Porecelain

Flawless skin starts with the fine print.

Fatty acids

The comedogenic numbers circulating for lauric, myristic, palmitic, stearic and capric acid come from a paper that built a second table for these acids alone and read nine of them twice over: once in a volatile organic solvent, once in sunflower oil. Four of the five change band on Fulton's own key between the two readings. That comparison is his own, reported without a concentration, a rabbit count or a statistic, and it is not a controlled vehicle study — and the number the charts print does not come from it at all. That one is from his routine run, in propylene glycol.

5 rated 0 Clogging 5 Caution 0 Clear 0 never assayed

What they are

Five free saturated fatty acids, distinguished from one another by the length of the carbon chain — capric acid at ten carbons, lauric at twelve, myristic at fourteen, palmitic at sixteen, stearic at eighteen. Chemically the grouping is not arbitrary. They are the same molecule with the same acid group on the end and a different number of carbons behind it, which is exactly the variable the comedogenicity literature reached for when it tested them. Three identity points, because on this family the label and the tested material come apart in three separate ways, and every number below belongs to the free acid. FIRST, a free fatty acid is not the same substance as the same fatty acid bound into a fat. Coconut oil is roughly half laurate by fatty-acid profile, but in coconut oil that laurate is esterified to glycerol inside a triglyceride. Free lauric acid is a different chemical species and it takes hydrolysis to get from one to the other. Fulton's table has no coconut oil row in any case — he tested coconut butter — but even if it did, an oil's reading would not transfer to the acid inside it, and the acid's reading does not transfer to the oil. SECOND, what a label declares is often not what the finished product carries. Where a formula also contains an alkali or an amine, the acid named on the list is neutralised in the pot into a soap — sodium, potassium or triethanolamine laurate, myristate, palmitate or stearate. Those soaps are their own materials. Fulton scored several of them separately, and the numbers are not the same; see below. How often it happens depends on the formula, and we hold no figures on it. THIRD, commercial "stearic acid" is usually not the pure C18 compound but a grade-dependent mixture of stearic and palmitic acid. Fulton's stearic acid row is the only row in this family carrying his asterisk, defined in the paper as "results depend on source of raw material", which is what a variable blend would produce. And the point the oil hub had to learn in public, which applies here too. FULTON'S TABLE AND OUR FAMILY ARE DIFFERENT SETS. In his Table I these acids run together with their derivatives, in the order the printed table gives them, and the twenty-seven rows that follow them — from Ascorbyl palmitate to Tridectyl neopentanoate — are esters, which we file under ESTER and ANTIOXIDANT and which the member table below will not show you. Three free acids run in the same stretch that we have never rated at all: caprylic (C8), eicosanoic (C20) and behenic (C22). They are not members of this family either, and we name them below only as rows in his table.

What the evidence actually says

Here is the fact this page exists for, and it is printed, not inferred. Fulton's 1989 survey is where nearly every fatty-acid comedogenic number in circulation comes from. It contains three tables. Table I is the famous one, the 220-row list the charts copy. Table III is headed, in full — "Effects of the Solvent on Comedogenicity and/or Irritancy of Fatty Acids". It is a table Fulton built for this family and no other, and he built it to read the same acids under two conditions. Table III has nine rows and two data columns, and both columns are comedogenicity: once dissolved in an organic solvent — the table's footnote gives that as "Ethyl ether or acetone" — and once in sunflower oil. On the scale of 0 to 5, for the five acids this family holds, the readings are: capric acid 2 of 5 in solvent and 3 of 5 in oil; lauric acid 3 and 4; myristic acid 1 and 3; palmitic acid 0 and 2; stearic acid 0 and 2. Set those beside Fulton's own published key, which almost no chart reproduces. "A minimal grade of 0 to 1 is not considered significant. Grade 2 to 3 is borderline. However, a grade of 4 to 5 is uniformally reproduceable and considered positive." Now read the table again. Myristic, palmitic and stearic acid move from his insignificant band into his borderline band. LAURIC ACID — the one every list names as the comedogenic fatty acid — moves from borderline to positive. Four of the five members of this family cross a band on Fulton's own interpretive scale between the two readings. Only capric acid stays put, and it moves within borderline, 2 to 3. The bound belongs in the same breath as the finding: this is one investigator's own two-condition comparison, reported with no concentration, no rabbit count, no duration and no statistics, and it is not a controlled vehicle study. Table I is not a third arm of that comparison. It is the separate, routine run — Fulton's stated method is roughly 10% of the ingredient in propylene glycol — and it is the run every chart copies. So the number a reader meets is the propylene-glycol number, and the two-condition comparison Fulton printed for these acids and no others sits in a different table that no chart carries. One thing worth printing, because it is easy to assume otherwise: none of the liquids involved scores anything on its own. Table I gives propylene glycol 0 of 5, acetone 0 of 5, ethyl ether 0 of 5, and sunflower oil 0 of 5. THE SECOND FAMILY-LEVEL FINDING IS ABOUT THE SOAPS, and it is the one that bears on reading a label. What the salt does is a different question from what the acid does, and Fulton tested it. Table I scores lithium stearate 1 of 5, magnesium stearate 1 of 5 and zinc stearate 0 of 5, against stearic acid at 2-3* of 5 — and he states the reading plainly: "When metallic bases, such as lithium, magnesium, and zinc stearate, are added to the fatty acids, the metal appears to prevent the comedogenic reaction." The organic salt goes the other way. He prints a separate row, "Stearic acid: TEA", at 3 of 5 — which is the top of stearic acid's own printed range rather than a grade above it — and reports a small experiment behind it: "In testing different ratios [4:1, 1:1, 1:4] of stearic acid to triethanolamine (stearic acid:TEA) in a cold cream base, all combinations were found to be comedogenic." That last one is a finished cold cream base rather than the isolated pair, so it is evidence about a formula, not a clean reading of a salt. None of these four materials is a member of this family and we hold no rating for any of them; they are named here because they are what several of these acids become, and because the counter-ion changes the answer in both directions. ONE OTHER LABORATORY READ THE SAME SERIES, AND IT DISAGREES WITH HIM ON THREE OF OUR FIVE. Kanaar, in Dermatologica in 1971, applied free fatty acids to the external ear canal of the rabbit — eighteen years before Fulton, a different site, a different laboratory, no numeric scale at all. The abstract, which we have read verbatim and hold in our corpus, reports: "Laurinic acid and oleic acid induced marked, and caprinic acid moderate inflammation and follicular keratosis. Myristinic, palmitinic and stearinic acid induced no distinct alterations." Those older Latinate names are lauric, capric, myristic, palmitic and stearic acid. So the two series agree at C10 and C12 and part company above it — where Fulton printed 3, 2 and 2-3* for myristic, palmitic and stearic acid, Kanaar found nothing distinct from any of the three. Two rabbit assays, two answers, on no common scale. This is the family's whole evidential position and it is why every member sits at Caution. THE CHAIN-LENGTH SHAPE, reported as what it is. Fulton opens his own paper with it: the first of four "generalizations" he says "can be deduced by examining the results" is "medium-chain-length fatty acids are more potent than short- or long-chain fatty acids in producing follicular keratosis." That sentence is his, in his Synopsis, and it is a summary of the table you have just read rather than a separate finding. His Table I run does show the shape — caprylic 1, capric 2, lauric 4, myristic 3, palmitic 2, stearic 2-3*, eicosanoic 2, behenic 0, all on 0 to 5 — and it survives into both columns of Table III, peaking at lauric acid in each. What it does not survive cleanly is the short end. Caproic acid, six carbons, appears only in Table III, and in sunflower oil it scores 2 of 5, level with palmitic and stearic acid. On the evidence in front of us this is a consistent direction across one investigator's readings of one series, with an exception at the short end in the carrier that raised everything. It is not a law, we are not going to promote it into one, and the mechanism Fulton offers for it — a molecular-weight and HLB argument about a penetration window — is an empirical hypothesis about his own ingredient set that we do not adopt anywhere on this site. There is human data on this exact series, and it is worth knowing precisely what it measured. Kellum, in Archives of Dermatology in 1968, applied the even-numbered free fatty acids from C2 to C16 to human skin under occlusive patch, and reported greater irritancy or penetration or both in the C8 to C14 range, most of all at C12. That is human skin, this family, and the same peak at lauric acid twenty-one years before Fulton — but the endpoint is irritation and barrier penetration, not comedones, and it covers four of our five members rather than all of them, since stearic acid at C18 is outside its range. Irritation is not follicular plugging. We read it at abstract level through the PubMed record and did not obtain the paper. It is the strongest thing in this family that is not a rabbit, and it does not answer the question this site asks.

What the evidence doesn’t tell you

What the Table III comparison establishes is narrow, and it cuts both ways. It shows that these numbers moved when the liquid changed, in one laboratory, in one paper. It does not tell you which of the readings a finished product resembles, and it does not license reading the lower one. A moisturiser is not ethyl ether. If anything the reading that ought to worry a reader is the one taken in sunflower oil, since a nonvolatile lipid is closer to what most leave-on products are than a solvent that flashes off — and that is the column where four of the five acids score higher. We are not going to hand anyone a reassurance the table does not contain. Table III is also thinly documented. Fulton reports no concentration for it, no rabbit count, no duration and no statistics; it is nine rows and two data columns with a one-line footnote. Where the footnote says "Ethyl ether or acetone" it does not say which acid got which. The comparison is internally consistent and it is the author's own, and it has not been replicated by anyone we can find. Nor is there a second Fulton reading to set against it. We hold his earlier 1984 table in full from a print copy, and it contains no free fatty acid row at all — no capric, no lauric, no myristic, no palmitic, no stearic acid — so the 1984 paper supplies no second reading of any member of this family, and the conflict it creates elsewhere in the corpus does not reach here. The paper contradicts itself once, on a row outside this family, and it is fair to hold that against the rest. Behenic acid is printed 0 of 5 in Table I and 1 of 5 in both columns of Table III. Fulton offers no explanation and we will not invent one. The same paper also prints the C20 acid under two different names, "Eicosanoic acid" in Table I and "Archidic acid" in Table III, at 2 of 5 and at 1 and 2 of 5. Neither acid is a member of this family, but a table that disagrees with itself about one of its own rows is a table to quote carefully. All of it is still rabbit ears, and Fulton bounded his own model in the paper: he wrote that "not everything that irritates this model will also irritate human skin", and he called the survey "not at all definitive but simply designed to stimulate research". Draelos and DiNardo argued in 2006 that isolated-ingredient rabbit scores say little about what a finished product does on a person; a 2025 review in JAAD Reviews found that no standardised comedogenicity test exists and that "non-comedogenic" is an unregulated claim. Those objections apply to the low readings in this family exactly as hard as to the high ones. On human evidence, the bounded statement and its exceptions in the same breath. We identified no study that applied any one of these five acids alone, at a stated concentration, to human skin and counted comedones. Two things exist that are close to it and neither closes the gap. Kellum 1968 is human skin and this series, but it measured irritation and penetration rather than comedones, and we read it in abstract. And lauric acid has been on human faces: Kozan, Guner and Akyol reported in Dermatologic Therapy in 2020 on ninety acne patients, thirty of whom used a twice-daily leave-on combining 4% niacinamide, 1% gallic acid and 1% lauric acid for eight weeks. Facial acne was graded. It settles nothing about lauric acid — three actives, no vehicle arm, and a paper whose title says retrospective while its methods describe patients randomly divided — but it exists, it is cited at length on that member's page, and a family page that said otherwise would be repeating the failure this site was built to catch. We have also identified, but not read, a second human chain-length irritancy series, Stillman and colleagues in Contact Dermatitis in 1975; we name it so the gap in our reading is visible rather than quiet. Kanaar we hold at abstract level only. We read the abstract verbatim through EuropePMC and did not obtain the paper, so its concentration, vehicle, animal count and any quantitative comparison do not reach us. His result is a word, not a score, and it cannot be converted into one or set against Fulton's numbers on a shared scale. Two laboratories disagreeing is a different position from one laboratory being unconfirmed, and neither can check the other. Finally, the honest caution about this page itself. A family-level generalisation is exactly the kind of claim this site exists to distrust, and that does not stop applying because we are the ones making it. What is established here is that ONE INVESTIGATOR'S readings of ONE SERIES moved with the liquid the acid was carried in, and that he printed both readings. It is not established that vehicle dependence is a general property of comedogenicity testing, that any particular product is safe or unsafe, or that the same thing would happen on a human face. The five ratings under this hub rest on single 1989 rabbit readings — four of them in Fulton's borderline band, and lauric acid's 4 in his positive band — contradicted in part by a 1971 rabbit assay, and that is the whole of it.

All 5 fatty acids we rate

The legacy score is printed exactly as it appears in the paper, including the ranges, the asterisks and the parentheses. Those marks are not decoration. They are the paper telling you what its own number is worth.

Sources

  1. Fulton JE Jr. "Comedogenicity and irritancy of commonly used ingredients in skin care products." J Soc Cosmet Chem 40:321-333 (1989) The primary source for essentially every fatty-acid comedogenic number in circulation. Design — ingredient mixed in propylene glycol at a 9 to 1 dilution unless otherwise indicated, 1 ml applied once daily to the inner surface of one ear of New Zealand albino rabbits, three rabbits per assay, five days a week for two weeks, follicular keratosis measured with a micrometer and converted to a 0 to 5 grade. Read here off the complete journal scan rather than through a chart quoting it, which is how the two items this page turns on became visible: Table III, "Effects of the Solvent on Comedogenicity and/or Irritancy of Fatty Acids", which reads nine acids under two conditions and prints two data columns, both of them comedogenicity; and the grade key, which converts those readings into bands and which the charts drop. What it cannot support — one laboratory, one animal model, three rabbits per assay, no replication, and the author's own verdict that the survey is "not at all definitive but simply designed to stimulate research".
  2. Kanaar P. "Follicular-keratogenic properties of fatty acids in the external ear canal of the rabbit." Dermatologica 142(1):14-22 (1971). PMID 5548538 The only reading of this series we hold that is not Fulton's, and the reason this family is not a single-laboratory rating. A rabbit experiment eighteen years earlier, in the external ear CANAL rather than on the inner ear surface, with NO NUMERIC SCALE — it grades in words. Cited at abstract level and labelled as such: we read the abstract verbatim through EuropePMC and did not obtain the paper, so concentration, vehicle and animal count do not reach us. It agrees with Fulton on lauric and capric acid and points the other way on myristic, palmitic and stearic acid. It cannot be placed on Fulton's scale and we do not give it a number.
  3. Kellum RE. "Acne vulgaris. Studies in pathogenesis: relative irritancy of free fatty acids from C2 to C16." Arch Dermatol 97(6):722-726 (1968). PMID 4231576 Human skin, this exact series, and the wrong endpoint — which is precisely why it is here. Even numbered free fatty acids from C2 to C16 applied under repeated occlusive patch to human skin; it reports that the acids penetrated the epidermal barrier and produced inflammation, with the greatest irritancy or penetration in the C8 to C14 range and most of all at C12. That is the same peak Fulton found in rabbits, twenty-one years earlier, in people. It counted no comedones, measured no follicular keratosis, and stops at C16, so it covers four of this family's five members and cannot corroborate a comedogenic score for any of them. Read at abstract level through the PubMed record; we did not obtain the paper.
  4. "Final Report on the Safety Assessment of Oleic Acid, Lauric Acid, Palmitic Acid, Myristic Acid, and Stearic Acid." J Am Coll Toxicol 6(3):321-401 (1987) An industry-funded, independently panelled toxicological safety review covering four of this family's five members, and NOT a comedogenicity study. What we read is the report as published in J Am Coll Toxicol in 1987; we hold no record of the status line printed on the document itself, so we do not describe it as settled, tentative or superseded. A later CIR re-review of the fatty acids and their salts is often referred to; we have not established its date or its status, so we do not cite it here. Useful for what these materials are in commerce — that "stearic acid" of trade is a grade-dependent mixture rather than the pure compound, and that these acids function as cleansing surfactants and soap precursors more often than as emollients. Its conclusion is that they are safe as used. Safety is not noncomedogenicity, and the report counts no comedones.
  5. Draelos ZD, DiNardo JC. "A re-evaluation of the comedogenicity concept." J Am Acad Dermatol 54(3):507-512 (2006) A "Current Issues and Opinion" piece that tested FINISHED PRODUCTS on small human upper-back panels, not a same-substance rabbit-versus-human validation of anything in this family. Cited because it cuts against the low readings in Table III exactly as hard as against the 4 in Table I, and a page that used it in one direction only would be choosing its evidence.
  6. "Comedogenicity in cosmeceuticals: a review of clinical relevance, regulatory gaps, and future directions." JAAD Reviews (2025) A recent peer-reviewed narrative review of the field, finding that no standardised comedogenicity test exists and that "non-comedogenic" is an unregulated claim. A synthesis of the legacy literature rather than a new measurement, and independent of Fulton only in authorship, since the corpus it surveys includes him.
  7. CosDNA, acne.org, Face Reality, Banish and Sofie Pavitt comedogenic ingredient lists Where a reader will already have met the numbers 4, 3 and 2 for these acids. They are consumer lists rather than measurements, and none of the ones we looked at printed Fulton's grade key, his Table III, or a second reading of any of these acids. We checked what they print; we did not trace who copied whom from whom, and we are not going to assert a citation chain we did not follow. Acne.org states a multi-study rule for its list and cites primary papers, so describing it as a straight republication of one assay would be inaccurate.

‹ All ingredients