Follicular openings cover roughly 0.1% of the skin's surface, yet they are not a rounding error. When half the hair follicles in excised skin were physically plugged, permeation of hydrophilic model compounds fell by about 50%, while non-ionised lipophilic compounds lost under 20%. The follicular route therefore carries between almost nothing and about half of total uptake — the molecule decides, not the formulation.
Published 13 October 2026. Last reviewed 13 October 2026.
Definitions, so the rest of this reads without context
- Follicular (transfollicular) route. Passage of an applied substance into the skin through the hair follicle opening and down the follicular canal, bypassing the continuous horny layer at the surface.
- Stratum corneum. The outermost layer of the epidermis, roughly 10–20 µm thick on most body sites, and the principal resistance to anything applied topically.
- Percutaneous penetration. Movement of a substance into the layers of the skin. It is not the same as transdermal delivery, which describes passage through the skin into the systemic circulation and is, in both the European Union and the United States, the territory of medicinal products.
- log Ko/w (log P). The base-ten logarithm of a substance's partition coefficient between octanol and water. Negative values mean the substance prefers water; values above about 3 mean it prefers oil.
- Permeability coefficient (P). The steady-state flux of a substance across a membrane normalised to the concentration driving it, expressed in cm/s. It is the quantity permeation experiments actually measure.
How much skin is follicle?
Mohd and colleagues state that follicular openings occupy about 0.1% of the whole skin area, rising towards 10% on the scalp and the perioral face (Pharmaceutics, 2016). Direct counts on human forearm skin, taken by cyanoacrylate skin stripping, give a density of 37.4 ± 10.0 follicles/cm² (range 17–59, n = 55) in one sample and 40.38 ± 13.29 follicles/cm² (range 17–84, n = 80) in a second; density decreases from proximal to distal, highest on the forehead and lowest on the calf (Scientific Reports, 2017).
Two things follow immediately. First, the between-person spread is wide: a range of 17 to 84 follicles/cm² on the same anatomical site is a factor of five. Second, forearm density predicted density on the thigh (r = 0.808), upper arm (r = 0.746) and calf (r = 0.671), but not on the forehead, neck, abdomen or lower back. Facial follicular density cannot be inferred from a forearm measurement, which is where most instrumental skin research is done.
What happens when the follicles are closed
The cleanest available measurement of the follicular contribution is subtractive: measure permeation through intact skin, then block the follicles and measure again. Mohd and colleagues did this on excised pig ear skin in vertical diffusion cells of 1.77 cm² effective area at 32 °C, counting a mean of 56.1 ± 2.5 follicles in that area and plugging 30 of 60 follicles with a 1:1 mixture of silicone grease and α-cyanoacrylate adhesive, with Nile red as a visual check on the seal.
The resulting reduction in permeability coefficient, using the paper's own compound labels, fell into three clear bands.
| Compound class | Compounds (paper's labels) | Reduction in P |
|---|---|---|
| Hydrophilic markers and ionised drugs | FD-4, Ca-Na, FL-Na, ionised LC, ionised AMP, ISMN | about 50% |
| Intermediate | non-ionised AMP, ionised IP, ISDN | 20–40% |
| Non-ionised lipophilic drugs | non-ionised BP, non-ionised IP, non-ionised LC | under 20% |
The gradient is orderly rather than anecdotal: the reduction falls by roughly 10% per 1.0 log Ko/w unit across the range from 0 to 4, and compounds above 400 Da showed reductions of around 50%. The ionisation effect is visible within a single molecule. For AMP on pig ear skin, P was 8.70 × 10−9 cm/s through non-plugged skin against 4.73 × 10−9 cm/s plugged at pH 3.0, a ratio of 0.54; at pH 7.4 the same comparison gave 1.61 × 10−7 cm/s against 1.43 × 10−7 cm/s, a ratio of 0.89. The same molecule used the follicle when charged and largely ignored it when neutral.
Independent work by the sandwich method, cited in the same paper, put the follicular contribution at 34% to 60% for compounds with log Ko/w from −1.05 to 2.29.
Where the published arithmetic should not be followed
Mohd and colleagues scale their reduction ratios to a whole-skin estimate by doubling them, on the reasoning that half the follicles were plugged. For FD-4, whose tabulated reduction is 59%, that operation returns about 118%, which cannot describe a physical contribution; the paper's discussion separately reports a contribution "over 90%" from aqueous solution. The measured reduction ratios are the defensible numbers; the doubled whole-skin extrapolations are not, and this article does not use them. Stating that plainly costs nothing and is the difference between a citable document and a promotional one.
What a simulation adds, and the caveat it carries
A finite-element model of the two routes reaches a strikingly similar place by a different road. Sebastia-Saez, Lian and Chen report that with a water-filled follicle, the follicular contribution is close to 50% for compounds with log Kow below 1 — and that this holds regardless of molecular weight, protein binding or ionised fraction — falls as lipophilicity rises, and becomes negligible above log Kow 6 (Pharmaceutical Research, 2024). Between log Kow 1 and 6 the two routes are roughly balanced.
With a sebum-filled follicle the ordering inverts: the contribution exceeds 20% in every case simulated, and lipophilic compounds now benefit more than hydrophilic ones. The authors' operational recommendation follows from this — characterise what is actually filling the follicle before running a permeation experiment, because in vitro work behaves as though the follicle were water-filled while topical products are used on skin where it is not.
The caveat is theirs and it is substantial: in their simulations, skin permeability computed with the follicular route closed was the variant closest to experimental stratum corneum permeability data. A model that reproduces the measurements best by switching the follicle off is not a model that licenses confident claims about follicular delivery.
Sebum is not a detail
Whether the follicle is water- or sebum-filled is therefore the hinge of the whole question, and it is site-specific. Sebaceous gland density reaches 400–900 glands/cm² on the scalp and forehead and is considerably lower on the limbs; an ultrastructural study cited in the same work reported sebum content of 1 µg/cm² on the leg against 200 µg/cm² on the forehead (Scientific Reports, 2018). Sampling forearm, chest and forehead in ten adults, the same study found the sebum markers C16:1 and squalene rising in the order arm < chest < head while the epidermal marker C24:0 and cholesterol fell — the follicular environment and the surface lipid composition change together.
| Quantity | Value | Site |
|---|---|---|
| Hair follicle density | 37.4 ± 10.0 /cm² (range 17–59) | Forearm |
| Hair follicle density | Highest measured of nine sites | Forehead |
| Hair follicle density | Lowest measured of nine sites | Calf |
| Sebaceous gland density | 400–900 glands/cm² | Scalp and forehead |
| Surface sebum content | 1 µg/cm² | Leg |
| Surface sebum content | 200 µg/cm² | Forehead |
What this licenses, and what it does not
Three statements are supported by the measurements above. The follicular route is a real pathway and, for small hydrophilic or ionised molecules, a quantitatively major one. Its size is set by the permeant's water–oil preference and by the follicle's contents, not by marketing category. And the figure varies by anatomical site across a range wide enough that a single number for "follicular penetration" is not meaningful.
Three statements are not supported. That a cosmetic formulation can be shown to exploit this route without having measured it on that formulation. That a measured permeation of a model compound through excised pig skin predicts the behaviour of a different molecule in a different vehicle on living human facial skin. And that penetration, however demonstrated, is itself a clinical outcome — the follicular-targeting literature concerns drug products developed and authorised as such.
Formulation at TTS Clinical follows pharmacy-compounding criteria, which in this context means the question "through which route, at what rate, and measured how" is treated as answerable in principle and unanswered until measured, rather than assumed.
Related work from this laboratory
The pH and oxidative-stability constraints that govern one widely used class of percutaneously delivered actives are set out in the open-access technical report Topical Vitamin C: the pH Constraint on L-Ascorbic Acid and the Oxidative Stability of Tetrahexyldecyl Ascorbate (TTS Clinical Scientific Team, 2026; DOI 10.5281/zenodo.22491021, CC BY 4.0).
References
- TTS Clinical Scientific Team (2026). Topical Vitamin C: the pH Constraint on L-Ascorbic Acid and the Oxidative Stability of Tetrahexyldecyl Ascorbate, v1.0. DOI 10.5281/zenodo.22491021 (concept DOI, all versions: 10.5281/zenodo.22491019). CC BY 4.0.
- Mohd F, Todo H, Yoshimoto M, Yusuf E, Sugibayashi K. Contribution of the Hair Follicular Pathway to Total Skin Permeation of Topically Applied and Exposed Chemicals. Pharmaceutics 2016;8(4):32. DOI 10.3390/pharmaceutics8040032. PMID 27854289.
- Sebastia-Saez D, Lian G, Chen T. In Silico Study on the Contribution of the Follicular Route to Dermal Permeability of Small Molecules. Pharmaceutical Research 2024;41:567–576. DOI 10.1007/s11095-024-03660-y. PMID 38351229.
- Jönsson EH, Bendas J, Weidner K, Wessberg J, Olausson H, Backlund Wasling H, Croy I. The relation between human hair follicle density and touch perception. Scientific Reports 2017;7:2499. DOI 10.1038/s41598-017-02308-9. PMID 28566678.
- Ludovici M, Kozul N, Materazzi S, Risoluti R, Picardo M, Camera E. Influence of the sebaceous gland density on the stratum corneum lipidome. Scientific Reports 2018;8:11500. DOI 10.1038/s41598-018-29742-7.
Scope
This article describes findings published in the peer-reviewed literature and the measurement conditions under which they were obtained. It does not constitute a claim about any product, is not a statement of cosmetic efficacy, and is not medical advice. Cosmetic products are not intended to diagnose, treat, cure or prevent any disease.
Research Team, TTS Clinical. 13 October 2026.