Experimental Dermatology 2006: 15: 483–492 Copyright # 2006 The Authors.
Journal compilation # 2006 Blackwell Munksgaard
Blackwell Munksgaard . Printed in Singapore
doi: 10.1111/j.1600-0625.2006.00437.x
Transepidermal water loss reflects
permeability barrier status: validation in
human and rodent in vivo and ex vivo
models
Fluhr JW, Feingold KR, Elias PM. Transepidermal water loss reflects
permeability barrier status: validation in human and rodent in vivo and
ex vivo models.
Exp Dermatol 2006: 15: 483–492. # 2006 The Authors. Journal
compilation # 2006 Blackwell Munksgaard
Abstract: Permeability barrier function is measured with instruments
that assess transepidermal water loss (TEWL), either with closed- or
open-loop systems. Yet, the validity of TEWL as a measure of barrier
status has been questioned recently. Hence, we tested the validity of this
measure by comparing TEWL across a wide range of perturbations,
with a variety of methods, and in a variety of models. TEWL rates with
two closed-chamber systems (VapoMeter and H4300) and one closed-
loop system (MEECO) under different experimental in vivo conditions
were compared with data from four open-loop instruments, i.e. TM 210,
TM 300, DermaLab and EP 1. The instruments were compared in vivo
both in humans and hairless mice skin subjected to different degrees of
acute barrier disruption. The values obtained with bioengineering
systems were correlated with absolute water loss rates, determined
gravimetrically. Measurements with both closed and open systems
correlated not only with each other, but each method detected
different degrees of barrier dysfunction. Although all instruments
differentiated among gradations in TEWL in the mid-range of barrier
disruption in vivo, differences in very low and very high levels of
disruption were less accurately measured with the H4300 and DermaLab
systems. Nevertheless, a high Pearson correlation coefficient (r) was
calculated for data from all instruments vs. gravimetrically assessed TEWL.
Together, these results verify the utility of TEWL as a measure of
permeability barrier status. Moreover, all tested instruments are reliable tools
for the assessment of variations in permeability barrier function.
Introduction
‘Barrier’ is defined in the Merriam-Webster
Online Dictionary (1) as either (a) ‘something
material that blocks or is intended to block pas-
sage’ or (b) ‘a natural formation or structure that
prevents or hinders movement or action’. The
assessment of epidermal permeability barrier
function, routinely employs measurements of
transepidermal water loss (TEWL), which have
been presented to provide information about
Abbreviations: SC, stratum corneum; TEWL, transepidermal water loss
EXPERIMENTAL DERMATOLOGY
ISSN 0906-6705
Joachim W. Fluhr1, Kenneth
R. Feingold2 and Peter M. Elias2
1
Department of Dermatology and Allergology,
Friedrich-Schiller-University Jena, Germany;
2
Dermatology and Medical (Metabolism)
Services, Veterans Affairs Medical Center, and
Departments of Dermatology & Medicine,
University of California, San Francisco, CA, USA
Key words: barrier function – epidermal
permeability barrier – surface temperature –
transepidermal water loss
Joachim Fluhr, MD
Department of Dermatology and Allergology
Skin Physiology Laboratory
Friedrich Schiller University
Erfurter Strasse 35
Jena 07740
Germany
Tel.: þ49 36 41 937 399
Fax: þ49 36 41 937 435
e-mail: Joachim.Fluhr@med.uni-jena.de
Accepted for publication 20 March 2006
permeability barrier status under either normal,
experimentally perturbed, or diseased conditions
(2). The degree of barrier disruption, again
assessed as TEWL, also correlates with drug
penetration of different compounds, in con-
cordance with an ‘outside-in’ concept of barrier
function (3–5). Recently disease activity has been
assessed using TEWL quantification (6–8). In
addition, the regulation of lipid biochemical
pathways affects not only TEWL (9–13), but
also drug penetration into and through the
epidermal compartment. Specifically, increased
and/or delayed uptake of drugs is evident when
483
Fluhr et al.
biochemical inhibitors or enhancers of lipid
synthesis are utilized. Moreover, this correlation
between drug penetration and TEWL has been
shown to be exponential (14,15).
Yet, based upon ex vivo studies in devitalized
skin by Chilcott et al. (16) the value of assess-
ment of epidermal barrier function through
quantification of TEWL was challenged
recently. Resolution of this issue is of pressing
importance to researchers of skin biology and
pathophysiology, we abrogated barrier status,
both in vivo and under ex vivo conditions,
comparing TEWL data from various bio-
engineering instruments with true water loss,
assessed gravimetrically. In addition, we evalu-
ated specific measures of TEWL, comparing
TEWL levels obtained with several different,
commercially available instruments (open
chamber, closed chamber and closed loop). As
the principle and the construction of open
chamber devices are quiet similar, comparable
values should be expected under comparable
conditions. Yet, as noted by Rogiers (17):
‘Only a limited number of comparison between
different types of instruments have been
described in the literature’.
These devices all measure water evaporation
from the skin surface; i.e. across the epidermis,
thus quantifying epidermal permeability barrier
function. Moreover, as SI units for these instru-
ments are provided in g/m2h (except the
MEECO, which measures ppm/0.5 cm2h),
their recorded values should be comparable.
Yet, differences could still result from differences
in either measurement principles or meas-
urement time frames. We therefore compared
four open-chamber (Evaporimeter, Tewameter
TM 210 and TM 300, DermaLab), two closed
chamber (VaporMeter and H4300) systems as
well as one closed-loop system (MEECO), in
order to comprehensively compare commercially
available instruments intended to measure
TEWL. Our results show that all of the instru-
ments are useful methods for the assessment of
epidermal barrier function under different in vivo
conditions, both in normal rodents (hairless
mice) and in humans. Furthermore, data from
these biophysical instruments were validated in
an ex vivo model that simulates the in vivo situ-
ation, whereas minimizing possible variability
because of external stressors (e.g. changes in
temperature or relative humidity). Our results
both validate TEWL as a measure barrier func-
tion and the utility of all these instruments to
assess permeability status under a variety of dif-
ferent conditions.
484
Materials and methods
Study design
Study conditions. All measurements were performed in clima-
tized rooms at a mean temperature of 22.1 C (22–23 C min–
max) and a mean relative humidity of 46.7% (44–52% min–max)
for in vivo and ex vivo studies. Both human and animal studies
were undertaken at the same time of the year (spring). Sequential
tape strippings (TSs) were performed with 22 mm D-Squame1-
100 tapes (CuDerm, Dallas, TX, USA). Petrolatum was pur-
chased from Walgreen Pharmacy (Deerfield, IL, USA).
Ex vivo validation of the mouse model by gravimetric assay. For
the ex vivo experiments, hairless mice(details as in the in vivo
studies) were sacrificed by inhalation of an overdose of
halothane prior to the immediate excision of full-thickness skin
from both flanks. The subdermal fat was removed from the
excised skin, and the full-thickness skin pieces were placed
epidermis-side outward on NaCl 0.9% w/w soaked filter paper
in Petri plastic dishes. The surface area of each skin piece was
photographed and the dimension was quantified with a
computer-assisted measuring system (SCION Image beta 4.02:
NIH-based image analyze shareware: www.scioncorp.com/
frames/fr_scion_products.htm) in order to calculate the water
loss/cm2. The edges of skin pieces were sealed with petrolatum
in order to prevent lateral water loss. The ambient temperature
and relative humidity were kept constant, as described for the
in vivo studies above. The skin was placed inside a microbalance
(Mettler-Toledo, Columbus, OH, USA) and the values for
TEWL were assessed every 15 min over 1 h. on skin from 7 to
12 different mice and correlated with the measured loss of weight
attributed to TEWL. The values were calculated as g/m2 h.
In vivo mouse model. Male hairless mice (Skh1/Hr), 8–12 weeks
old, were purchased form Charles River laboratories
(Wilmington, MA, USA) and fed Purina mouse diet and
water ad libitum. All of the studies were performed after proto-
col approval by the Animal Care Committee at the VAMC, San
Francisco. Epidermal barrier function was altered by two diver-
gent approaches: (a) lowering of the TEWL values in intact skin
by applying petrolatum or (b) increasing TEWL values by
sequential TS with standardized (D-Squame1) tapes. The mod-
els were grouped in two grades of barrier perturbation:
(1) Partially intact barrier function with mild disruption by TS
(2); and
(2) Moderately damaged barrier function (TS – 3), severely
damaged (TS – 6) or very severely (TS – 8). The 8 TS
was equivalent to complete removal of the stratum disjunc-
tum, leaving one to two layers of residual stratum corneum,
confirmed by conventional histology (data not shown).
All studies were performed in parallel on the same hairless
mice: one flank for baseline barrier measurements and subse-
quent petrolatum treatment, whereas the other flank was tape-
stripped (TS) and measured after sequential TS at the intervals
described above. To control for another external variable, skin
surface temperature (AT) was also measured at baseline, after
both 2 and 6 TS in order to assess the possible influence of
temperature on assessed values with the different instruments.
Minor barrier perturbations in humans in vivo
Measurements were performed on the volar forearm of five
males and six normal female volunteers, with a mean age of
34 years. (range: 20–49 years), after an adaptation time of at
least 15 min in a climate-controlled room. All volunteers pro-
vided written, informed consent prior to participation in the
 Transepidermal water loss as measure of permeability barrier
study. We mildly damaged the forearm SC by sequential TS with
D-Squame1 tapes (10). Although skin surface temperatures
were measured at baseline and after 10 TS, temperatures did
not vary significantly at any level of barrier function in the same
subjects. Reproducibility of results was assessed by performing 10
separate measurements of each perturbed skin site, with subsequent
calculations of the coefficient of variation (CV).
Measurement Devices
Open chamber instruments
Evaporimeter. The ServoMed evaporimeter (Servomed,
Stockholm, Sweden) is based upon a vapor pressure gradient
estimation principle (18). [The probe weighs 104 g and has a
cylindrical open chamber measuring system, diameter 12 mm,
height 15 mm and skin area of 1.13 cm2. Two sensor units are
placed at 3 and 6 mm distance from the skin surface. The
standard version of the instrument has a digital display
showing water evaporation (0–300 mg/m2 h with two ranges of
sensitivity), relative humidity (RH) (0–100%) and water vapor
pressure (0-500 mmHg). The probe can be adapted with a chimney
extension in order to reduce air turbulence and with a metallic
shield.] The underlying principle is based upon the measurement
of a vapor pressure gradient, subjacent to the skin surface
that drives the rate of water exchange across the skin. The
probe consists of an open cylinder placed perpendicular to
the skin surface. For each data point, the local relative
humidity and temperature are measured by paired sensors,
and the vapor pressure at each point is computed. The
difference between the vapor pressure at the two points along
the gradient relates directly to rates of evaporative water loss
from that skin site.
Tewameter–TM 210. The measurement of water evaporation
rates with the Tewameter TM 210 (Courage and Khazaka,
Cologne, Germany) is based upon diffusion principles in an
open cylinder system dm/dt ¼  DA dp/dx (where A represents
the surface in m2, m the water transported in g, t the time in
hours, D the diffusion constant 0.877 g/m h, p the vapor
pressure of the atmosphere in mmHg, x the distance from the
skin surface to point of measurement in meters) according to
Ficks law. [The probe weighs 25 g and has a cylindrical open
chamber measuring system, diameter of 10 mm, height of
20 mm, and covered skin area of 0.79 cm2. The two sensor
units are placed at 3 and 8 mm distance from the skin surface.
A special probe holder with a clamp is furnished with the
device, which allows a fixed positioning of the probe on the
skin surface.] The vapor density gradient is measured indirectly
by two pairs of sensors (temperature and relative humidity)
inside a hollow cylinder, and the resulting data are analyzed
by a microprocessor. Measurement values are given in g/m2 h. The
standard version of the instrument shows the following experimental
parameters: TEWL (0–90 g/m2 h), ambient relative humidity
(0–100%) and temperature (0–50 C) at the level of the sensors.
Tewameter–TM 300: Although the measurement principle of
the TM 300 is the same as for the TM 210, the former differs in
operation as follows: the probe weighs 90 g with a diameter of
10 mm, height of 2 cm and covers a surface area of 0.79 cm2.
Measurements are started by pressing a button either in the
probe or on the PC. In addition to TEWL values, calculations of
SSWL value (skin surface water loss after occlusion) and assess-
ment of partial water evaporation pressures (pd) are possible.
The measurement technology is completely built into the probe
itself. As the probe of the TM 300 is preheated, it quickly
reaches a steady state where values should be taken (17,19),
allowing for faster measurements in comparison to the TM 210.
The TM 300 is available either as a stand-alone device, or as
part of a Multiprobe Adapter (MPA) (Courage and Khazaka),
which then must be attached to a PC.
DermaLab. The DermaLab (Cortex Technology, Hadsund,
Denmark) can be equipped with a TEWL probe. The
measurement principle is similar to the Tewameter devices
(see above): The probe consists of two combined humidity/
temperature sensors with a diameter of 10 mm covering an
area of 0.79 cm2 mounted in a diffusion chamber. The water
pressure measured in the open chamber is used to calculate the
evaporated water over a constant skin area. The values of the
different measurements are given in g/m2/h. The maximum
obtainable value is 250 g/m2/ h. Measurement duration can be
selected for different time lengths from 1 to 250 s.
Closed chamber devices
VapoMeter. The VapoMeter (Delfin Technologies, Kuopio,
Finland) is a small closed-chamber system that calculates
water evaporation rates. [The device is portable, weighs
150 g and ambient air flow does not disturb measurements.
The inner diameter of the VapoMeter is 11 mm and covers
an area of 95 mm2. An additional surface adapter with a
smaller surface area (4.5 mm diameter; 16 mm2 surface area)
can be inserted in order to measure smaller areas (e.g. nail
surface).] Increases in relative humidity correlate with
TEWL, as the instrument compares the ambient humidity
with the humidity within the chamber after a short filling
time of 10 s (20,21). Compared with open chamber techni-
ques, measurement times are standardized and relatively
short (10 s). Although continuous measurements are not
possible, because each measurement period is so short, up
to three measurements per minute can be performed at low
evaporation rates. However, at higher evaporation rates, the
fixed measurement period might not represent steady-state
values, because the instrument does not monitor changes in values
over longer measurement times. The values of the different
measurements are expressed in g/m2 h, with a maximum obtainable
value of 300 g/m2 h.
H4300. The H4300 is a closed, hand-held, portable chamber
system (22), which consists of a monitor and a probe. The probe
includes thermo- and humidity sensors. The instrument displays
the following experimental parameters: evaporative quantity
(0–300 g/m2/h), relative humidity (0–100%) and probe
temperature (5–45 C). The evaporative quantity (i.e. water
loss) is calculated with the HR 4300 closed-type instrument in
accordance with a predetermined algorithm, and the results
reflect the response curves for the temperature and humidity
sensors after application to the skin surface.
Closed-loop devices
MEECO. Water-loss determinations also can be performed
with an electrolytic moisture analyser (MEECO, Warrington,
PA, USA). The results are presented in p.p.m. of water loss/
0.5 m2 h, which corresponds approximately to 10 g/m2h
(9,23,24). This instrument utilizes a closed-loop system that
is filled with dry nitrogen gas, and the sensor measures the
moisture captured in an opening of the loop. The metal cap
has to be firmly pressed against the skin for analyses; and in
doing this, fragile skin can potentially be damaged, and local
blood flow can be diminished. Furthermore, excess pressure may
cause local discomfort. Because of the design of the probe, with
its attached in- and out-going small tubes, measurement sites are
largely limited to the extremities in humans.
485
Fluhr et al.

Skin surface temperature a MEECO ***

100 *** ***
In all these studies, skin surface temperature was assessed with a
battery-driven HI 93530 K-Thermocouple Thermometer from
75

(mean ± SEM)
Hanna Instruments (Bedfordshire, UK).

g/m2 h
50
Statistical analyses and data presentation
25 ***
n.s.
Statistical analyses were performed using Prism 3 software (Graph ***
Pad Software Inc., San Diego, CA, USA). Normal distributions 0
ANOVA P < 0.0001 P < 0.0001
were tested before calculating comparisons with a paired t-test.
When three or more experimental groups were compared, H4300 ***
b
ANOVA analyses were performed, followed by a pair-wise, post- ** ***
hoc, a-adjusted Bonferroni Test. Values are given as mean + stan- 150

(mean ± SEM)
dard error of the mean (SEM) (Figs 1 and 2). The relationships

g/m2 h
between gravimetrically measured water loss and the values 100
assessed with the different devices were calculated. Pearson correl-
ation coefficients were calculated and the statistical significance of
50 **
correlations were determined. Correlations between the different n.s.
devices were determined from both mouse and human in vivo ***
studies. A statistically significant difference was set at P < 0.05 0
ANOVA P < 0.0001 P < 0.0001
(*). P < 0.01 was labelled as (**) and P < 0.001 as (***).
Reproducibility was tested by analyzing with 10 repeated
measurement of one damaged skin site, and subsequently calcul- c VapoMeter ***
n.s.
ated as the coefficient of variation (CV). 250
***

200

(mean ± SEM)
Results 150

g/m2 h
Gravimetric data validate in vivo measurements of 100
***
TEWL 50
n.s.
***

The key issue addressed in this study relates to 0

the validity of TEWL measures as an indicator of
barrier function. Hence, we correlated gravi-
metric measurements on freshly excised normal
skin ex vivo with measures of TEWL with the
different devices. The gravimetric values correl-
ated significantly with the values for all instru-
ments tested (Table 1). Correlations could not
be calculated for the DermaLab device because
of five missing values. The closest correlation were
for EP1 (r ¼ 0.8076) > TM 210 (r ¼ 0.7666) >
VapoMeter (r ¼ 0.7630) > TM 300 (r ¼ 0.7557) >
H4300 (r ¼ 0.7082) > MEECO (r ¼ 0.6825)
(Table 1a). In contrast, surface temperature did
not correlate with gravimetric values, as skin
temperatures remained constant over the entire
test period (P > 0.05; Pearson r < |0.28|)
(Table 1b). These results demonstrate that
TEWL measurements, with both open and closed
systems, correlate significantly with absolute
rates of water loss, assessed gravimetrically.
Both closed and open systems detect minor barrier
perturbations in murine skin
We next assessed whether small changes of epi-
dermal barrier function, induced by petrolatum
treatment of intact human skin (as a parameter
for inducing a slight decrease in evaporative
waterloss) or very mild (2 TS) disruption
ANOVA P < 0.0001 P < 0.0001
Mild (2×)
Untreated
Moderate (3×)
Petrolatum
Severe (6×)
Very severe (8×)
Figure 1. In vivo mouse model: closed chamber and closed-loop
systems. In the low-range values, all the three closed systems
are capable of differentiating between both untreated and
petrolatum-treated areas and the mild barrier damage
[2 tape-stripped (TS)]. Also the high-value ranges can be
differentiated by these three instruments: moderate (3 TS),
severe (6 TS) and very severe (8 TS). (a) Measurements
with the MEECO revealed the following differences ANOVA
for the low range: P < 0.0001; for the high range: P < 0.0001
(n ¼ 12) (b) Measurements with the H4300 showed the follow-
ing differences ANOVA for the low range: P < 0.0001; for the
high range: P < 0.0001 (n ¼ 12). (c) Measurements with the
VapoMeter resulted in the following differences ANOVA for
the low range: P < 0.0001; for the high range: P < 0.0001
(n ¼ 12).
could be detected accurately as significant
changes in TEWL. Figure 1 displays the data
from three closed chamber systems (left sections
of Fig. 1a–c), whereas Fig. 2 shows the in vivo
data for the open chamber systems (left sections
of Fig. 2a–d). All three closed chamber devices
detected significant differences between baseline
and mild barrier damage, but not between base-
line vs. petrolatum-treated areas. Likewise, none
of the four open chamber systems could differ-
entiate significantly between baseline and lowered
TEWL on petrolatum-treated, normal skin (left

486
 Transepidermal water loss as measure of permeability barrier

a TM 210 b ** Untreated vs. 10× TS (human)

TM 300
** n.s.
100 30
(mean ± SEM) 60 90

(mean ± SEM)
50 80 **
70 25 **
g/m2 h

g/m h
40 60

2
30 ** 50 ***
n.s. 40 n.s. *** **
***

(mean ± SEM)
20 30 20
10 20

g/m2 h
10
0 0
ANOVA P < 0.0001 n.s. ANOVA P < 0.0001 P < 0.005 15
* *
c EP1 ** d DermaLab *** 10 *
n.s. n.s.
* **
60 60
n.s .
(mean ± SEM)

(mean ± SEM)
50 50 5
g/m2 h

40
30
***
20 n.s.
***
10
0 0
ANOVA P < 0.0001 P < 0.0001
Untreated
Vaseline
Severe (6×)
Very severe (8×)
Figure 2. In vivo mouse model: open chamber systems. In the
low-range values, all the four open chamber systems are capable
of differentiating between both untreated and petrolatum-
treated areas and the mild barrier damage [2 tape-stripped (TS)].
Also the high-value ranges can be differentiated by these instru-
ments: moderate (3 TS), severe (6 TS) and very severe
(8 TS) except the older Tewameter TM 210. However the
difference between severe (6 TS) and very severe (8 TS)
barrier damage was only detected by the Evaporimeter EP1.
(a) Measurements with the Tewameter TM 210 showed the
following differences: ANOVA for the low range P < 0.0001; for
the high range: P > 0.05 (n ¼ 12). (b) Measurements with the
Tewameter TM 300 revealed the following differences:
ANOVA for the low range P < 0.0001; for the high range:
P < 0.005 (n ¼ 12). (c) Measurements with the Evaporimeter
EP1 showed the following differences: ANOVA for the low
range: P < 0.0001; for the high range: P < 0.0001 (n ¼ 12).
(d) Measurements with the DermaLab revealed the following
differences: ANOVA for the low range: P < 0.0001; for the high
range: P < 0.0001 (n ¼ 12).
section of Fig. 2a–d). However, all 4 open chamber
devices were able to detect significant differences
between baseline and mildly damaged skin (2 TS).
Together, these studies show that TEWL measure-
ments are capable of detecting small increments in
barrier function in murine skin in vivo, independent
of the specific detection methodology employed. In
contrast, further reductions from normal barrier
function could not be detected, presumably because
basal evaporation rate is already very low in normal
skin.
Both closed and open systems detect moderate-to-
severe perturbations of murine skin
In the next group of studies, we assessed the
sensitivity of TEWL measurements after more
drastic perturbations induced by skin stripping
(moderate, severe and very severe damage) to
the barrier. The MEECO and the H-4300 signifi-
cantly differentiated all three degrees of barrier
g/m2 h 40
30
** 0
20 n.s. MEECO VapoMeter TM300 EP1
**
10 H4300 TM210 DermaLab
ANOVA P < 0.0001 P < 0.0001 ANOVA: P < 0.0001
Mild (2×) Untreated
Moderate (3×)
10 Tape strippings
Figure 3. In vivo study on the ventral forearm of healthy human
volunteers: baseline values were compared with those after a
mild damage of the stratum corneum by 10 sequential tape
stripping with D-Squame1 tapes. A significant difference
(P < 0.05 and lower; ANOVA P < 0.0001) between baseline
and disrupted barrier could be detected with all instruments
except the MEECO (n ¼ 10).
damage (Table 1). The new TM 300 (right section
of Fig. 2b), as well as the DermaLab (right section
of Fig. 2d), also were capable of differentiating
between moderate TS (3) and both severe
damage (6) and very severe damage (8) in hair-
less mouse skin. The relatively old, EP 1 device also
detected significant differences between all three
stages of moderate through severe barrier damage,
but in contrast to the closed-chamber systems and
the older TM 210 device did not reveal differences
between moderately severe and severe barrier
damage (right section of Fig. 2a).
Correlations between the seven devices were
calculated with the Pearson test (Table 2). A sig-
nificant correlation were calculated for all tested
instruments, with a Pearson coefficient of at least
r > 0.83 for all instruments. However, the TM
210 correlated less well (r between 0.8353 and
0.9160) then the other devices, which showed
higher correlations (r up to 0.9879). Although,
surface temperature showed small, but signifi-
cant differences for untreated skin (baseline:
33.1 C) compared with mild barrier impairment
(2 TS: 31.0 C) and the severely disrupted sites
(6 TS: 30.6 C), these differences could be
attributed to the extended anaesthesia time
required for progressive barrier disruption.
These results show that both closed and open
systems are sensitive to incremental external per-
turbations of murine SC, in a moderate to very
severe range of barrier disruption.

487
Fluhr et al.

Table 1. Gravimetric measurements (water evaporation) correlate significantly with the device values

MEECO H4300 VapoMeter TM 210 TM 300 DermaLab EP1 Temperature

10 12 12 12 12 7 12 12
(a) Gravimetric TEWL
P (two-tailed) 0.0297 0.0099 0.0039 0.0036 0.0045 n.d. 0.0015 0.3335
Pearson 0.6825 0.7082 0.763 0.7666 0.7557 0.8076 0.306
(b) Temperature
P (two-tailed) 0.3836 0.7416 0.5294 0.6236 0.4443 0.6822 0.731 –
Pearson 0.2769 0.1066 0.2018 0.1581 0.2442 0.1906 0.1111 –

(a) The gravimetric measurement as an independent parameter of evaporation of water on freshly excised skin in a mouse ex vivo model showed a significant
correlation with the values from six devices (DermaLab not evaluated due to missing values) with a Pearson correlation coefficient r > 0.682. No correlation
between the surface temperature and the gravimetric measurement was detectable thus excluding a temperature effect. (b) No correlation between the surface
temperature and the device values was seen.

Both closed and open systems detect minor
perturbations of human SC in vivo
Figure 3 shows data from an in vivo study on the
ventral forearm of healthy human volunteers.
Baseline values were compared with measurements
after mild damage of SC induced by 10 TS of the
ventral forearm with standardized D-Squame1
tapes. Significant differences between basal and
disrupted sites could be detected with all instru-
ments, except the MEECO. These differences
were more pronounced with the closed chamber
system, H-4300, and both Tewameters (TM 210
and TM 300). The choice of model did not sig-
nificantly influence skin surface temperature
except for the closed chamber system H-4300,
which also correlated significantly with changes
in surface temperature (Table 3b).
The correlations between the seven different
instruments were calculated based on the 22 values
obtained from the 11 individuals both on untreated
and mildly barrier damaged skin. All instruments
displayed significant correlations with each other
(Table 3a). However, data for the closed loop,
MEECO system, correlated less well with the H-
4300 and the TM 210 (Pearson coefficient < 0.600).
The correlation values of all other instruments
were highly significant (P < 0.01 or less). Finally,
we tested the reproducibility of measurements after
10 different measurements of individual damaged
skin sites, and subsequently calculated the CV
from a total of 22 measurements. The CV was
highest for DermaLab (18.0) followed by the
MEECO (16.6) values. The H-4300 (12.4),
VapoMeter (12.8), TM 210 (12.0) and the EP 1
(13.5) showed values of around 13, whereas the
TM 300 (1.9) had the lowest CV values, indicating
the best reproducibility. In summary, with the
exception of the MEECO, both closed and
open systems detect minor barrier perturbations
of human SC in vivo. All systems correlated well
with each other, and the values were not tem-
perature dependent, except for the H-4300.
Finally, all of these instruments demonstrated
very good reproducibility.
Discussion
Current models of epidermal permeability barrier
function assume that the central parameter of this
function reflects evaporative water loss originating
from deeper layers of the skin, i.e. TEWL (25,26).
Every experimental change, whether in vivo or
in vitro, should respect the physiological back-
ground of this fragile tissue compartment, and
the metabolic processes that maintain functional
homeostasis. A literature search (www.pubme-
d.org) with the key words ‘TEWL’ and ‘skin’
show a high number of Medline-listed publications

Table 2. Significant correlation between all instruments in the mouse in vivo model

Pearson r MEECO H4300 VapoMeter TM 210 TM 300 DermaLab EP1

MEECO X 0.9746 0.9560 0.9160 0.9149 0.9253 0.9247

H4300-S *** X 0.9811 0.8991 0.9513 0.9536 0.9596
VapoMeter *** *** X 0.8759 0.9750 0.9716 0.9797
*** *** ***
TM 210 *** *** *** X 0.8353 0.8361 0.8544
TM 300 *** *** *** *** *** X 0.9879 0.9875
DermaLab *** *** *** *** *** *** X 0.9847

All comparisons: ***P < 0.001.

All instruments have a significant correlation with each other in the mouse model (n ¼ 68) with very high Pearson correlation coefficients r > 0.91 except for the
TM 210 vs. the other open chamber instruments, H4300 and VapoMeter with slightly lower correlation coefficients.

488
 Transepidermal water loss as measure of permeability barrier

during recent years (1999: 41; 2000: 47; 2001: 71;
2002: 66; 2003: 50; 2004: 60). Yet, TEWL has been
measured as far back as 1965 (27). In basic research
(28–38), as well as in applied clinical (39–46),
pharmacological (47–50), and cosmetic studies
(51–54), quantification of changes in epidermal
barrier function is perceived to possess direct rele-
vance for both assessment of skin physiology,
pathophysiology, as well as product performance.
Yet, the current model of quantitative assess-
ment of epidermal barrier function, utilizing
TEWL as a measuring parameter, was challenged
recently by Chilcott and co-workers (16). In their
models, the penetration of a lipophilic agent, sulfur
mustard, was assessed across both heat-split
human epidermis and non-pigmented pig skin,
that had been stored for up to 14 days, with pene-
tration studies extended over 96 h postheat separ-
ation. Thus, not only was the skin structure
potentially compromised, but the additional
potentially deleterious effects of heat separation
on the extracellular lipid matrix and corneocytes
as major components of the epidermal barrier
function, were not addressed. Although the
authors claimed that structural integrity was main-
tained under these conditions, they provided no
functional, structural or ultrastructural evidence
in support of this assertion. Finally, Chilcott relied
solely on penetration of externally applied sub-
stances, whereas barrier function implies not only
outside-in, but also inside-out functions (25).
Nevertheless, they concluded, in contrast to a
wide range of in vivo and in vitro data published
over three decades, that TEWL does not reflect
actual permeability barrier status.
Extensive prior work has demonstrated that the
extent of epidermal physiological responses to
external stressors correlates well with TEWL.
Specific examples include: DNA synthesis and epi-
dermal proliferation (55,56), lipid synthesis (23,57–
59) and lipid enzymatic activity and activation state
(60,61).
Therefore, to directly address this issue, we com-
pared gravimetric data from an ex vivo model with-
out major alterations due to cold, heat, structural
defects or other physico-chemical insults, thus
insuring the integrity of samples for the assessment
of epidermal barrier function. In our study, the
open chamber, the closed chamber and the
closed-loop systems all accurately and reproduci-
bly measured TEWL, as shown by the high degree
of correlation with gravimetric data. However, our
studies did reveal certain limitations among the
different instruments, and therefore one has to
choose the type of system carefully (open or closed
system), and the most appropriate instrument for
the planned studies. One conclusion from our study
is the need to calibrate the different devices in order
for them to objectively measure the amounts of
water loss per surface area and time in SI units.
Few comparative studies with different devices are
published and most of them lack an outside validation
parameter for the measured water evaporation
(22,62–67). The present study is the most complete
comparison of commercially available devices for
measuring TEWL, utilizing open chamber, closed
chamber and closed loop systems. Furthermore,
human in vivo models, mammalian in vivo models
and an ex vivo model were chosen in order to cover
the broad spectrum of possible research fields regard-
ing epidermal barrier function, but each device had
specific drawbacks and limitations e.g. the tempera-
ture dependency of the H4300 measurements in the
human in vivo measurements (Table 3). We were able

Table 3. Significant correlation between all instruments in the human in vivo model is independent from temperature (except H4300)

Pearson r

P-value MEECO H4300 VapoMeter TM 210 TM 300 DermaLab EP1

(a)
MEECO X 0.4700 0.7064 0.5598 0.6269 0.6763 0.8383
H4300 * X 0.7029 0.7633 0.6475 0.7602 0.5858
VapoMeter *** *** X 0.8926 0.7833 0.9050 0.8780
TM 210 * *** *** X 0.7325 0.8639 0.7702
TM 300 ** ** *** *** X 0.9371 0.8365
DermaLab ** *** *** *** *** X 0.8735
EP1 *** ** *** *** *** *** X
(b) Temperature
Pearson r 0.2793 0.497 0.2646 0.3637 0.4025 0.4184 0.2861
P-value n.s. 0.0258 n.s. n.s. n.s. n.s. n.s.

All comparisons: *P < 0.05, **P < 0.01,***P < 0.001.

(a) All instruments have significant correlation with each other in the human in vivo model (n ¼ 22). However, for the closed-loop system MEECO, a relatively weak
correlation was detected with the other closed system H-4300 and the TM 210 (Pearson coefficient < 0.600). The correlation values of all the other instruments
were highly significant (P < 0.01 or smaller). (b) We also tested whether the results were influenced by any changes in the surface temperature and interestingly
the values of the closed chamber system H-4300 correlated significantly with those of the surface temperature.

489
Fluhr et al.
to provide evidence, that both closed and open sys-
tems are sensitive to minor barrier perturbations to
murine stratum corneum in vivo.
Our results demonstrate that TEWL measure-
ments correlate significantly with absolute rates
of water loss assessed gravimetrically indicating
the intension of quantifying the amount of eva-
porating water at the skin surface as a maker for
barrier function was reached. All systems correl-
ated well with each other and the values were not
temperature dependent except for H4300 and all
instruments showed good reproducibility in the
human in vivo model.
Acknowledgements
This study was supported by NIH grants AR 050629, AR 39448
(PE), RR00173 and HD29706, and the Medical Research
Service, Department of Veterans Affairs, San Francisco, CA.
Special thanks to Courage & Khazaka, Delfin Company and
Gary Grove for technical support.
References
1. Merriam-Webster. Merriam-Webster Online
Dictionary. Available at http://www.m-w.com/cgi-bin/
dictionary?book=Dictionary&va=barrier2005.
2. Pinnagoda J, Tupker R A, Smit J A, Coenraads P J,
Nater J P. The intra- and inter-individual variability
and reliability of transepidermal water loss measure-
ments. Contact Dermatitis 1989: 21: 255–259.
3. Tsai J C, Guy R H, Thornfeldt C R, Gao W N,
Feingold K R, Elias P M. Metabolic approaches to
enhance transdermal drug delivery. 1. Effect of lipid
synthesis inhibitors. J Pharm Sci 1996: 85: 643–648.
4. Curdy C, Naik A, Kalia Y N, Alberti I, Guy R H. Non-
invasive assessment of the effect of formulation excipi-
ents on stratum corneum barrier function in vivo. Int J
Pharm 2004: 271: 251–256.
5. Panchagnula R, Bokalial R, Sharma P, Khandavilli S.
Transdermal delivery of naloxone: skin permeation,
pharmacokinetic, irritancy and stability studies. Int J
Pharm 2005: 293: 213–223.
6. Angelova-Fischer I, Bauer A, Hipler U C et al. The
objective severity assessment of atopic dermatitis
(OSAAD) score: validity, reliability and sensitivity in
adult patients with atopic dermatitis. Br J Dermatol
2005: 153: 767–773.
7. Schmuth M, Fluhr J W, Crumrine D C et al. Structural
and functional consequences of loricrin mutations in
human loricrin keratoderma (Vohwinkel syndrome
with ichthyosis). J Invest Dermatol 2004: 122: 909–922.
8. Tomita Y, Akiyama M, Shimizu H. Stratum corneum
hydration and flexibility are useful parameters to
indicate clinical severity of congenital ichthyosis. Exp
Dermatol 2005: 14: 619–624.
9. Menon G K, Feingold K R, Mao-Qiang M, Schaude M,
Elias P M. Structural basis for the barrier abnormality
following inhibition of HMG CoA reductase in murine
epidermis. J Invest Dermatol 1992: 98: 209–219.
490
10. Mao-Qiang M, Feingold K R, Elias P M. Inhibition of
cholesterol and sphingolipid synthesis causes paradox-
ical effects on permeability barrier homeostasis. J Invest
Dermatol 1993: 101: 185–190.
11. Mao-Qiang M, Feingold K R, Jain M, Elias P M.
Extracellular processing of phospholipids is required
for permeability barrier homeostasis. J Lipid Res 1995:
36: 1925–1935.
12. Holleran W M, Man M Q, Gao W N, Menon G K,
Elias P M, Feingold K R. Sphingolipids are required for
mammalian epidermal barrier function. Inhibition of
sphingolipid synthesis delays barrier recovery after acute
perturbation. J Clin Invest 1991: 88: 1338–1345.
13. Holleran W M, Takagi Y, Menon G K, Legler G,
Feingold K R, Elias P M. Processing of epidermal
glucosylceramides is required for optimal mammalian
cutaneous permeability barrier function. J Clin Invest
1993: 91: 1656–1664.
14. Sekkat N, Kalia Y N, Guy R H. Porcine ear skin as a
model for the assessment of transdermal drug delivery
to premature neonates. Pharm Res 2004: 21: 1390–1397.
15. Sekkat N, Kalia Y N, Guy R H. Development of an in
vitro model for premature neonatal skin: biophysical
characterization using transepidermal water loss. J Pharm
Sci 2004: 93: 2936–2940.
16. Chilcott R P, Dalton Ch, Emmanuel A J, Allen C E,
Bradley S T. Transepidermal water loss does not correlate
with skin barrier function in vitro. J Invest Dermatol
2002: 118: 871–875.
17. Rogiers V. EEMCO guidance for the assessment of
transepidermal water loss in cosmetic sciences. Skin
Pharmacol Appl Skin Physiol 2001: 14: 117–128.
18. Nilsson G E. Measurement of water exchange through
skin. Med Biol Eng Comput 1977: 15: 209–218.
19. Pinnagoda J, Tupker R A, Agner T, Serup J. Guidelines
for transepidermal water loss (TEWL) measurement. A
report from the Standardization Group of the European
Society of Contact Dermatitis. Contact Dermatitis
1990: 22: 164–178.
20. Nuutinen J, Alanen E, Autio P, Lahtinen M R,
Harvima I, Lahtinen T. A closed unventilated chamber
for the measurement of transepidermal water loss. Skin
Res Technol 2003: 9: 85–89.
21. Nuutinen J, Harvima I, Lahtinen M R, Lahtinen T.
Water loss through the lip, nail, eyelid skin, scalp skin
and axillary skin measured with a closed-chamber eva-
poration principle. Br J Dermatol 2003: 148: 839–841.
22. Tagami H, Kobayashi H, Kikuchi K. A portable device
using a closed chamber system for measuring transepi-
dermal water loss: comparison with the conventional
method. Skin Res Technol 2002: 8: 7–12.
23. Grubauer G, Feingold K R, Harris R M, Elias P M.
Lipid content and lipid type as determinants of the
epidermal permeability barrier. J Lipid Res 1989: 30:
89–96.
24. Orsmark K, Wilson D, Maibach H. In vivo transepider-
mal water loss and epidermal occlusive hydration in
newborn infants: anatomical region variation. Acta
Derm Venereol 1980: 60: 403–407.
25. Elias P M, Feingold K R. Coordinate regulation of
epidermal differentiation and barrier homeostasis.
Skin Pharmacol Appl Skin Physiol 2001: 14 (Suppl. 1):
28–34.
26. Chuong C M, Nickoloff B J, Elias P M et al. What is
the ‘true’ function of skin? Exp Dermatol 2002: 11:
159–187.
 Transepidermal water loss as measure of permeability barrier
27. Bettley F R, Grice K A. A method for measuring the
transepidermal water loss and a means of inactivating
sweat glands. Br J Dermatol 1965: 77: 627–638.
28. Bikle D D, Chang S, Crumrine D et al. Mice lacking
25OHD 1alpha-hydroxylase demonstrate decreased
epidermal differentiation and barrier function. J Steroid
Biochem Mol Biol 2004: 89–90: 347–353.
29. Owada Y, Suzuki I, Noda T, Kondo H. Analysis on the
phenotype of E-FABP-gene knockout mice. Mol Cell
Biochem 2002: 239: 83–86.
30. Jensen J M, Schutze S, Neumann C, Proksch E.
Impaired cutaneous permeability barrier function, skin
hydration, and sphingomyelinase activity in keratin 10
deficient mice. J Invest Dermatol 2000: 115: 708–713.
31. Fluhr J W, Mao-Qiang M, Brown B E et al. Functional
consequences of a neutral pH in neonatal rat stratum
corneum. J Invest Dermatol 2004: 123: 140–151.
32. Zettersten E, Man M Q, Sato J et al. Recessive x-linked
ichthyosis: role of cholesterol-sulfate accumulation in
the barrier abnormality. J Invest Dermatol 1998: 111:
784–790.
33. Schmuth M, Man M Q, Weber F et al. Permeability
barrier disorder in Niemann-Pick disease: sphingomye-
lin-ceramide processing required for normal barrier
homeostasis. J Invest Dermatol 2000: 115: 459–466.
34. Schmuth M, Yosipovitch G, Williams M L et al.
Pathogenesis of the permeability barrier abnormality
in epidermolytic hyperkeratosis. J Invest Dermatol 2001:
117: 837–847.
35. Kao J S, Fluhr J W, Man M Q et al. Short-term
glucocorticoid treatment compromises both permeabil-
ity barrier homeostasis and stratum corneum integrity:
inhibition of epidermal lipid synthesis accounts for
functional abnormalities. J Invest Dermatol 2003: 120:
456–464.
36. Ghadially R, Reed J T, Elias P M. Stratum corneum
structure and function correlates with phenotype in
psoriasis. J Invest Dermatol 1996: 107: 558–564.
37. Ghadially R, Brown B E, Sequeira-Martin S M,
Feingold K R, Elias P M. The aged epidermal perme-
ability barrier. Structural, functional, and lipid bio-
chemical abnormalities in humans and a senescent
murine model. J Clin Invest 1995: 95: 2281–2290.
38. Moskowitz D G, Fowler A J, Heyman M B et al.
Pathophysiologic basis for growth failure in children
with ichthyosis: an evaluation of cutaneous ultra-
structure, epidermal permeability barrier function, and
energy expenditure. J Pediatr 2004: 145: 82–92.
39. Sugarman J L, Fluhr J W, Fowler A J, Bruckner T,
Diepgen T L, Williams M L. The objective severity
assessment of atopic dermatitis score: an objective
measure using permeability barrier function and stra-
tum corneum hydration with computer-assisted esti-
mates for extent of disease. Arch Dermatol 2003: 139:
1417–1422.
40. Chamlin S L, Kao J, Frieden I J et al. Ceramide-
dominant barrier repair lipids alleviate childhood atopic
dermatitis: changes in barrier function provide a sensi-
tive indicator of disease activity. J Am Acad Dermatol
2002: 47: 198–208.
41. Hachem J P, De Paepe K, Sterckx G, Kaufman L,
Rogiers V, Roseeuw D. Evaluation of biophysical and
clinical parameters of skin barrier function among hos-
pital workers. Contact Dermatitis 2002: 46: 220–223.
42. Agner T, Serup J. Skin reactions to irritants assessed by
non-invasive bioengineering methods. Contact
Dermatitis 1989: 20: 352–359.
43. Elsner P, Burg G. Irritant reactivity is a better risk
marker for nickel sensitization than atopy. Acta Derm
Venereol 1993: 73: 214–216.
44. John S M, Uter W, Schwanitz H J. Relevance of multi-
parametric skin bioengineering in a prospectively-
followed cohort of junior hairdressers. Contact
Dermatitis 2000: 43: 161–168.
45. Fluhr J W, Pfisterer S, Gloor M. Direct comparison of
skin physiology in children and adults with bioengineer-
ing methods. Pediatr Dermatol 2000: 17: 436–439.
46. Lavrijsen A P, Oestmann E, Hermans J, Bodde H E,
Vermeer B J, Ponec M. Barrier function parameters in
various keratinization disorders: transepidermal water
loss and vascular response to hexyl nicotinate. Br J
Dermatol 1993: 129: 547–553.
47. Lehmann L, Keipert S, Gloor M. Effects of microemul-
sions on the stratum corneum and hydrocortisone pene-
tration. Eur J Pharm Biopharm 2001: 52: 129–136.
48. Alberti I, Kalia Y N, Naik A, Bonny J D, Guy Rh. In
vivo assessment of enhanced topical delivery of terbina-
fine to human stratum corneum. J Control Release
2001: 71: 319–327.
49. Choi E H, Lee S H, Ahn S K, Hwang S M. The
pretreatment effect of chemical skin penetration enhan-
cers in transdermal drug delivery using iontophoresis.
Skin Pharmacol Appl Skin Physiol 1999: 12: 326–335.
50. Benfeldt E, Serup J, Menne T. Effect of barrier pertur-
bation on cutaneous salicylic acid penetration in human
skin: in vivo pharmacokinetics using microdialysis and
non–invasive quantification of barrier function. Br J
Dermatol 1999: 140: 739–748.
51. Halkier-Sorensen L, Thestrup-Pedersen K. The efficacy
of a moisturizer (Locobase) among cleaners and kitchen
assistants during everyday exposure to water and deter-
gents. Contact Dermatitis 1993: 29: 266–271.
52. Loden M, Andersson A C. Effect of topically applied
lipids on surfactant-irritated skin. Br J Dermatol 1996:
134: 215–220.
53. Hill S, Edwards C. A comparison of the effects of bath
additives on the barrier function of skin in normal
volunteer subjects. J Dermatolog Treat 2002: 13: 15–18.
54. Fluhr J W, Gloor M, Bettinger J, Gehring W. On the
influence of bath oils with different solvent characteris-
tics and different amounts of a non-ionic tenside on the
hydration and barrier function of the stratum corneum.
J Cosm Sci 1998: 49: 343–350.
55. Proksch E, Feingold K R, Elias P M. Epidermal HMG
CoA reductase activity in essential fatty acid deficiency:
barrier requirements rather than eicosanoid generation
regulate cholesterol synthesis. J Invest Dermatol 1992:
99: 216–220.
56. Proksch E, Feingold K R, Man M Q, Elias P M. Barrier
function regulates epidermal DNA synthesis. J Clin
Invest 1991: 87: 1668–1673.
57. Grubauer G, Feingold K R, Elias P M. Relationship of
epidermal lipogenesis to cutaneous barrier function.
J Lipid Res 1987: 28: 746–752.
58. Williams M L, Feingold K R, Grubauer G, Elias P M.
Ichthyosis induced by cholesterol-lowering drugs.
Implications for epidermal cholesterol homeostasis.
Arch Dermatol 1987: 123: 1535–1538.
59. Grubauer G, Elias P M, Feingold K R. Transepidermal
water loss: the signal for recovery of barrier structure
and function. J Lipid Res 1989: 30: 323–333.
60. Proksch E, Jensen J M, Elias P M. Skin lipids and
epidermal differentiation in atopic dermatitis. Clin
Dermatol 2003: 21: 134–144.
491
Fluhr et al.
61. Proksch E, Elias P M, Feingold K R. Regulation of
3-hydroxy-3-methylglutaryl-coenzyme A reductase activ-
ity in murine epidermis. Modulation of enzyme content
and activation state by barrier requirements. J Clin
Invest 1990: 85: 874–882.
62. De Paepe K, Houben E, Adam R, Wiesemann F,
Rogiers V. Validation of the VapoMeter, a closed
unventilated chamber system to assess transepidermal
water loss vs. the open chamber Tewameter. Skin Res
Technol 2005: 11: 61–69.
63. Barel A O, Clarys P. Study of the stratum corneum
barrier function by transepidermal water loss measure-
ments: comparison between two commercial instru-
ments. Evaporimeter Tewameter Skin Pharmacol 1995:
8: 186–195.
492
64. Grove G L, Grove M J, Zerweck C, Pierce E.
Comparative metrology of the evaporimeter and the
DermaLab TEWL probe. Skin Res Technol 1999: 5: 1–8.
65. Pinnagoda J, Tupker R A, Coenraads P J, Nater J P.
Comparability and reproducibility of the results of
water loss measurements: a study of 4 evaporimeters.
Contact Dermatitis 1989: 20: 241–246.
66. Scott R C, Oliver G J, Dugard Ph, Singh H J.
A comparison of techniques for the measurement of
transepidermal water loss. Arch Dermatol Res 1982:
274: 57–64.
67. Schulz A, Elsner P, Burg G. Quantification of irritant
contact dermatitis in vivo: comparison of the Dermatest
system with the evaporimeter. Contact Dermatitis 1991:
24: 235–237.