Quantitative evaluation of skin barrier function using water evaporation time related to transepidermal water loss

Abstract

Background

Transepidermal water loss (TEWL) is often used as an index for skin barrier function. The skin barrier tester, SBT‐100 (Rousette Strategy Inc), measures the TEWL, water evaporation time, and time constant by contacting the skin and diffusing water into the closing measurement chamber. However, the relationship between the TEWL and time constant has not been sufficiently investigated. This study involved analyzing the underlying measurement principle and obtaining data through two experiments.

Materials and methods

The TEWL and time constant were measured using SBT‐100. Experiment 1 produced a simple simulation model for continuous water evaporation from the skin using a moisture‐permeable film. In experiment 2, four skin sites of 43 healthy volunteers were examined from May to September 2018.

Results

In experiment 1, the TEWL increased and time constant decreased, following an increase in humidity in the external environment. Both parameters demonstrated significant negative correlation (drying: ρ = −0.832, p < 0.001). For the 43 healthy volunteers who participated in experiment 2, their TEWL increased and time constant decreased in summer. For all skin measurement sites, both data demonstrated significant negative correlation (forehead: ρ = −0.909, p < 0.001; back of the left hand: ρ = −0.829, p < 0.001; left lateral elbow: ρ = −0.896, p < 0.001; left lateral malleolus: ρ = −0.865, p < 0.001).

Conclusion

Results indicated that the time constant is significantly correlated with TEWL. Furthermore, the time constant can be used as a parameter for evaluating skin barrier function.

1. INTRODUCTION

The most important function of the skin is to form an effectual barrier between the “interior” and “exterior” of bodies. ¹ , ² This physical barrier primarily consists of the stratum corneum (SC), which is the thin outermost layer (3–20 μm thick) of the skin. ¹ It acts as the primary barrier against transcutaneous penetration of harmful substances and microbes while being capable of tolerating mechanical forces. ¹ , ³ Further, the SC regulates water evaporation from the interior of bodies, known as transepidermal water loss (TEWL). ¹ Regulation of the TEWL by the SC depends on several factors, such as the level of intercellular lamellar lipids, diffusion path length, natural moisturizing factor, and filaggrin. ² TEWL is the amount of water that passively evaporates through the skin into the external environment and is used to characterize the skin barrier function. Although the barrier function of the SC is within the normal range, previous studies reported that the TEWL could vary extensively with skin surface temperature, ⁴ , ⁵ relative humidity, ⁶ , ⁷ and circadian rhythms. ⁸ , ⁹ The rate of water loss through the skin in healthy subjects showed significant differences by region of the body. ¹⁰ , ¹¹ Atopic dermatitis (AD) is a condition caused, among others, due to epidermal dysfunction. ¹² , ¹³ Barrier failure in AD is evidenced by enhanced TEWL and an increase in viral and fungal skin infections compared with their occurrence in normal skin. ¹⁴ , ¹⁵

Many devices have been developed to measure the TEWL for assessing the skin barrier function, both in the clinical and research domains. These devices are either opening or closing types. The Tewameter (Courage + Khazaka, Cologne, Germany) is one of the opening type devices. It measures water loss due to diffusion from the skin surface to the atmosphere in real‐time using Fick's law. ¹⁶ , ¹⁷ However, air regurgitates from the top of the chamber in the opening type devices. The VapoMeter (Delfin Technologies, Kuopio, Finland) is one of the closing‐type devices. It calculates water loss using the change in humidity of the covered chamber. ¹⁶ , ¹⁷ Compared with the opening type devices, closing‐type devices have little influence on air regurgitation and airflow due to body movement. However, the closing‐type devices are easily influenced by the humidity in the covered chamber, and they must be calibrated with an initial humidity value. A past study reported that the VapoMeter provided significantly lower TEWL readings at the forearm compared with the Tewameter. ¹⁸ Conversely, other research group reported that the Tewameter was more sensitive or discriminative to skin barrier damage and able to detect significantly smaller differences than the VapoMeter. ¹⁹

This study aims to evaluate a skin barrier tester, SBT‐100, developed by Rousette Strategy Inc. This device is a closing‐type attachment probe that measures water loss from the skin surface. In particular, the characteristic function of SBT‐100 is to calculate the water loss time, time constant, and evaporation into the closing chamber using a simple mathematical model based on a primary delay system. We have considered the time constant to be one of the indexes for skin barrier function due to its dependency on skin resistances. However, the relationship between TEWL and the time constant and measurement accuracy are yet to be investigated in‐depth. This study evaluates the skin barrier tester, SBT‐100, by measuring the moisture transpiration in the simple simulated model and the healthy subjects; the study is conducted in collaboration with the developers of the device. Through this study, we expect this device to contribute to evaluating skin barrier function in clinical dermatology and the cosmetic field.

2. METHODS

2.1. Measurement system

Figure 1 shows the overview of the measurement system and specifications of the probe tip. The measurement system consists of SBT‐100 and a personal computer with the installed analysis software developed by Rousette Strategy Inc. The outer diameter and height of the SBT‐100 probe are 25 and 100 mm, respectively, and the probe comes in direct contact with the skin. A microcomputer plate (PIC16F1829, Microchip Technology Inc., Chandler, USA) and USB to serial communication interface plate (AE‐FT231X, Akizuki Denshi Tsusho Co., Ltd., Tokyo, Japan) are placed inside this probe. As shown in Figure 1B, the temperature–humidity sensor (SHT21, Sensirion AG, Stäfa, Switzerland) and noncontact infrared temperature sensor (MLX90615, Melexis NV, Ieper, Belgium) are in two plastic chambers (diameter, 8 mm; depth, 6 mm) in the tip of the probe. The signals from both the sensors with a sampling time of 0.5 s are saved on the personal computer connected via the USB cable. The measuring and recording time per trial is approximately 20 s.

FIGURE 1.

Quantitative evaluation system and the measurement probe. (A) The measurement probe, SBT‐100, connects to a personal computer, and the tip of this probe comes in contact with the skin. The outer diameter and height of the SBT‐100 are 25 and 100 mm, respectively. (B) The temperature–humidity sensor and noncontact infrared temperature sensor are located in two plastic chambers (diameter, 8 mm; depth, 6 mm) at the tip of the SBT‐100.

2.2. Measurement principles for TEWL and time constant

The average TEWL is approximately 300–400 ml d⁻¹ and is called insensible perspiration. ¹ An assumption is that relative humidity under the skin is 100% as normal skin is filled with water. As shown in Figure 2A, this water is diffused based on the humidity difference between the measurement chamber and skin when the probe is brought in contact with the skin. Based on this assumption, the rate of water evaporation can be defined as follows:

$$Q=\left(W-H_a\right)/R$$ (1)

FIGURE 2.

Measurement principle of the SBT‐100. (A) Water diffuses due to the humidity difference between the skin and measurement chamber. Water evaporation, Q, is determined by the skin barrier resistance, R difference between skin humidity, W, and measurement chamber humidity, H. (B) The time constant, τ, is the time required for the relative humidity in the measurement chamber to rise exponentially through approximately 63.2% of the equilibrium state. τ is determined by the skin barrier resistance, R, the chamber volume, V, and the constant, α.

where Q is the rate of water evaporation (g m⁻² h⁻¹), W (%) is the relative humidity under the skin, H_a (%) is the initial value of relative humidity inside the measurement chamber, and R (%·g⁻¹ m² h) is the skin barrier resistance. When the measurement probe is brought in contact with the skin for time =  t  s, the accumulated water volume (F (g m⁻²)) can be defined as follows:

$$F=\int Q\left(t\right)dt$$ (2)

The amount of water vapor E (absolute humidity, g m⁻³) per 1 m² of the skin surface can be defined as follows:

$$E=F/V$$ (3)

where V is the volume of the measurement chamber (0.3 cm³).

Figure 2B illustrates the change in relative humidity in the measurement chamber when the probe is brought in contact with the skin. The curve is hypothesized to be a primary delay system. When the probe is in contact with the skin for t s, the relative humidity inside the chamber, H(t), can be defined as follows:

$$H\left(t\right)=W-\left(W-H_a\right){\times \mathrm{e}}^{-t/\tau }$$ (4)

Further, the saturated water vapor pressure and saturated water vapor volume in the measurement chamber can be calculated using the temperature in the chamber T(t), and the water vapor pressure in the measurement chamber can be calculated using H(t). H(t) is converted to E(t) using three values, as follows:

$$E\left(t\right)=W-\left(W-H_a\right){\times \mathrm{e}}^{-t/\tau }$$ (5)

Here, the TEWL is calculated by the following method: First, the saturated vapor pressure S_P in the measurement chamber is defined as follows:

$$S_P=6.1078\times {10}^{\frac{7.5\times T}{T+237.3}}$$ (6)

where T is the temperature (°C) inside the measurement chamber. Next, the saturated water vapor content S_V in the measurement chamber is defined as follows:

$$S_V=\frac{217\times S_P}{T+273.15}$$ (7)

Finally, the absolute humidity E (g m⁻³) is defined as follows:

$$E=S_V\times H_a/100$$ (8)

where H_a is the relative humidity (%) inside the measurement chamber. The sampling time of the sensor is 0.5 s, and E (t ₁) and E (t _1.5) can be calculated using Equations (6)–(8). The difference between E (t ₁) and E (t _1.5) can be defined as the TEWL over 0.5 s for the volume of the measurement chamber (V, 0.3 cm³). As this value is obtained under the conditions with a chamber cross‐section area A of 0.5 cm² and measurement time of 0.5 s, the TEWL (g m⁻² h⁻¹) is calculated by unit conversion.

A time constant, τ, is the time required for the relative humidity in the measurement chamber to rise exponentially to approximately 63.2% of its equilibrium state, as shown in Figure 2B. τ, commonly used in engineering, is calculated using the electrical resistance and capacitance of a circuit. Accordingly, we defined τ as follows:

$$\tau =R\times V\times \alpha$$ (9)

where V is the volume of the measurement chamber (0.3 cm³), R is the skin barrier resistance (%·g⁻¹ m² h), and α is a coefficient. Figure 2B shows the humidity change in the measurement chamber that is the secondary delay system. However, Equation (9) is based on a primary delay system. As the response time of the temperature–humidity sensor cannot be ignored, the time constant is calculated using the point of the slope with maximum humidity change in the measurement chamber and not the initial slope. Next, τ is determined by R and α as V is a constant. Furthermore, α is the coefficient that considers the response time in the temperature–humidity sensor and water diffusion coefficient, among other parameters.

2.3. Experiment 1

2.3.1. Simple experimental model

Figure 3 shows a simple experimental model that simulates continuous water loss from the skin surface. A small circular container (diameter = 42 mm, height = 32 mm) filled with hot water (volume = 10 ml, temperature = 80°C) was heated to 37°C using a hot plate. A partially closed space was produced by covering this container with a large circular container (diameter = 84 mm, height = 112 mm) having two air holes. A semipermeable waterproof film (Multi Fix, ALCARE Co., Ltd., Tokyo, Japan) was attached to the upper air holes of the large container. This film was used as a substitute for skin and transpired water vapor. The relative humidity of the partially closed space was almost 100% due to moisture generated in the small container. Temperature–humidity sensors (SHT31, Sensirion AG, Stäfa, Switzerland) were placed in the large container and the atmosphere. The former sensor was connected to the microcomputer (Arduino UNO R3, Arduino Holding) via the lower air holes to compute and record the temperature and relative humidity values.

FIGURE 3.

Experimental model simulating continuous water evaporation from the skin surface. A semipermeable waterproof film was used as a substitute to skin that transpired water vapor from the plastic container to the atmosphere. Water vapor was generated by having heating hot water (volume = 10 ml; temperature = 80°C) at 37°C by using a hot plate. The SBT‐100 probe was brought in contact with the waterproof film and measured the water vapor was measured. The temperature and relative humidity of the plastic container and atmosphere were measured using two temperature–humidity sensors.

2.3.2. Protocol in experimental 1

The semipermeable waterproof film and temperature–humidity sensor were attached to the large container. Then, the large container and the small container filled with water were placed on the hot plate. After 3 min of heating, we measured the TEWL and time constant from the film on the upper air hole using SBT‐100. Next, a second measurement was done after 6 min of heating. We measured the parameters at 3‐min intervals in this experiment until 15 min elapsed. Therefore, the data were obtained five times per cycle. The mean of these measurements was used for statistical data analysis. Additionally, the TEWL and time constant were measured under two conditions: (1) SBT‐100 was dried by using silica gel for 20 s (17 sets) each before measurement, and (2) SBT‐100 was used without drying (23 sets).

2.4. Experiment 2

2.4.1. Subjects and measurement sites

The subjects in this experiment were 43 healthy volunteer students in their 20s, 22 males, and 21 females. The measurement sites on the skin were the forehead, back of the left hand, left lateral elbow, and left lateral malleolus. Written informed consent for participation in this study was obtained from all the subjects. They had no visible skin diseases. The Ethics Committee of Kyorin University approved the study in advance (approval number: 29‐93).

2.4.2. Protocol in experimental 2

This experiment was performed monthly from May to September 2018 for each subject. Over 5 months, the temperature and relative humidity in the measurement room were 25–26°C and 50%–65%, respectively. After recording their basal information, such as identification number, sex, and age, before the experiments, the subjects rested for 10 min in the measurement room. After 10 min, the measurement probe of SBT‐100 was dried using silica gel for 20 s. Next, 4 s after starting the application installed on the personal computer, we brought the probe in contact with the back of the left hand. Following the previous procedure, measurement was performed for a total of three times at each site. The mean of two approximate values from the three measurements was used for statistical data analysis. Throughout the experiment, the measurement was taken in the order: (1) back of the left hand, (2) left lateral elbow, (3) forehead, and (4) left lateral malleolus for each subject.

2.5. Statistical analysis

Statistical analysis was performed using SPSS statistics version 26.0 for Windows (IBM Inc., Armonk, NY, USA). The obtained data were checked for normal distribution and homogeneity of data variances using the Shapiro–Wilk's test and Levene's test, respectively. In experiment 1, the data were analyzed using the Mann–Whitney test to check for the effect of the drying process on the measurement probe. Additionally, the influence of indoor humidity on measurement data was analyzed using Welch's analysis of variance (ANOVA). Consequently, when significant effects were detected, post hoc multiple comparisons were conducted using the Games–Howell method. In experiment 2, the data obtained over months were analyzed using the repeated‐measures ANOVA or Friedman's test in the case of normal distribution and non‐normal distribution, respectively. Consequently, when differences in the data obtained over months were detected, post hoc multiple comparisons were conducted using the Bonferroni method. The relationships between the TEWL and time constant in each experiment were assessed using Spearman's correlation coefficients. A p‐value of <0.05 was considered statistically significant.

3. RESULTS

For experiment 1, the temperature and relative humidity inside the simple experimental model was 30.09 ± 1.26°C and 97.59% ± 0.83%, respectively. Table 1 shows the TEWL and time constant values obtained using two experimental processes (drying and nondrying). The number of data values (N) obtained by the two processes was 81 and 110 for drying and nondrying, respectively. Consequently, both data were concluded to be of the non‐normal distribution type. The TEWL values were higher in drying than in nondrying, as calculated using the Mann–Whitney test (U = 1144.00, p < 0.001). The Mann–Whitney test revealed that the time constant values were higher in nondrying than in drying (U = 5215.50, p = 0.044). Figure 4 shows the relationship between the TEWL and time constant according to the experimental process. The two measurement values were negatively correlated with a statistical significance (drying process: ρ = −0.832, p < 0.001). In contrast, both the parameters were neither correlated not statistically significant (nondrying process: ρ = −0.099, p = 0.301). Figure 5 shows the TEWL and time constant of the drying process obtained according to indoor humidity. The indoor relative humidity was categorized into five groups: RH 1, less than 30% (N = 11); RH 2, 30%–40% (N = 19); RH 3, 40%–50% (N = 23); RH 4, 50%–60% (N = 21); RH 5, 60% or more (N = 7). The means of the TEWL in each category, RH 1–RH 5, were 2.49, 2.81, 3.68, 4.21, and 4.52 g m⁻² h⁻¹, respectively, as shown in Figure 5A. All values observed normal distribution (RH 1, p = 0.154; RH 2, p = 0.841; RH 3, p = 0.409; RH 4, p = 0.970) except for RH 5 (p = 0.042). However, Levene's test did not confirm the homogeneity of data variances (p = 0.026). Additionally, based on Welch's analysis, the significant primary effect of relative humidity was apparent in the TEWL values (p < 0.001). Multiple comparisons showed that the TEWL values were larger for high relative humidity conditions than low relative humidity conditions and confirmed the differences between the five groups (RH 1–RH 3, p < 0.001; RH 1–RH 4, p < 0.001; RH 1–RH 5, p = 0.001; RH 2–RH 3, p < 0.001; RH 2–RH 4, p < 0.001; RH 2–RH 5, p = 0.001). The mean time constants in each category, RH 1–RH 5, were 203.11, 159.22, 75.75, 49.06, and 37.00 s, respectively, as shown in Figure 5B. The time constant data observed normal distribution (RH 1, p = 0.657; RH 2, p = 0.221; RH 3, p = 0.949; RH 4, p = 0.325; RH 5, p = 0.422), whereas the homogeneity of data variances was not confirmed (p < 0.001). Based on Welch's analysis, the significant primary effect of relative humidity was apparent in the time constant values (p < 0.001). Multiple comparisons showed that the time constant values were smaller for high relative humidity conditions than for low relative humidity conditions and confirmed the differences among all groups (p < 0.001), including RH 4 and RH 5 (p = 0.007).

TABLE 1.

Comparison of the transepidermal water loss (TEWL) and time constant in the simple experimental model for cases with and without drying for SBT‐100

	TEWL (g m−2 h−1)		Time constant (s)
	Mean (±SD)	Significant difference	Mean (±SD)	Significant difference
With drying (N = 81)	3.52 (±0.90)	U = 1144.00 (p < 0.001)	102.36 (±60.87)	U = 5215.50 (p = 0.044)
Without drying (N = 110)	2.44 (±0.48)		119.20 (±61.57)

FIGURE 4.

Relationship between the transepidermal water loss (TEWL) and time constant with and without drying the probe with silica gel. The data are expressed as a scatter diagram with drying (●; N = 81) and without drying (○; N = 110). (A) The TEWL and time constant had a strong negative correlation and statistical significance (drying: ρ = −0.832, p < 0.001). (B) The two parameters were not correlated and not statistically significant (non‐drying: ρ = −0.099, p = 0.301).

FIGURE 5.

Comparison of (A) the transepidermal water loss (TEWL) and (B) time constant for each of the five indoor relative humidity categories (RH 1–RH 5) in the drying type data of experiment 1. Data are shown for each humidity category and expressed as box plots indicating the TEWL (gray) and time constant (white). The box limits are the 25th and 75th percentiles. The whiskers denote the minimum and maximum values, and the crosses indicate the mean values. The horizontal line within each box is the median value. (A). The mean of the TEWL in each category, RH 1 to RH 5, were larger for high relative humidity conditions than low relative humidity conditions. Statistical significance was confirmed among the five groups (RH 1–RH 3, p < 0.001; RH 1–RH 4, p < 0.001; RH 1–RH 5, p = 0.001; RH 2–RH 3, p < 0.001; RH 2–RH 4, p < 0.001; RH 2–RH 5, p = 0.001). (B). The mean time constants in each category, RH 1–RH 5, were smaller for high relative humidity conditions than for low relative humidity conditions and were confirmed to be statistically significant for all groups (p < 0.001), including RH 4 and RH 5 (p = 0.007). The statistically significant differences across the five categories were not indicated in both figures.

Table 2 shows the TEWL and time constant obtained for 43 healthy volunteer students according to measurement sites and months. For the forehead, from May to September, the means of the TEWL were 4.06, 4.32, 4.97, 5.21, and 4.92 g m⁻² h⁻¹, respectively; the mean time constant values were 121.66, 94.43, 77.30, 72.00, and 91.34 s, respectively. For the back of the left hand, the mean TEWL values were 3.76, 4.28, 4.44, 4.72, and 4.81 g m⁻² h⁻¹; the mean time constant values were 107.95, 79.83, 70.00, 61.71, and 79.21 s. For the left lateral elbow, the mean TEWL values were 2.64, 3.26, 3.71, 4.05, and 3.45 g m⁻² h⁻¹; the mean time constant values were 151.00, 100.45, 81.92, 74.49, and 108.26 s. For left lateral malleolus, the mean TEWL values were 3.23, 3.52, 3.89, 4.23, and 4.08 g m⁻² h⁻¹; the mean time constant values were 124.26, 92.50, 71.45, 65.20, and 89.91 s. Although both the measurement values differed at each measurement site, the TEWL had the lowest values in May and increased during the summer months. Conversely, the time constant data showed an opposite trend. Table 3 shows the statistical analysis of the monthly data in the subjects according to each measurement site. Consequently, monthly statistical differences were obtained for each measurement site.

TABLE 2.

Comparison of the transepidermal water loss (TEWL) and time constant in healthy subjects according to the measurement sites and months

		TEWL (g m−2 h−1)	Time constant (s)
		Mean (±SD)	Mean (±SD)
Forehead	May	4.06 (±1.61)	121.66 (±29.76)
	Jun.	4.32 (±1.49)	94.43 (±23.54)
	Jul.	4.97 (±2.61)	77.30 (±19.19)
	Aug.	5.21 (±1.46)	72.00 (±30.01)
	Sep.	4.92 (±1.94)	91.34 (±30.94)
Back of left hand	May	3.76 (±1.41)	107.95 (±31.42)
	Jun.	4.28 (±1.30)	79.83 (±24.85)
	Jul.	4.44 (±1.77)	70.00 (±22.61)
	Aug.	4.72 (±1.48)	61.71 (±20.86)
	Sep.	4.81 (±2.03)	79.21 (±33.89)
Left lateral elbow	May	2.64 (±0.55)	151.00 (±33.28)
	Jun.	3.26 (±0.73)	100.45 (±30.33)
	Jul.	3.71 (±0.82)	81.92 (±27.17)
	Aug.	4.05 (±1.18)	74.79 (±31.30)
	Sep.	3.45 (±0.97)	108.26 (±39.57)
Left lateral malleolus	May	3.23 (±1.48)	124.26 (±35.33)
	Jun.	3.52 (±0.95)	92.50 (±31.28)
	Jul.	3.89 (±0.91)	71.45 (±15.63)
	Aug.	4.23 (±1.31)	65.20 (±26.76)
	Sep.	4.08 (±1.68)	89.91 (±37.74)

TABLE 3.

Statistical analysis of the monthly data in healthy subjects

		TEWL (g m−2 h−1)				Time constant (s)
		Jun.	Jul.	Aug.	Sep.	Jun.	Jul.	Aug.	Sep.
Forehead	May	1.000	0.002	<0.001	0.011	0.012	<0.001	<0.001	<0.001
	Jun.		0.052	0.001	0.204		0.037	0.001	1.000
	Jul.			1.000	1.000			1.000	0.820
	Aug.				0.882				0.052
Back of left hand	May	0.052	0.007	<0.001	0.007	<0.001	<0.001	<0.001	0.001
	Jun.		1.000	1.000	1.000		0.316	0.004	1.000
	Jul.			1.000	1.000			0.399	0.577
	Aug.				1.000				0.040
Left lateral elbow	May	0.027	<0.001	<0.001	0.013	<0.001	<0.001	<0.001	0.002
	Jun.		0.267	0.223	1.000		0.120	0.007	1.000
	Jul.			1.000	0.442			1.000	1.000
	Aug.				0.375				0.127
Left lateral malleolus	May	0.442	0.002	0.001	0.011	0.017	<0.001	<0.001	0.001
	Jun.		0.820	0.562	1.000		0.064	0.011	1.000
	Jul.			1.000	1.000			1.000	0.656
	Aug.				1.000				0.170

Abbreviation: TEWL, transepidermal water loss.

Figure 6 shows the relationship between the TEWL and time constant from May to September according to the measurement sites. Accordingly, the TEWL and time constant for all four locations, the forehead (ρ = −0.909, p < 0.001), back of the left hand (ρ = −0.829, p < 0.001), left lateral elbow (ρ = −0.896, p < 0.001), and left lateral malleolus (ρ = −0.865, p < 0.001), have strong negative correlation. As a result of regression analysis, all the four locations showed negative regression curves, and the coefficient of determination R ² was more than 0.6: forehead (R ² = 0.736, p < 0.001); back of the left hand (R ² = 0.696, p < 0.001); left lateral elbow (R ² = 0.758, p < 0.001); and left lateral malleolus (R ² = 0.646, p < 0.001).

FIGURE 6.

Relationship between the transepidermal water loss (TEWL) and time constant according to each measurement sites. The monthly data are expressed as a scatter diagram: blue plot for May, orange plot for June, gray plot for July, yellow plot for August, and green plot for September. For all measurement sites, the TEWL and time constant strong negative correlation: (A) forehead, ρ = −0.909, p < 0.001, (B) back of the left hand, ρ = −0.829, p < 0.001, (C) left lateral elbow, ρ = −0.869, p < 0.001, and (D) left lateral malleolus, ρ = −0.865, p < 0.001

Figure 7 shows the relationship between the time constant and estimated water evaporation resistance based on the obtained data in experiment 1. The data are expressed as a scatter diagram, and both the parameters show strong positive correlation and statistical significance (ρ = 0.926, p < 0.001).

FIGURE 7.

Relationship between the time constant and estimated water evaporation resistance according to the obtained data in experiment 1. The data are expressed as a scatter diagram, and both parameters show strong positive correlation (ρ = 0.926, p < 0.001). The curvilinear regression equation is R′ = 0.106τ + 2.486.

4. DISCUSSION

This study aimed to conduct two experiments using the skin barrier tester, SBT‐100, and to verify the usefulness of this device for evaluating skin barrier function based on the obtained data. We concluded that SBT‐100 must be dried before measurement. For healthy subjects, the TEWL values obtained from four measurement sites increased in summer compared with spring. Conversely, the time constant values decreased in summer than in spring. Additionally, the two measurement values showed a significant negative correlation by Spearman's correlations analysis.

Experiment 1 measured continuous water loss via the semipermeable film that imitated the skin surface. This experiment was conducted with and without drying SBT‐100. The obtained data were compared. Consequently, the TEWL and time constant values showed significant differences between the two processes (Table 1). Without drying, the TEWL value did not increase or decrease significantly regardless of the indoor relative humidity. Typically, TEWL is affected by the relative humidity of the external environment. ⁶ , ⁷ Thus, we considered that the true value of the TEWL is obtained by drying SBT‐100 before use. In particular, the initial humidity value in the SBT‐100 chamber is an important factor as it calculates the water loss and time based on the humidity difference between the subcutaneous regions and the chamber inside the measurement probe. Moreover, the obtained TEWL after drying showed a negative correlation to the time constant, unlike without drying (as shown in Figure 4). From these results, we conclude that SBT‐100 must be dried before measurement to obtain accurate data.

Figure 5 shows the obtained data from experiment 1 with the drying process, categorized by indoor humidity. With an increase in the indoor humidity, the TEWL increased, and the time constant decreased. When water evaporation increased, we assumed that humidity in the probe chamber of the device reached the equilibrium state rapidly, reducing the time constant. Consequently, the TEWL and time constant showed a strong negative correlation. Conversely, we assumed that a higher relative humidity difference between the subcutaneous region and indoor environment increased the obtained TEWL value. However, as the TEWL values were small for lower indoor humidity conditions than higher indoor humidity conditions, we conclude that the discrepancy between our hypothesis and experimental results is due to the humidity change in the measurement chamber. The results of our analysis showed that the mean initial relative humidity in the probe chamber from RH 1 to RH 5 were 40.95%, 45.71%, 62.22%, 68.50%, and 71.09%, respectively. Moreover, the mean rates of change of chamber humidity before and after measurement were 2.86%, 3.86%, 6.23%, 7.89%, and 9.94%, respectively. These results suggest that the silica gel used significantly affects the measurement chamber under higher indoor humidity and causes the water to evaporate easily. However, as humidity in the measurement probe depends on the indoor environment, this problem cannot be ignored. Thus, the TEWL value may be influenced by the relationship between the silica gel and indoor humidity, and to obtain more accurate data, the measurement probe should be dried before the experiment.

Considering Equation (9), as α is a constant, τ is assumed to be proportional to R. Using Equation (1), R is calculated using the humidity difference between the skin and external environment, and TEWL, as follows:

$$R^{\prime }=\beta \times \left(W-H_a\right)/\mathrm{TEWL}$$ (10)

where R′ is the resistance of water evaporation (%·g⁻¹ m² h), that is, the skin barrier resistance, and β is a coefficient. Although the humidity response of the model is theoretically synonymous with a high‐order delay system, we considered it to be the response of a simple primary delay system. To investigate the validity of our consideration, the linearity between the estimated R′ based on the results in experiment 1 and the obtained τ is analyzed. When β is assumed to be 1, Figure 7 shows a significant positive correlation between R′ and τ (ρ = 0.926, p < 0.001). Using both the parameters, a curvilinear regression equation was obtained as follows:

$$R^{\prime }=0.106\tau +2.486$$ (11)

This linear model has a coefficient of determination, R ², of 0.934 (p < 0.001). These results indicate that the τ hypothesized in this study has a strong linear relationship to the water evaporation resistance, that is, skin barrier resistance. Therefore, we consider that the time constant, τ, is a coefficient that indicates skin barrier function.

In experiment 2, the TEWL values at the four parts showed significant differences over several measurement months (as shown in Tables 2 and 3). Compared with a previous study, ¹⁰ the TEWL value at the forehead, back of the left hand, left lateral elbow, and left lateral malleolus was 3–4 times, 3 times, 2.5 times, and 1.6 times smaller, respectively. One of the causes for this difference may be the type of measurement chamber used. Many TEWL measuring devices have an opening measurement chamber, ¹⁷ whereas SBT‐100 has a closing measurement chamber. The closing‐type device calculates the TEWL based on the change in temperature and humidity in the chamber. This type device accumulates water vapor in the chamber over time. Hence, the humidity difference between the skin and closed chamber is small. Consequently, water evaporation from the skin is reduced. In particular, as the developed SBT‐100 is a small chamber, we expect that the humidity inside the chamber is easy to modify. Accordingly, we believe that the obtained TEWL values in SBT‐100 were lower as compared with past studies.

The previous study reported that both the readings obtained by the closing and opening measurement chamber device for the TEWL of the forearm were not equivalent. ¹⁰ Moreover, varied measurement data, despite the same type of measurement chamber, were also reported. ¹⁰ Hence, comparing and evaluating TEWL measurement data would be difficult because of variations due to the measurement method and external environment. Table 2 shows that the TEWL in healthy subjects was higher in the summer season (primarily August) than in the other seasons. Typically, the TEWL decreases due to a drop in the water vapor pressure gradient. ¹ Similar to experiment 1, this result may have been triggered by the difference in the drying effect of silica gel. In addition, past studies reported that the TEWL was higher in the summer season. ²⁰ , ²¹ This may be because more water transpired to inhibit the rise in body temperature due to the high temperature and humidity in the summer season. Moreover, sweat being included in the TEWL values in comparison with other seasons is very likely. However, other studies reported that the TEWL was higher in the fall or winter seasons. ²² , ²³ Therefore, interpreting the TEWL measurement data is difficult as many variables affect the TEWL value (e.g., age, ¹¹ skin temperature, ⁴ circadian rhythms, ⁹ and skin condition ¹² , ¹⁴ ). Therefore, we concluded that TEWL should be evaluated as a relative value by comparing the obtained data before and after the experiment or therapy rather than analyzing it as an absolute value.

This study proposed the time constant (skin barrier index) as an assessment index for skin barrier functions. The time constant was calculated from the point with the most humidity change in the measurement chamber. As shown in Figure 2B, the humidity change in the measurement chamber is similar to a secondary delay system as the response time of the temperature–humidity sensor used is slow. However, this study considered the time constant as a primary delay system to develop a simple measurement principle. Here, the constant α in Figure 2B can be affected by the measurement data but cannot be controlled by the system. Sensor delay and moisture diffusion are considered α. The SHT21 used is a compact temperature–humidity sensor with excellent versatility. Typically, a sensor responds to input signals with a delay. This delay, known as the response time or response speed, refers to the time taken to reach approximately 63% of the maximum value of the output signal. The response time of SHT21 is approximately 8 s, and we believe that the actual measured value includes an error due to the sensor delay. The water vapor diffuses in the gas phase at 2.56 × 10⁻⁵ m² s⁻¹ (1 atm, 25°C), ²⁴ and this coefficient depends on temperature and pressure. In the closing‐type device, the environment inside the chamber varies with water evaporation. Hence, the moisture diffusion coefficient during measurement can vary. Thus, the TEWL was calculated by using a simple formula comprising of the temperature and relative humidity inside the chamber. Additionally, as water vapor accumulates in the chamber over time, the humidity difference between the skin and chamber is small, and evaporation from the skin is reduced. Accordingly, we believe that the TEWL may be lower.

As a result of this study, the TEWL and time constant show a strong negative correlation (Spearman's correlations, ρ ≥ −0.83). Thus, the time constant is expected to be useful as a new evaluation index for skin barrier function. We believe that SBT‐100 provides information not only for clinical dermatology, including AD, itch, dry skin, and so on but also for cosmetics. For example, in the above diseases associated with itching, scratching the skin can be associated as a response to itching reducing the barrier function. Using the proposed device in this study, we can provide quantitative information regarding skin barrier damage. For cosmetics, we expect that the SBT‐100 can evaluate the effect of newly developed moisturizers on dry skin. In addition, we can provide useful skin care information for various skin issues, including prolonged mask‐wearing, using this device.

We recognize that this study has a few limitations. First, this device cannot distinguish between insensible perspiration and sweating. Thus, to inhibit the influence of sweating, a TEWL measurement must be performed after the subject has rested well. Second, experiment 1 simulated continuous water loss from the skin surface using one semipermeable waterproof surgical film. However, human skin has a layered structure that adjusts water evaporation according to the external environment, unlike the artificial film. Hence, whether or not the simulated data arein accordance with the subject data is unclear. Third, all the subjects were Kyorin University students in their twenties. As a result of the experiments, a significant correlation between the TEWL and time constant was reported. We expect the time constant to be a new evaluation index for the skin barrier. However, TEWL has been shown to change with age. ¹¹ To observe the relationship the TEWL and time constant with age, we require subjects over a wide age range. Fourth, this study standardized the closed chamber in SBT‐100 using silica gel, based on the development of the sweat device. ²⁵ , ²⁶ Consequently, the TEWL and time constant showed statistical significance. However, concluding that silica gel did not affect the measurement data in this study is difficult. In the future, we recommend reverification of chamber standardization.

5. CONCLUSIONS

This study created a simulation model for continuous water evaporation and measured the TEWL and time constant using the skin barrier tester, SBT‐100. Concomitantly, we performed experiments using 43 healthy volunteer subjects using this device. The result of experiments showed that the TEWL and time constant were higher and lower, respectively, in a high humidity environment than in a low humidity environment. A strong negative correlation was observed between both data. SBT‐100 has some issues regarding TEWL and time constant calculation, such as the influence of sweating and the drying process using silica gel. However, the compactness and practicality of this device should contribute to the evaluation of skin barrier function in clinical dermatology and the cosmetic field.

CONFLICTS OF INTEREST

The authors declare that there are no conflicts of interest that could be perceived as prejudicing the impartiality of the research.

Seno S, Shimazu H, Kobayashi H, Kogure E, Watanabe A, Isoyama T. Quantitative evaluation of skin barrier function using water evaporation time related to transepidermal water loss. Skin Res Technol. 2023;29:1–11. 10.1111/srt.13242

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author on reasonable request.