RANKING OF SURFACTANT-HUMECTANT SYSTEMS 605 grader was not made aware of the contents of the patches containing aqueous contacting solutions i, ii, iv, and v. One should note that the range, 0-2, included all the visual skin dryness scores reported by the expert grader (see Table I). Although these scores corre­ spond to the mild range, the expert grader was nevertheless able to effectively discrimi­ nate between the observed low levels of visual skin dryness. Measurement of skin erythema using a chromameter. Skin erythema was measured instrumen­ tally by a Minolta CR-200 chromameter that is based on a standardized reflectance technique using a tristimulus system (18). The tristimulus system makes use of color reading values on three independent axes: (i) L * axis, reflecting the tone of lightness/ darkness, with higher values indicating lightness and lower values indicating darkness (ii) a* red/ green axis, reflecting the extent of redness/ greenness, with higher values indicating more red tone and lower values indicating less red tone and (iii) b* blue/ yellow axis, reflecting the extent of blueness/yellowness, with positive values indicating a yellowish tinge and negative values indicating a bluish tinge. Specifically, the color reading values were translated into the L *a *b* coordinates whose spacing correlates closely with color changes perceived by the human eye. For the evaluation of skin erythema using a chromameter, only values along the a* red/green axis that can capture the extent of redness (erythema) of the skin were considered. Sets of three a* readings from each of the volar forearm test sites were taken at baseline, as well as on Day 2 (approximately 18-20 hours after patch removal), and the average a* value was calculated for each site. Increased a* values along the red/ green axis, relative to the baseline measurements, indicate that the patch containing the aqueous contacting solution has induced skin redness (erythema) (18). Additional details can be found in references 18 and 19. THEORETICAL DEVELOPMENT OF AN IN VITRO TEST TO RANK AQUEOUS SURFACTANT-HUMECTANT SYSTEMS The central objective for developing an in vitro ranking metric is to rank aqueous surfactant-humectant contacting solutions i, ii, iv, and v, relative to the in vitro PBS control (iii), based on the extent of their perturbation to the skin aqueous pores. For this purpose, we chose skin electrical current induced by aqueous contacting solutions i-v, relative to the in vitro PBS control (contacting solution iii), as the preferred in vitro ranking test. Specifically, we adopted the following in vitro ranking metric (RM): IE RM=­ Ic (1) where E denotes the enhancer (that is, aqueous contacting solutions i, ii, iv, and v), C denotes the control (that is, the in vitro PBS control (iii)), IE denotes the skin electrical current induced by EJ and l e denotes the skin electrical current induced by C. Note that RM in equation 1 corresponds to the enhancement in the skin electrical current. The in vitro skin electrical current measurements ( 6,41) show that these measurements are: (a) extremely sensitive to small changes in the extent of skin barrier perturbation, (b) highly reproducible, and (c) simpler to implement, less time-consuming, and safer than typical skin permeability measurements, which make use of radioactive materials

606 JOURNAL OF COSMETIC SCIENCE and involve relatively complex assaying procedures. In addition to benefits a -above, we will show that the in vitro skin electrical current measurements correlate well with the in vivo skin barrier measurements reported here (see the Results and Discussion section). As shown previously, skin electrical current measurements can also be related to average skin electrical resistivity, R, values, which, in turn, can be combined with mannitol skin permeability, P, values, in the context of the hindered-transport aqueous porous pathway model (5,6,9,41). Specifically, by analyzing Log Pas a function of Log R, two important characteristics of the skin aqueous pores can be obtained: (a) the average pore radius, r pore' and (b) the porosity-to-tortuosity ratio, elT. In order to determine if aqueous contacting solutions i-v induce skin barrier perturbation by increasing the average pore radius and/or the pore number density (number of pores/unit area) of the aqueous pores in the SC, it is important to consider the relationship between r pore and e/T. Specifically, � - (Np)(7rr ore ) ( � )- ( Np ) _ 'ITr2ore p - 7rpr2ore p T a total T atota!T (2) where NP is the number of aqueous pores contained within a SC cross-sectional area of atota/J and p is the number of tortuous pores/unit area = (N/atota!T) = pore number density. Interestingly, equation 2 shows that e/T increases linearly with p and quadrati­ cally with r pore · Because aqueous contacting solutions i-v may modify either r pore and/or p, an analysis of the ranking metric to obtain mechanistic insight on the extent of perturbation of the skin aqueous pores should incorporate changes in both r pore and p. Once r pore and e/T are determined using the theoretical analysis involving the Log P and Log R values (5,6,9,41), equation 2 can be used to obtain p. The skin permeability (P) of a hydrophilic permeant, such as mannitol, can be modeled by considering transport of the hydrophilic permeant through the skin aqueous pores (5-9,41). Specifically, this results in the well-known relationship between P and the aqueous pore characteristics, given by references 5-9 and 41: (3) where v is the permeant (p) infinite-dilution diffusion coefficient, L is the thickness of the SC, and HCA ;, ) is the hindrance factor experienced by permeant pas it partitions into the SC from the aqueous contacting solution and diffuses across the SC The hindrance factor, H(A ;, ), is a nonlinear function of"- ;, , where "- ;, is the ratio of the permeant radius, r p , and the average pore radius, r pore ' that is, "- p = r/r por e (5-10,41). By combining equations 2 and 3, it follows that P is a function of both p and r JJure for a specific hydrophilic permeant, such as mannitol. Specifically, (4) Because P is inversely proportional to {RH(l\.)IH(l\. p )J, 5 and R is inversely proportional to the skin electrical current, /, P is directly proportional to {IH(l\. 1 )/H(l\.)} (see refer­ ences 5-9 and 41, as well as equation 5 below). Therefore, the ranking metric adopted 5 Note that A. = r/r pore ' where i represents the current-carrying ion (5).