114 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS 500 400 lOO I 0 5 10 15 20 25 :50 :55 40 Temperature øC. Figure 3. TEWL rate vs temperature at steady-state conditions (after 3 hours). Bars refer to standard error of measurement. If the reciprocal temperatures in degrees Kelvin are plotted vs the natural logarithm of the TEWL rate in the temperature ranges between 12 and 37øC, linearity is observed (Figure 4). It is difficult to explain the lack of linearity at 5øC. The water permeation may have been disrupted at this near-freezing temperature. The energy of activation was calculated by slope analysis. Our calculated value of 13.1 Kcal/mole compares favorably with an activation energy value of 15 Kcal/mole previously reported by Scheuplein for human skin (9). MOISTURIZER STUDY Four experimental cells and two blank cells were prepared for each moisturizing agent tested. Ten •L of each agent was applied to the top of the membrane surface, prior to attachment of the receptor cell filled with desiccant and placement of the unit into the 32øC heating block. Plots of TEWL-vs-time values for membranes treated with various potential moisturizers are shown in Figures 5 and 6. The figures show individual TEWL values vs time over a 24-hour time period for the untreated sample and skin treated with agents. The untreated sample shows a result similar to that reported in Figure 2. Transepidermal water loss reaches steady-state conditions after three hours, except for that sample treated with 25% glycerin. When compared to the untreated sample, castor oil is the only agent that exhibited a "burst" effect. Skin treated with sesame oil, mineral oil, and 25% glycerin in water all appear to inhibit this effect. Sato and Nagai (37) compared the effects of different

METHOD FOR TRANSEPIDERMAL WATER LOSS 115 O-• 5.0 4.0 m m mm •mmm m• ecipr0cl of Kelvin Temperature X10 z Figure 4. Natural logarithm of TEWL vs reciprocal of absolute temperature x 1000. emollient materials on TEWL using a silastic membrane. There was a good correlation between the TEWL rate for this in vitro model and the TEWL rate for an in vivo human skin model. They found that the decrease of TEWL rate that resulted from emollient application was inversely proportional to the polarity of the material applied. In other words, the more polar the material applied to the membrane, the greater the TEWL rate. Mineral oil is the least polar substance that we tested and it inhibited TEWL the most. At steady state, the effect of castor-oil-treated skin on TEWL (Figure 6) was similar to that of sesame-oil-treated skin (Figure 5). Glycerin, which acts to bind water to the skin, produced an interesting effect. During the first two hours glycerin acted as an occlusive agent, since TEWL values were close to that of mineral-oil-treated skin. After four hours, however, the glycerin-tissue com- posite was observed to become saturated with water. At the six-hour time point, TEWL rates were almost double those of untreated skin. This is comparable to the results obtained by Reiger and Deem (30) who found an increase in the TEWL rate upon application of 25% glycerin in water. Reiger and Deem showed that hurnectants, in general, increase TEWL rate at low relative humidities.