are not interchangeable on an ingredient list.

Which ceramides you can actually buy

The gap between the subclass list and the catalogue is wide. Almost all commercial ceramide is made by fermentation to phytosphingosine followed by acylation, which is why the phyto subclasses dominate what is for sale.

Most small makers end up buying a pre-dispersed blend rather than a pure powder. A typical one lists three or four ceramides alongside cholesterol, phytosphingosine and an emulsifier such as a lactylate, at a total ceramide content of a few percent of the blend. It is easier to handle and much cheaper per gram of blend, but the emulsifier in it exists to build lamellae in a water-containing product. Dropped into a waterless base it has nothing to emulsify, and it will not disperse the ceramide any better than heat and stirring would.

Cholesterol, the fatty acid, and what 3:1:1 really says

Cholesterol (INCI Cholesterol) is a crystalline solid melting at 147 to 150 C under the pharmacopoeial method. It is used at 0.5 to 2 percent, dissolves in warm oil at those levels and tends to seed crystals above roughly 2 percent as a balm cools. Most cosmetic cholesterol is refined from wool grease, so it shares a feedstock with lanolin and is not vegan; plant sterols are chemically different molecules and do not substitute.

The free fatty acid is the part that gets fudged. Skin barrier fatty acids are overwhelmingly very long chain saturates, C22 behenic and C24 lignoceric being the commonest. What sits on a maker's bench is C16 palmitic or C18 stearic. Stearic acid at 0.2 to 1 percent is a reasonable stand-in for the arithmetic, but it is not the same molecule the barrier uses, and nothing about that gap is fixed by calling the blend physiological.

The 3:1:1 idea comes from work on acetone-disrupted and tape-stripped skin at the University of California, San Francisco in the 1990s. Two findings matter and only one gets quoted. Incomplete mixtures, two of the three lipids without the third, delayed barrier recovery. Complete equimolar mixtures allowed normal recovery, and recovery was accelerated further when any one of the three was raised up to threefold. Not the ceramide specifically: any of them. A cholesterol-dominant or fatty-acid-dominant 3:1:1 performed as well in that model as the ceramide-dominant version the industry adopted.

Worked example. To build an equimolar trio around 0.5 percent Ceramide NP in a 100 g batch: 0.5 g of ceramide is 0.86 mmol, so you need 0.86 mmol of cholesterol (0.33 g) and 0.86 mmol of palmitic acid (0.22 g). That is 0.5 percent, 0.33 percent and 0.22 percent by weight. Scale it for your batch with the batch calculator rather than by hand.

Getting them into a wax base without grit

Pure ceramides are high-melting crystalline powders. Reported melting points for the synthetic phyto ceramides sit around 95 to 110 C depending on chain length and subclass, above any temperature at which you should be holding a balm. Suppliers describe them as dispersible rather than soluble in triglycerides and mineral oil, and the practical consequence is straightforward: heat alone will not give you a clear solution, and whatever you do dissolve at 90 C will try to come back out at 40 C.

Three approaches work in an anhydrous base, in increasing order of reliability. Pre-melt the ceramide into a small quantity of a polar ester or a liquid emollient with a high solvency, hold at 80 to 85 C until clear, then let that premix down into the main phase. Use a supplier blend already dispersed in a carrier. Or accept a dispersion, keep the level low, and cool fast so the crystals stay small. The general behaviour of solutes in a waterless system is covered in solubility in anhydrous formulas, and the effect of cooling rate on crystal size in crystallisation and cooling rate.

Use levels and what they cost per tin

Published cosmetic use of ceramides runs from about 0.01 percent, which is a label-facing trace, to about 1 percent, which is as much as most bases will hold without hazing. Serums and clinical-style emulsions cluster at 0.1 to 0.5 percent. In a balm the practical ceiling is lower, because there is no water phase to help and because a wax network shows crystals up more readily than a cream does.

Pure ceramide sits one to three orders of magnitude above beeswax per gram at small-quantity trade pricing, which is why 0.01 percent exists as a use level at all. Blends are far cheaper per gram but contain only a few percent ceramide, so the cost per gram of actual ceramide often lands in the same place. Put your own supplier figure through the cost calculator before you commit to a level, because this is one of the few balm ingredients that can outrun the packaging in the cost sheet.

Phytosphingosine and pseudoceramides

Phytosphingosine (INCI Phytosphingosine) is the sphingoid base itself, without the fatty acid attached. It is used at roughly 0.02 to 0.2 percent, costs less per gram than a finished ceramide, has documented antimicrobial activity against skin bacteria, and is proposed to act as a precursor the skin can acylate itself. It is also a high-melting solid with the same dispersion problem, and it is more irritating than a ceramide as the level climbs, so the low ceiling is not arbitrary.

Pseudoceramides are synthetic analogues that copy the amide and the two chains without the sphingoid backbone. Names such as hydroxypropyl bispalmitamide MEA or cetyl-PG hydroxyethyl palmitamide identify them on a list. They melt lower, disperse more easily and are used at 0.5 to 3 percent, so they can do something structural in a balm that a 0.05 percent ceramide cannot. What they lack is the clinical record of the real thing: there is supportive trial work, but it is thin, small and mostly from the manufacturers. Treat them as sensory and cost decisions with a plausible mechanism, not as equivalents.

What the barrier repair evidence supports

The strongest evidence is for the general case rather than the specific one. Moisturisers help atopic skin: they reduce flare frequency, lengthen time to flare and cut steroid use, which a Cochrane review of emollients and moisturisers for eczema established across many trials at low to moderate certainty. The same review did not find convincing evidence that any one moisturiser outperforms another. Lipid-containing products were not shown to beat plain ones.

The lipid mixture work is real, specific and narrower than its use in marketing suggests. It was done on experimentally disrupted skin, dosed in a propylene glycol and ethanol vehicle that evaporates and leaves lipid on a stripped surface, and it measured recovery of water loss over hours to days. None of that describes a wax balm applied to intact skin. What can honestly be said: the three lipid classes together behave differently from any one or two of them, and incomplete mixtures can be worse than none. What cannot be said: that adding a trace of Ceramide NP to an oil blend reproduces the result. The consumer-facing version of this argument, including where it becomes a medical claim, is in balm for eczema evidence.

Does a ceramide add anything to a product that is already all lipid?

Ask what limits the product. A balm works by occlusion: it reduces water loss through a film, which is measurable and useful, and which is why it belongs in the occlusive column in occlusive, emollient, humectant. Occlusion is not limited by a shortage of ceramide in the jar. Adding 0.05 percent of anything to a 100 percent lipid system changes the film's water-holding behaviour by an amount nobody has demonstrated.

The argument for adding one is different and narrower: not to seal better, but to deliver a lipid class the surface is missing. That requires the ceramide to partition out of your base into the stratum corneum, and as lipid penetration into skin sets out, a triglyceride and wax vehicle is a poor release medium for a high-melting crystalline solid. If you want to attempt a genuine physiological mixture, do it properly: all three classes, in moles, at 0.5 to 1 percent ceramide, in a base built to release them, which is the reasoning behind the barrier repair face balm. At 0.01 percent in a lip tube, you are buying a line on the ingredient list, and the honest comparison there is with the well-evidenced cheap materials in ingredients for chapped lips.

Labelling it without a claim you cannot support

Ceramides are cosmetic ingredients, not approved actives anywhere. They have no US over-the-counter monograph, so a US product that claims to treat or prevent a skin disease with them is an unapproved drug regardless of the level used. In the EU and UK, claims must satisfy the common criteria in Commission Regulation (EU) No 655/2013: legal compliance, truthfulness, evidential support, honesty, fairness and informed decision making. Evidential support is the one that bites, because it means holding evidence for the claim as made about the product as sold, not evidence about the ingredient in someone else's study.

Safe territory: naming the ingredient, stating the percentage, describing the product as rich in lipids or as a barrier balm in a sensory sense. Unsafe territory: repairs the skin barrier, restores the skin's ceramides, treats eczema, or any implication of a therapeutic effect. Structure and function language sits between the two and moves depending on jurisdiction, which is what cosmetic versus drug claims exists to unpick. On the list itself, an ingredient below 1 percent may be placed in any order after those above 1 percent, so a 0.05 percent ceramide belongs near the end; the INCI list builder will place it correctly.

The decision rule

Add ceramides when three things are true at once: the product is a face or hand balm where a customer will read the list closely, you can afford 0.3 percent or more so the inclusion is not decorative, and you are prepared to add cholesterol and a saturated fatty acid alongside it because the mixture is the point. Otherwise leave them out and spend the money on the base, the packaging or the fill weight.

The uncomfortable part is that no one has shown a wax balm to be a good vehicle for delivering barrier lipids, and there is a body of work suggesting incomplete lipid mixtures can be worse than none at all on damaged skin. That is not a reason for alarm at 0.05 percent, where the dose is too small to do much in either direction. It is a reason to be sceptical of your own product page. A well-made occlusive balm already does the thing balms do well, and if a customer is choosing yours because of the ceramide line, the honest position, and the one argued in do expensive balms work better, is that the base matters more than the trace.

Reviewed and updated 6 September 2026. Spotted an error? Tell us and we will fix and log it.