Porecelain

Flawless skin starts with the fine print.

Butters

Two of these three butters carry a 4 out of 5 from Fulton's 1989 rabbit-ear table - printed 4*, under a section heading whose asterisk means results depend on the source of the raw material - and the paper does not present either number as a discovery. It presents them as confirmation of something already believed: "Clinically, natural oils such as cocoa butter and coconut butter have long been known to cause problems with pomade acne. This is confirmed in the rabbit ear assay." The clinical impression he is confirming is his own: the pomade-acne study he points at is reference 7 in his own bibliography and he co-wrote it, though nothing we hold establishes that the pomades in it contained either butter. The third butter, shea, has no row in his table at all.

3 rated 1 Clogging 1 Caution 1 Clear 1 never assayed

What they are

A butter is a physical description, not a chemical one. It is a fat that is solid in the jar and melts at or near skin temperature, which is why formulators reach for one — it gives a balm structure and lays down an occlusive film. The word says nothing about what the fat is made of. This family holds three members. Cocoa butter is the fat pressed from the seed of Theobroma cacao. Shea butter is pressed from the kernel of Vitellaria paradoxa, and appears on labels under the older botanical name Butyrospermum Parkii. Coconut butter is the difficult one — as our page on it sets out, the common name covers at least three distinct cosmetic identities, and Fulton never said which he had. The composition does not line up with the ratings, and it is worth seeing that early. Cocoa butter runs 22.9-26.1% palmitic, 34.7-37.3% stearic and 36.2-39.9% oleic by gas chromatography (Padilla and colleagues, 2000). Shea butter runs 2.6-9% palmitic, 25.6-50.2% stearic and 37.1-62.1% oleic (CIR's 2017 final report). Those two are closer to each other than their ratings are — the clearest difference is palmitic acid, about a quarter of cocoa butter and under a tenth of shea. Coconut-derived fat is the outlier, being lauric-rich rather than C18-dominated. So "butter" groups two long-chain fats with a mid-chain one on the basis of how they behave at room temperature. ONE CATEGORY NOTE, because this is a site about materials being confused with one another. Fulton had no butters section. Both of the scored members sit in his Section VI, headed "VI. Oils*", which also contains squalane, mineral oil and hydrogenated polyisobutane — materials we file as HYDROCARBONS, in their own family, and which the table below will not show you. His section and our family are different sets, and where we say "Fulton's oils section" we mean his, not ours. Shea butter is a separate case again: it is not filed elsewhere in his table. It is absent from it.

What the evidence actually says

Here is the sentence this page exists to print. It is in the Results and Discussion of Fulton's 1989 paper, and it is the only sentence in that discussion that names either butter. "Clinically, natural oils such as cocoa butter and coconut butter have long been known to cause problems with pomade acne. This is confirmed in the rabbit ear assay." Read the order of it. The belief comes first and the assay comes second. The 4 is not offered as the thing that established that these butters cause trouble; it is offered as the thing that agreed with a clinical impression already in the literature. Every list that reprints "cocoa butter 4", asterisk stripped, reverses that order, presenting the number as the origin of the concern rather than as its corroboration. So the obvious question is whose impression it was, and the answer is in his own bibliography rather than in anything we need to quote. Reference 7 of the 1989 paper is Plewig, Fulton and Kligman, "Pomade acne", Archives of Dermatology 101, 1970, and the discussion that leads into the butter sentence cites that reference while recounting the 1970 examination he took part in. We have not read that recounting off the scan ourselves, so we do not reproduce it here, and the point does not need it. The reference line is checkable by anybody: he is a co-author of the observation his own assay confirms. We are not calling that misconduct, and it was not hidden — it is printed in the paper, in the discussion and again in the bibliography, with a reference number attached. It was ordinary practice: a clinician who had seen something in a clinic built a bench model and checked. But it changes what a butter 4 is. It is not a measurement that arrived independently at the same conclusion as the clinical literature. It is the same investigator testing his own hypothesis, and reporting agreement. That is one line of evidence presented as two, and it is exactly the collapse this site exists to notice. There is a second thing in that reference. The clinical observation was about HAIR POMADES — finished products rather than ingredients. The textbook account we cite describes pomade acne as a follicle-based eruption following prolonged use of oil-based hair preparations, and it names no butter. Nineteen years later, "natural oils such as cocoa butter and coconut butter" appear as the example, and a rabbit-ear grade appears beside them. We have not obtained the 1970 paper itself and cannot tell you what those pomades contained; the dermatology literature that describes it characterises them as oil-based hair preparations used by Black men, and the secondary accounts we read disagree with each other on how many subjects there were, so we print no subject count at all. THE FAMILY'S EVIDENCE IS NOT EVENLY DISTRIBUTED, and the confirmation sentence covers only two thirds of it. Written out, so the unevenness is visible in one place: COCOA BUTTER is one of the better-evidenced ingredients on this site. Fulton 1989 gives it 4* on a 0-5 scale, the asterisk attaching at the heading of the section his row sits in and meaning that results depend on the source of the raw material. Kligman and Kwong 1979 gave it 3 on a 0-3 scale — the maximum on that scale, in both rabbits and by both methods. Nguyen and Maibach 2007 listed it among the first nine of fourteen materials "deemed positive"; no per-material number is readable from the abstract, which is all of that paper we have read, so we carry the classification and not a score. And it has two human studies, which disagree: Baek 2022 found significantly more microcomedones than untreated skin in six people, and Lee 2015 found cocoa butter's 47% rise indistinguishable from the 54% under a patch containing nothing at all, in eight. COCONUT BUTTER has one reading. Fulton's 4* out of 5, carrying that same section asterisk, on a material he never identified — no supplier, no trade name, no INCI name, no hydrogenation status. And here is the thing this hub has to say out loud, because the reader-facing retraction next door exists for it: FULTON HAS NO COCONUT OIL ROW. Coconut butter is the only coconut row in all 220, and the 4 the lists print against coconut oil is this asterisked row wearing a different name. Within the legacy corpus we hold, no other assay scored coconut butter; Morris and Kwan's 1983 survey, the one independent contemporaneous assay on the same 0-5 scale, has no coconut butter row and no butter row of any kind. SHEA BUTTER does not appear in the assay corpus we searched, in either direction. Not in Fulton, not in the eleven studies acne.org catalogues, and not in the 2017 CIR final report on thirteen shea-derived ingredients — the document that would carry a comedogenicity result if one existed. AND ONE ABSENCE COVERS ALL THREE OF THEM. Fulton's other table, the 1984 one we have read in full from a 400dpi render of a print copy, contains no cocoa butter row, no coconut butter row and no shea butter row. A second reading from the same investigator could have existed for any of the three. None does. Print Fulton's key beside those numbers, because almost nobody does. He wrote that "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." Neither butter is borderline. Both of those asterisked 4s sit in his positive band, and nothing on this page moves them out of it - though the asterisk is his own statement that where the raw material came from can move the result.

What the evidence doesn’t tell you

What the confirmation reading does not do is make the 4s wrong. A hypothesis a researcher already believed can be a true hypothesis, and Fulton's butters went through the same rabbit-ear procedure as his other 200-odd rows: roughly 10% in propylene glycol, applied to the inner ear of New Zealand albino rabbits, five days a week for two weeks, follicles scored for keratosis under a micrometer. There is nothing in the paper suggesting the butter ears were read differently from any other ear. What the provenance changes is how much weight "confirmed" should carry, not the grade. It also tells us nothing about whether the butters were in the pomades. We have not read Plewig, Fulton and Kligman 1970, and we have not read the 1989 paper's recounting of it off the scan either. What we can read is the sentence that does name the butters, and it says only that they "have long been known" to cause the problem — a claim about received clinical opinion, with no citation attached to it beyond the pomade study itself. Whether the 1970 subjects' pomades actually contained cocoa or coconut butter is a question we cannot answer from the documents we have. Both 4s carry a qualifier the lists strip, and it is not ours — it is Fulton's. The rows sit under his section heading "VI. Oils*", and he defines that asterisk once, at the foot of the table, as "Results depend on source of raw material." Our oil hub sets out what that asterisk does across his whole section; it applies to these two rows as much as to the sesame and the corn oil. A 4 the author says depends on which batch arrived is not a fixed property of an ingredient. Do not read the family as a unit. Sharing the word "butter" on a label tells you about texture. Cocoa and shea are compositionally close and rated differently, and the difference that decides them is evidentiary rather than chemical: one has three animal readings and two conflicting human studies, and the other has nothing we can find. Shea's Clear is a default built on an absence, not a test it passed — and the compositional argument that defends it cuts the other way too, since oleic acid is up to 62.1% of shea butter, and oleic acid is one of the fatty acids Kanaar 1971 reported inducing marked inflammation and follicular keratosis in the rabbit ear canal - a qualitative finding, on a different site, with no scale attached to it. On human evidence, the bounded statement is this: we identified no comedogenicity study that applied any of these three butters to facial skin. What exists for cocoa butter is two occluded upper-back patch studies, six and eight participants, pointing opposite ways; Fulton himself argued that back skin is relatively insensitive compared with facial skin. We are stating that carefully because the hub above this one got the equivalent sentence wrong: the oil family page asserted a family-wide absence of human evidence that our own coconut oil page disproved two clicks away. A family-level generalisation is precisely the kind of claim this site exists to distrust, and being the ones making it is not a reason to relax. AND ONE THING A HUB CAN SEE THAT NO MEMBER PAGE CAN. Cocoa butter is rated Caution on this site and coconut butter is rated Clogging, and the number under both is the same asterisked 4, from the same section of the same table, read by the same investigator in the same programme. On the evidence set out above the ranking is inverted: cocoa butter is the member with a second animal reading at the top of its scale, a third animal classification and two human studies, while coconut butter has one reading of a material Fulton never identified. Whatever justifies the split between them, it is not the rabbit-ear number, and a reader comparing the two rows is entitled to know that. We record it here rather than quietly reconcile it, because that decision lives in the ratings data and a hub is not the place to overturn it. Finally, the scales. Fulton's 4* is on 0-5. Kligman and Kwong's 3 is on 0-3, where it is the top of the scale, not the middle of one — written as a bare number beside Fulton's 4* it would read as the milder result, and it is the more severe one. Nguyen and Maibach's scale is 0-4 and we print no number from it at all, because none is readable in the abstract we read. Three assays, three scales, and one of them we can only report as a classification.

All 3 butters 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. "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 comedogenic number in circulation, and the source of everything load-bearing on this page. Rabbit-ear assay, roughly 10% in propylene glycol, inner ear, five days a week for two weeks, follicular keratosis read under a micrometer; scale 0-5 with the author's own key, 0-1 "not considered significant", 2-3 "borderline", 4-5 "considered positive". Table I, Section VI ("Oils*"), scores Cocoa butter 4* and Coconut butter 4*, the asterisk attaching at the section heading and defined as "Results depend on source of raw material". It has no coconut oil row. Two things on this page are taken from the BODY of the paper rather than the table, and no consumer list reprints either: the sentence beginning "Clinically, natural oils such as cocoa butter and coconut butter"; and the bibliography, in which the 1970 pomade study is reference 7. That discussion also recounts the 1970 study in the first person, a passage we have not read off the scan and therefore do not quote. Fulton also describes the survey as "not at all definitive but simply designed to stimulate research."
  2. Plewig G, Fulton JE, Kligman AM. "Pomade acne." Arch Dermatol 101:580-584 (1970) The clinical observation Fulton's butter scores are reported as confirming, and reference 7 of his 1989 bibliography. WE HAVE NOT OBTAINED THE FULL TEXT and are not going to pretend otherwise: what we hold is the reference line in Fulton's 1989 bibliography, his citation of the study in that paper's discussion, and secondary descriptions in the dermatology literature. We carry no figure from it - no subject count, no proportion - because the 1989 recounting that supplies those numbers is itself a passage we have not read off the scan. Cited for one point only, which does not depend on the full text — that Fulton co-authored the observation his own assay later confirmed. It is a clinical survey of men using finished hair products, not an assay of any ingredient, and it can support no score for anything.
  3. Onwudiwe O, Callender VD. "Pomade Acne", in Acneiform Eruptions in Dermatology, Springer (2014) A textbook chapter, cited only to characterise what pomade acne is and what the 1970 paper is taken to have shown — a follicle-based eruption after prolonged use of agents containing high-melting-point hydrocarbons, described in the literature as affecting Black men using oil-based hair preparations. It is a secondary account, not a study, and it names no butter. Cited because we could not read the 1970 primary source and would rather say where our characterisation of it came from.
  4. Morris WE, Kwan SC. "Use of the rabbit ear model in evaluating the comedogenic potential of cosmetic ingredients." J Soc Cosmet Chem 34:215-225 (1983) An independent rabbit-ear assay on the same 0-5 scale, six years before Fulton's table, and not a republisher of him — its authors contradict him in print on sodium lauryl sulfate. Cited here entirely for what its row list does NOT contain: reading it in full, there is no cocoa butter row, no coconut butter row and no shea butter row. The one contemporaneous survey that could have given this family a second opinion tested none of it. That establishes an absence of corroboration, not a competing number.
  5. Kligman AM, Kwong T. "An improved rabbit ear model for assessing comedogenic substances." Br J Dermatol 100:699-702 (1979) A separate rabbit-ear paper on a DIFFERENT SCALE, 0-3, comparing a histological against a whole-mount method across sixteen materials, two rabbits each. Cocoa butter scored 3 in all four readings — both rabbits, both methods — which is the maximum of that scale and not a middling result. It is genuine non-Fulton corroboration for one member of this family, and the only such corroboration that carries a number; the other, Nguyen and Maibach 2007, is a classification with no readable score. Note also that isopropyl myristate, the reference comedogen the whole literature is calibrated against, scored 1 out of 3 in these authors' hands.
  6. Kanaar P. "Follicular-keratogenic properties of fatty acids in the external ear canal of the rabbit." Dermatologica 142(1):14-22 (1971). PMID 5548538 Cited on this page for one sentence in the limits and nothing else. ABSTRACT READ VERBATIM, full text not read. It is the only reading in our corpus that bears on oleic acid, which is the largest fatty acid in shea butter, and it reports that oleic acid induced marked inflammation and follicular keratosis. Two cautions travel with it and both matter: the site is the rabbit's EXTERNAL EAR CANAL rather than the inner ear surface Fulton used, and the paper grades in words, not numbers, so no score of any kind can be taken from it. It says nothing about shea butter, or about any butter.
  7. Nguyen SH, Dang TP, Maibach HI. "Comedogenicity in rabbit: some cosmetic ingredients/vehicles." Cutan Ocul Toxicol 26:287-292 (2007) A later rabbit-ear paper on a 0-4 scale. ONLY THE ABSTRACT HAS BEEN READ, and it prints no per-material number, so nothing numeric from it appears on this page. What it does state is that of fourteen listed materials "the first nine were deemed positive", and cocoa butter is ninth on that list. Cited as a third animal reading for one member — enough to stop us saying that this family's numbers all come from one laboratory, which is what we would have written without it.
  8. Baek JH et al. "Early detection of microcomedones induced by cocoa butter using reflectance confocal microscopy." J Cosmet Dermatol 21:3016-3021 (2022) A dedicated human comedogenicity study, and one half of the family's contradictory human record. Six participants, patches on the upper back, reflectance confocal microscopy; the ratio of microcomedones to follicles rose significantly above the untreated site by two weeks, and the microcomedones were significantly larger at two and four weeks. The accessible record does not disclose concentration, amount, grade, supplier or vehicle. Its own stated conclusion is chiefly that confocal microscopy is a good way to detect microcomedones. Not a face, and not a butter other than cocoa.
  9. Lee E, Kim N, Jung A, Ha J. "Isopropyl Myristate and Cocoa Butter are not Appropriate Positive Controls for Comedogenicity Assay in Asian Subjects." J Cosmo Trichol 2(1):1000102 (2015) The other half, and the study a page arguing for the 4s would prefer not to cite. Eight Asian women with acne-prone upper-back skin, cocoa butter under an occlusive patch three times weekly for four weeks on the Mills and Kligman human model. Microcomedone activity rose 47% at the cocoa butter site, 55% under petrolatum and 54% under a patch carrying no test material at all, with no significant differences between any of them. Read honestly it shows a model that discriminated nothing rather than a butter that did nothing — eight people, one population, and the authors themselves call for larger studies. Cited because the family's human evidence does not agree with itself and a hub that showed only one side would be choosing its evidence.
  10. Cosmetic Ingredient Review. "Safety Assessment of Butyrospermum parkii (Shea)-Derived Ingredients as Used in Cosmetics" — FINAL REPORT, released 5 October 2017 The industry safety dossier on thirteen shea-derived ingredients, and the best composition data available for this family's third member — palmitic 2.6-9%, stearic 25.6-50.2%, oleic 37.1-62.1%, plus a 2-12% unsaponifiable fraction. Cited chiefly for a negative, and it is the strongest one available: the Expert Panel reviewed the full toxicological record and the report contains no comedogenicity data. It is a safety assessment, not a comedogenicity study, and the version cited is the final report rather than the tentative one we linked by mistake in an earlier draft of the shea butter page.
  11. Padilla FC, Liendo R, Quintana A. "Characterization of cocoa butter extracted from hybrid cultivars of Theobroma cacao L." Arch Latinoam Nutr 50(2):200-205 (2000) Gas-chromatography fatty-acid profile of cocoa butter — palmitic 22.9-26.1%, stearic 34.7-37.3%, oleic 36.2-39.9%. A food-science compositional analysis, with no comedogenicity endpoint of any kind. Cited only so the composition comparison across the family is drawn from a measurement rather than from a supplier's brochure, and for the inconvenient half of that comparison: cocoa and shea butter are compositionally closer than their ratings are.

‹ All ingredients