112 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS RESULTS AND DISCUSSION METHODOLOGY Calcium chloride was chosen as it is the most convenient desiccant available for study. Five individual measurements were made of the amount of anhydrous calcium chloride placed in the receptor compartment of the cell. The average weight recorded was 0.24 + 0.016 g. Cartilage-stripped hamster ear membrane is a new source for obtaining a fresh sample of full-thickness skin membrane. Percutaneous absorption theory states that the flux is inversely proportional to skin thickness (10). In actuality, we are referring to the thick- ness of the stratum corneum, as this is the rate-determining skin layer. Measured skin thickness was 0. 181 --- 0.003 mm. Since little or no variation in skin thickness was observed, it was assumed to be a constant when measuring TEWL. TIME STUDY Two blank and six experimental cells were prepared and placed into the heated alu- minum holding block at various temperatures. The receptor compartment containing the desiccant was exchanged for a fresh sample at the following time intervals: 0,0.5,1,2,3,4,5,6,7,8, and 20 hours. The TEWL flux was determined by calculating the cumulative counts at the particular time interval. A plot of the rate of TEWL-vs-time is shown in Figure 2. It was observed that at 0.5 hour and one hour, the rate increased rapidly from 340 Ixg/cm2/hr to 460 Ixg/cm2/hr, and then the TEWL decreased after two hours. Equilibrium was reached between the third and sixth hour. The apparatus operated in the presence of a dry atmosphere (less than 25 percent relative humidity), determined by a cobalt chloride paper technique 500 400 500 0 1 2 3 4 5 6 7 8 20 Time (hours) Figure 2. TEWL rate vs time. Bars refer to standard error of measurement.

METHOD FOR TRANSEPIDERMAL WATER LOSS 113 adapted from Solomon (36). Since the apparatus operates in the presence of a dry atmo- sphere, this transit time period results in an initial evaporation of water from the mem- brane. Initially, the membrane is fully hydrated, swollen, and is more permeable. This provides an "increased push" effect to pass water through the membrane towards the anhydrous desiccant. This transit time period is designated the "burst" effect. At steady state, the amount of water adsorbed onto the desiccant is in equilibrium with the amount above the membrane. Also, the membrane is approaching a drier state and is less permeable. Equilibrium TEWL rate after six hours was approximately 300 Ixg/cm2/hr at 32øC. Table I summarizes previously reported TEWL values based upon the various in vitro techniques currently employed for measurement. Blank (1,2) utilized full-thickness skin membrane from the human abdomen in his studies and reported TEWL values of between 130 and 270 Ixg/cm2/hr (23øC, 23% RH). Our value of 152 Ixg/cm2/hr (22øC, 20-25% RH) for the hamster skin is in excellent agreement with Blank's value for human skin. TEMPERATURE STUDY A plot of TEWL rate-vs-temperature is shown in Figure 3. An exponential increase in TEWL was observed, which implies that water vapor permeates by a diffusional pro- cess. The TEWL rate between 5 and 12øC, however, showed little change. Table I Summary of Transepidermal Water Loss (TEWL) Rate Data for Various In Vitro Skin Models Investi- gator TEWL Temper- Refer- Membrane rate ature Year ence used Method (•g/cm2/hr) øC Relative humidity in % Blank Mali Mali Reiger, Deem Wu Blank et al. This work 1952 (1,2) Human, full- Gravimetric 100-200 23 thickness skin 130- 270 23 from abdomen 400-800 35 1956 (8) Human, epider- Gravimetric, 500-600 21 mis from trunk desiccator 1956 (8) Human, epider- Gravimetric, 3000 21 mis from sole desiccator 1974 (30) Human, stratum Gravimetric 250- 300 21 corneum from abdomen 1983 (19) Fetal hog Saturated 251 21 periderm salt solutions 130 21 85 21 1984 (33) Human, stratum Saturated salt 600-1200 30 comeurn solutions and tritiated water 1987 Hamster ear, Tritiated water 250-300 32 full-thickness 152 22 skin 40-50 20-30 57 0 0 0-20 5O 4O 3O 0-80 20-25 20-25