182 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS 0.030 0.025 0.020- • o. o15 u. 0.010 0.005 0.000 I I I I I 0 1 2 3 4 5 Time (Weeks) Figure 5. Effect of storage temperature on the permeation of salicylic acid through micro-Yucatan pig skin (skin dermatomed before storage). ß 4øC, ¸ -15øC. * Statistically significant value from the initial result. Each error bar represents the standard error of the mean. 0.01 ___ 9.7 x 10 -4 mg/cm2/hour, respectively. The average value of the flux ratios was 269.84 ___ 23.63. Based on the values of the standard error of the means obtained, it is clear that the variability among pigs for the rates obtained was very limited. DISCUSSION The flux ratio reinforced in many instances the results obtained by the individual rates. At the two-week time point for skin stored before dermatoming, the individual flux values for both markers at both storage temperatures did not differ significantly from the baseline value (Figures 2, 3). However, the flux ratios for these data were significantly different from the reference values (Table I). On the other hand, at three weeks under the same conditions, there were significant differences in individual flux values but not in the ratios. Therefore, the most rigorous test for barrier integrity is to require that both the individual marker flux values and their ratios remain unchanged from the initial values. Storing skin at 4øC after dermatoming appears to be the most appropriate and conve- nient method of storage. In all cases, the flux did not change significantly over time from the initial values. One exception was noticed when the flux value of SA was significantly

ASSESSING SKIN BARRIER INTEGRITY 183 • 5 • 4 I I I I I 0 1 2 3 4 5 Time (Weeks) Figure 6. Effect of storage temperature on the permeation of water through micro-Yucatan pig skin (skin dermatomed and dried before storage). ß 4øC, O -15øC. Each error bar represents the standard error of the mean. higher after one week and then dropped down for the remainder of the study. This time point could be a result of an experimental error. Rosenquist et al. (4) observed that the viability of pig skin remained unchanged for 30 days when stored at 4øC. Although viability is not directly related to permeability, their results show a pattern consistent with the findings of our study. Skin stored before dermatoming at 4øC for more than two weeks changed appearance and odor. Such a change was not noticed in skin stored at -15øC. This change in physical appearance was reflected in the barrier properties of the skin. This is the only storage condition where the -15øC station, though not completely satisfactory, pre- served the skin better than the 4øC station. It was observed that the permeability of pig skin to water and SA did not change when it was measured from the back and left and right hams of the pig. In contrast, abdominal rat skin was more permeable to water and benzoic acid than back skin (7). Storing the dermatomed skin at - 15 øC changed the barrier properties of the skin. This was probably due to freezing of the water in the SC. This explanation was confirmed by