Same Senior Author, Nineteen Years Apart, Opposite Verdicts

Transepidermal water loss is how much water evaporates through the skin per hour. It is the standard proxy for how well the barrier works, it is easy to measure, and the claim that it differs by ethnic group has been in circulation for two decades.

It shapes how moisturizers are formulated and marketed. And it has now been imported into an entirely different literature — a 2025 review of pressure injury risk lists increased water loss in dark skin tones as a mechanism, in the flat present tense.

The dermatology review dedicated to that question concluded, three years earlier, that no conclusion can be made.

The two reviews

The 2003 review is where the claim comes from. Its key sentence: the majority of the evidence, six out of eight studies, indicates that transepidermal water loss is greater in Black skin compared with White skin. Its summary goes further — substantial evidence to support that Black skin has higher water loss, variable blood vessel reactivity, decreased surface pH and larger mast cell granules.

Six out of eight is a vote count. It ignores how many people were in each study, how big the effects were, how precise they were, and how well the studies were done. Eight studies, and the largest has a sample size in the tens.

The same review opens by saying that racial differences in skin have been minimally investigated by objective methods and the data are often contradictory, and states that findings on water content, lipid content and Asian skin are all inconclusive. Those sentences travel much less widely than the six-out-of-eight one.

In 2022, a review devoted specifically to whether skin of color has an enhanced barrier searched the literature again. Its conclusion, in full: the data show wide contradiction in results for all skin of color groups and white skin and, therefore, no conclusion can be made based on this question. The authors suggest the variation may be down to confounding variables such as anatomic site and sample size.

Henry Maibach is a senior author on both.

That is not an inconsistency to hold against him — it is a scientist updating in public, which is what should happen. What is worth holding against the field is that the 2003 sentence is still the one in circulation.

What the eight studies actually say

Sorted by direction, which is the only honest way to present them.

The internal contradictions are just as sharp on the other parameters. Surface pH: the 2003 review says Black skin is lower; a 2019 study of 50 South African nursing students found it significantly higher, at p below 0.0001; a 1998 study found no baseline difference but lower after tape stripping; a 2015 study found the three groups similar. Four sources, four positions.

Hydration: that 2019 study found stratum corneum hydration significantly lower in African participants. A 2015 study in the same country concluded that the two more darkly pigmented ethnic groups had superior skin hydration properties. Opposite directions, both in South Africa.

Four women per group

The facial ranking that circulates in cosmetic science — Caucasian barrier best, then Black African, then Chinese, then Indian — comes from a study that measured water loss and capacitance at 30 predefined sites on the forehead, cheek, jaw and eye area.

In sixteen women. Four per ethnic group.

Its own conclusion is more interesting than its ranking, and it is the part that gets dropped: in contrast to other reports, darkly pigmented skin does not always have a superior barrier function, and hydration differences are complex across different parts of the face.

The same group’s 2019 methodological review puts a bomb under the entire literature. Remarkable gradients of hydration, water loss, pH and sebum exist within short distances across a single face, and those gradients differ between ethnic groups. There are no correlations between capacitance, water loss, surface pH and individual typology angle. And most studies use single-point measurements on surrogate sites such as the volar forearm — which, they conclude, is inappropriate for facial skin.

Nearly every study in the table above measured the volar forearm. If the within-face variation is steep and the between-group difference is small, then the site you choose determines your answer. Which is exactly what happened in the 1996 study that found a difference on cheeks and legs and none on the forearm.

The result that does replicate, and what it is a result about

There is a clean finding in this literature, and it is in a paper whose title states it: skin type, but neither race nor gender, influences epidermal permeability barrier function.

The method was tape stripping — counting how many strips it takes to break the barrier, defined as water loss reaching 20 grams per square meter per hour.

Skin types II and III required 29.6 strips, plus or minus 2.4. Skin types V and VI required 66.7, plus or minus 6.9.

More than twice as many. And in the same fourteen people, comparing white against Asian participants, and men against women, there was no significant difference in either the number of strips needed or the rate of barrier recovery.

So the barrier finding that survives is a phototype finding, not a race finding, in a study that tested both and said which one worked. That distinction is the whole argument of this section, arriving with a number attached.

Note the sample: seven people in the darker group. A 2.25-fold effect in seven people is a starting point, not a settled quantity, and it deserved a replication it has not had in thirty years.

What a real meta-analysis looks like

There is no systematic review of ethnic differences in skin physiology as a whole. There is one true systematic review and meta-analysis of a single parameter, and it happens to answer the question cleanly.

Epidermal thickness: 142 studies included, 133 pooled, covering 37 skin areas, searched back to 1946.

The thinnest site pooled at 31.2 micrometers. The thickest, the sole of the foot, at 596.6.

And the verdict on the question this whole literature is about: differences between sexes and among ethnic origins were minor.

Anatomical site varies epidermal thickness nineteen-fold. Ethnicity moves it by a rounding error. The paper also found that measurement instrument matters — histology, optical coherence tomography and laser scanning microscopy agree, and high-frequency ultrasound systematically reads high and should not be used.

Site and instrument dwarf ancestry. That is the shape of this entire subject, stated once, properly, with 142 studies behind it. I went through the specific thickness claim in the piece on 1.39 against 1.41 millimeters.

Melanosomes, where the microscopy agrees with itself

This is the one area where the electron microscopy agrees with itself.

In dark skin, melanosomes inside keratinocytes are predominantly distributed individually — 88.9 percent. In light skin they are predominantly clustered in membrane-bound groups — 84.5 percent. Asian skin sits between, at 62.6 percent individual against 37.4 clustered.

Size follows the same gradient: dark skin largest, then Asian, then Caucasian.

Even here the numbers deserve reading properly. The mean sizes are 1.44 against 0.94 in the units reported, a difference of 0.50 — with standard deviations of 0.67 and 0.48. The spread within each group exceeds the gap between them, which is the recurring shape of everything in this field.

The melanin-content work adds a comparison nobody makes. The most lightly pigmented skin types have approximately half as much epidermal melanin as the most darkly pigmented. And within any individual, regardless of ethnicity, chronically sun-exposed skin has up to twice as much melanin as photoprotected skin.

The effect of sun exposure inside one person is the same order as the difference between the lightest and darkest groups. Any study that does not control site and sun exposure is measuring both at once.

That same paper contains a finding that contradicts a very common claim: the proportion of spheroidal melanosomes was under 10 percent in every skin type examined, including European. Pheomelanin is a minor component of epidermal melanin even in the lightest skin.

The people who built this literature, on this literature

I want to end with what the authors say themselves, because it is more damning than anything I would write.

From the standard 2006 review of ethnic skin types: most of these studies have been of small scale, and in some studies interindividual differences in skin quality overwhelm any racial differences.

From a nine-author expert panel in 2021, including the most senior figures in US skin-of-color dermatology: ethnic categories based on race and ethnicity are often not well defined and inconsistent across different studies, and the studies comparing barrier properties frequently had methodological flaws, mainly were small, and demonstrated conflicting results. They add that data on Hispanic, Latinx and South Asian populations are lacking, and that robust comparative studies are needed to understand these basic concepts.

Basic concepts. In 2021.

From a 2023 toxicology appraisal: potential differences in stratum corneum physiology and structure across race are derived from a small number of studies with a limited number of research subjects, self-reported race and ethnicity is prone to bias, and the existing studies show varying and sometimes contradictory results.

And from 2003, a point nobody has since addressed: the interpretation of these phenomena should consider anatomic and functional characteristics of ethnic groups alongside socioeconomic, hygienic and nutritional factors. Twenty-three years later, no study in this set adjusts for occupation, wet work, product use or water hardness — all of which move barrier measurements, and none of which is randomly distributed across the groups being compared.

Why this matters beyond the moisturizer aisle

A claim built on a vote count of eight small studies has now traveled into wound care, where it appears as a mechanism explaining why pressure injuries develop differently in dark skin — a field with its own serious measurement problem and every reason to want a biological explanation for it.

That is how a contested finding becomes load-bearing. It gets stated without its caveats in a review, the review gets cited by a different specialty that has no reason to know the caveats existed, and by the third hop it is a fact about a population that is being used to justify a clinical decision.

The honest position, in one line: the barrier evidence differs by phototype in the one study that tested phototype against race directly, by anatomical site enormously, by instrument systematically, and by ethnic category not reliably at all.

This article is educational and is not a substitute for care from someone who can look at your skin. Anything new, changing, painful or spreading should be assessed rather than researched.