614 (i) 1 wt% SDS JOURNAL OF COSMETIC SCIENCE (iii) In Vivo Control (iv) 10 wt% PG (v) 10 wt% G Figure 5. Skin erythema scores, with the error bars corresponding to standard errors, which were measured using a Minolta chromameter upon contacting human skin in vivo with aqueous contacting solutions i, ii, iv, and v, and for the in vivo control (iii). of the SC and disorder them, thereby increasing water loss from the skin, which in turn, results in skin dryness. The deviation from baseline in chromameter a* values, denoted hereafter as Lla*, as described in the Experimental section, are reported in Figure 5. Note that skin treated with 1 wt% SDS shows a positive deviation from the baseline, which indicates that SDS is indeed harsh to the skin and induces skin erythema (redness) (13-16,21-24,29,37). On the other hand, the Lla* value of skin treated with 1 wt% C 12 E6, is not statistically different (p 0.05) from zero, which indicates that C 12 E6 , is indeed mild to the skin and does not induce skin erythema (21-24). 13 In fact, a Student t-test did indicate a significant difference (p 0.05) between the Lla* values of skin treated with 1 wt% SDS and those of skin treated with 1 wt% C 12 E 6 . One 13 It is important to note that a mild surfactant is defined in the skin care literature as one that does not induce erythema (21-24). On the other hand, a harsh surfactant is defined as one that induces erythema (13-16,19-22,24-29,37). It is also important to keep in mind that a mild surfactant such as C 12 E6 induces skin barrier perturbation because it induces skin dryness by disordering intercellular lipid bilayers in the SC (see the in vivo TEWL measurements in Figure 3 and the visual skin dryness scores in Figure 4).

RANKING OF SURFACTANT-HUMECTANT SYSTEMS 615 should stress that the TEWL measurements and the visual skin dryness scores determine the extent of intercellular lipid bilayer perturbation in the SC, while the chromameter measurements of erythema (skin redness) determine the extent of denaturing of the keratins in the corneocyte/keratinocyte domains of the SC and the VE (viable epidermis), which leads to an increase in the cutaneous blood flow and associated skin redness (13-16,19-22,24-29,37). Therefore, the results in Figure 5 indicate that SDS induces a larger extent of erythema relative to C 12 E6, because SDS can interact strongly with the keratins of the corneocytes and the keratinocytes relative to C 12 E 6 . As a result, because SDS can interact with both the keratins and the intercellular lipid bilayers in the SC, while C 12 E6 can interact only with the intercellular lipid bilayers in the SC, SDS can induce a larger extent of skin barrier perturbation relative to C 12 E 6 ( 13-16,21-2 5 ,29). The in vitro skin barrier measurements discussed in the Experimental section corroborate this finding. In Figure 5, the bars corresponding to the Aa* values of the in vivo control (iii) and of the two aqueous humectant solutions (iv and v) are below the zero level, which indicates that no skin erythema was observed in these three cases. In addition, no statistical difference (p 0.05) was observed between the Aa* values of the in vivo control (iii), which is an untreated skin test site, and 10 wt% PG. However, a statistically significant difference (p 0.05) was observed between untreated skin and skin exposed to 10 wt% G. These results indicate that glycerol has a stronger beneficial impact on the skin barrier relative to propylene glycol, which is consistent with the in vitro ranking metric results discussed above (30-34,36). RELATIONSHIP BETWEEN THE IN VITRO AND IN VIVO SKIN BARRIER PERTURBATIONS INDUCED BY AQUEOUS SURFACTANT-HUMECTANT SYSTEMS The in vitro ranking metric (RM) analysis and the mannitol skin permeability measure­ ments indicate that the two surfactant, aqueous contacting solutions (i and ii) induce a significantly larger extent of skin barrier perturbation relative to the in vitro PBS control (iii) and relative to the two humectant aqueous contacting solutions (iv and v). The in vivo TEWL measurements, the visual skin dryness scores, and the chromameter mea­ surements of in vivo skin erythema (redness) are consistent with the in vitro results (see Figures 3-5). The in vitro ranking metric analysis also predicts that aqueous contacting solution i, containing 1 wt% SDS, induces the largest extent of skin barrier perturbation in vitro. Conversely, aqueous contacting solution v, containing 10 wt% G, was predicted to induce the smallest extent of skin barrier perturbation in vitro. A similar result was obtained from the chromameter measurements of in vivo skin erythema (see Figure 5). Note that the TEWL values (see Figure 3) and the visual skin dryness scores (see Figure 4) corresponding to 1 wt% C12E6 were found not to be statistically different (p 0.05) from those corresponding to 1 wt% SDS. In addition, the in vitro RM analysis and the mannitol skin permeability measurements indicate that aqueous contacting solution v, containing 10 wt% G, induces a smaller extent of skin barrier perturbation, and there­ fore better preserves the skin barrier relative to aqueous contacting solution iv, contain­ ing 10 wt% PG (see Table II). This result is consistent with the chromameter mea­ surements of in vivo skin erythema (see Figure 5). In view of the agreement between the in vitro ranking metric analysis and the in vivo skin barrier measurements summarized above, the methodology presented here to rank ague-