Skin Barrier Testing: TEWL Measurement, Corneometer & Clinical Improvement Data

Overview #

TEWL measurement is not a marketing metric. It is the closest proxy we have for actual barrier function, and when a brand asks us to substantiate a “barrier repair” claim, it is the first number we reach for. Corneometer readings tell you hydration status at the stratum corneum — useful, but incomplete on its own. The combination of TEWL + corneometer + clinical grading is what gives a dossier real weight, whether you are filing in the EU, submitting to the FDA, or registering with the NMPA. This article is written from the perspective of the team that runs those tests and prepares those dossiers.

What TEWL and Corneometer Data Actually Measure — and What They Don’t #

TEWL (transepidermal water loss) is measured in g/m²/h using a closed-chamber or open-chamber evaporimeter. We use the Tewameter TM 300 in our in-house testing suite. Normal facial skin sits roughly between 5–10 g/m²/h. Compromised barrier skin — the kind your “barrier repair” product is targeting — typically reads above 15 g/m²/h, and in clinical eczema populations we have seen values above 30 g/m²/h at baseline.

The Corneometer CM 825 measures capacitance-based skin hydration in arbitrary units (AU). Dry skin is generally below 30 AU. Healthy, well-hydrated skin sits between 45–60 AU. What the Corneometer does not tell you is why the skin is dry — whether it is a lipid deficiency, a ceramide imbalance, or just environmental humidity. That distinction matters when you are writing a claim.

Honestly, most brands underestimate how much the measurement protocol affects the data. Room temperature, relative humidity, acclimatization time — all of it shifts your baseline. We run all TEWL measurements at 21°C ± 1°C and 50% ± 5% RH, with a 30-minute acclimatization period. Deviate from that and your week-0 and week-8 numbers are not comparable.

The regulatory relevance here is direct. Under EU Cosmetics Regulation 1223/2009, any efficacy claim must be substantiated. TEWL and corneometer data, when collected under a controlled protocol with sufficient subject numbers, constitute acceptable instrumental evidence. The SCCS Scientific Opinion on claim substantiation is explicit that instrumental data must be paired with a defined methodology — not just a graph from a supplier.

For our barrier repair and sensitive skin product line, we require a minimum of all three instrumental methods before we will write a claim into a product brief. One method alone is not enough.

The Clinical Evidence Standard — What a Dossier Actually Needs #

Here is a study design we ran for a ceramide-niacinamide barrier cream brief. Double-blind, vehicle-controlled, n=42 subjects with self-reported sensitive skin and baseline TEWL above 14 g/m²/h. Duration: 8 weeks, twice-daily application. Primary endpoint: TEWL reduction at week 8 versus vehicle. Result: 28% reduction in TEWL in the active group versus 6% in the vehicle group (p<0.01). Secondary endpoint: Corneometer increase — active group showed +18 AU from baseline versus +4 AU in vehicle. Clinical grading of dryness improved by 1.4 points on a 5-point scale in the active group.

That is a clean dataset. It is also the minimum we consider acceptable for a “clinically tested” claim in the EU market.

What it does not cover — and this is where brands often push back — is the mechanism. The TEWL improvement tells you the barrier got better. It does not tell you whether it was the ceramides, the niacinamide, or the occlusive base doing the work. We are still not fully convinced the clinical evidence cleanly separates ingredient contributions in most multi-active formulas. The supplier data and our own stability and efficacy results do not always agree on this.

For FDA Cosmetics Guidelines, the bar is different. The FDA does not require pre-market clinical substantiation for cosmetic claims — but it does require that claims are truthful and not misleading. In practice, if you are making a “clinically proven” claim on a US label, you need the data to back it up if challenged. The FTC is the more active enforcement body here, and their standard is “competent and reliable scientific evidence.” A single n=42 study is borderline. Two independent studies is safer.

China NMPA is the most demanding of the three markets for special-function claims. Under NMPA Cosmetic Regulation, products making “repair” or “sensitive skin” claims may require human safety testing conducted at a NMPA-designated testing institution inside China. The testing timeline alone — from submission to approval — can run 6–12 months for a new registration. We have seen projects stall here because the brand assumed their EU clinical data would transfer directly. It does not always.

Where Most Brands Get This Wrong #

The brief comes in: “barrier repair cream, suitable for sensitive skin, clinically tested.” Three claims. Each one has a different substantiation requirement depending on the market. This is usually where projects go sideways.

“Suitable for sensitive skin” in the EU requires a dermatologist-tested or repeat insult patch test (RIPT) protocol at minimum. “Barrier repair” requires instrumental data — TEWL specifically. “Clinically tested” requires a defined study with a defined endpoint. You cannot use one study to substantiate all three claims unless it was designed to cover all three endpoints. Most supplier-provided studies are not.

We almost always push back on briefs that arrive with a single supplier efficacy sheet and an expectation that it covers everything. It does not.

The EU SCCS Scientific Opinion framework distinguishes between claims that are “cosmetic” and those that imply a therapeutic effect. “Repairs the skin barrier” sits close to that boundary. We have had EU-market products require claim rewording at the dossier review stage because the original language implied a medical function. “Helps support the skin’s natural barrier” clears the line. “Repairs damaged skin barrier” does not, in most EU member state interpretations.

A lot of clean beauty brands also underestimate how fragile low-pH preservative systems become when you are simultaneously trying to maintain the ceramide-rich lipid matrix that drives barrier repair efficacy. Drop below pH 5.0 and your ceramide crystallization risk increases. Go above pH 6.5 and your phenoxyethanol-based preservative system starts losing efficacy. The window is narrow. On our production line, we have seen emulsion phase separation at scale when the pH drifted by 0.3 units during a 200 kg batch — something that never appeared at 500 g lab scale because the mixing dynamics are completely different.

Regulatory Market Comparison: EU, US, and China #

The three major markets have meaningfully different requirements for barrier and sensitive skin products. Here is how they compare on the dimensions that matter most for dossier preparation:

The CPNP notification timeline of 2–4 weeks sounds fast. It is, for the notification itself. The safety assessment — the Cosmetic Product Safety Report (CPSR) — is what takes time. A well-prepared CPSR for a barrier repair product with multiple actives typically takes 4–8 weeks to complete, depending on the assessor’s queue and whether any ingredients require additional toxicological review.

For NMPA, the 3–12 month range is real. Ordinary cosmetics (普通化妆品) with no special-function claims sit at the faster end. Products claiming “repair,” “anti-sensitivity,” or similar functional language may be classified as special-use cosmetics (特殊化妆品), which triggers the longer pathway. We now require brand partners to confirm their China claim strategy before we finalize the formula — because the formula and the claim are not independent decisions in the NMPA system.

ICH Stability Guidelines apply primarily to pharmaceutical products, but the stability testing logic — accelerated conditions at 40°C/75% RH, real-time at 25°C/60% RH — is what we follow for cosmetic stability dossiers submitted to EU and NMPA. It is the recognized framework, even if not formally mandated for cosmetics.

Formulation Notes for Brand Partners #

What market? What are you expecting on-pack? Those are the first two questions we ask when a barrier repair brief lands on our desk.

If you are targeting the EU with a “clinically tested, suitable for sensitive skin” positioning, we need to plan for a RIPT protocol (minimum 50 subjects, 9-week induction + challenge design) and an instrumental study with TEWL and corneometer endpoints. Budget 12–16 weeks for testing before you have a complete dossier. The formula needs to be locked before testing starts — late-stage ingredient changes invalidate the data.

For the US market, the documentation burden is lower, but the FTC exposure is real. We recommend at minimum one controlled instrumental study (n≥30, 4–8 weeks) before making any “clinically proven” claim. If you are selling through a major retailer, their compliance team will ask for it.

China NMPA is the one that requires the most upfront planning. If your product concept involves any claim that could be read as “repair” or “anti-sensitivity,” we will flag it at brief intake and discuss whether to pursue the special-use pathway or reframe the claim to stay in the ordinary cosmetics category. That decision affects formula, timeline, and cost — sometimes significantly.

Our internal documentation package for barrier repair products includes: TEWL and corneometer study protocol and raw data, RIPT report, CPSR (EU), stability data to ISO Standards accelerated conditions, ingredient safety dossiers, and a claim substantiation matrix mapping each on-pack claim to its supporting evidence. We prepare this as a single submission-ready package, not a collection of loose files.

For our encapsulation technology platform — which we use for ceramide and active delivery in barrier formulas — the documentation package also includes encapsulation efficiency data and release kinetics, because NMPA reviewers have started asking for this on complex delivery systems.

Frequently Asked Questions #

Q: We want to say “clinically proven to reduce TEWL by 30%” on pack — is that defensible?

Defensible if the study was run under a controlled protocol with at least 30 subjects and the 30% figure is the mean reduction with statistical significance (p<0.05). If that number came from a supplier’s in-house study with n=10, we would not put it on pack. We have seen brands get challenged on exactly this by EU notified bodies.

Q: How long does it take to get a barrier repair product registered in China if we want to make a “repair” claim?

Plan for 9–12 months minimum from formula lock to approval if you are going the special-use cosmetics route. If you can reframe the claim to stay in ordinary cosmetics, the NMPA filing is a notification process and can move in 30–60 days. We almost always have this conversation at brief intake because the answer changes the project timeline significantly.

Q: Do we need a separate EU Responsible Person if we are already selling in the UK post-Brexit?

Yes. UK and EU are separate regulatory territories post-Brexit. You need an EU-based Responsible Person for CPNP notification and a UK Responsible Person for the UK SCPN system. Two separate entities, two separate dossiers. We can connect you with RP services in both territories.

Q: Our dermatologist-tested claim was approved for our US product — can we use the same language in the EU?

Not automatically. “Dermatologist-tested” in the EU requires the test to have been conducted under a defined protocol with a qualified dermatologist, and the claim must be substantiated in the product dossier. If your US testing was a simple dermatologist review rather than a formal RIPT or usage study, it will not satisfy EU claim substantiation requirements. We review this case by case.

Q: What is the minimum subject number for a TEWL study to be accepted in an EU dossier?

There is no hard regulatory minimum, but the SCCS guidance and standard practice point to n=30 as the floor for a statistically meaningful instrumental study. We prefer n=40–50 for primary claims. Below 30, a reviewer can reasonably challenge the statistical power, and we have seen dossiers returned for exactly that reason.

Have a product concept in mind? Contact our formulation team to request a complimentary brief review.