126 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS length, width, and thickness, and that the greatest response is in thickness. The stratum corneum also responded to changes in relative humidity by swelling or shrinking in the three dimensions. We postulated that under large changes in tempera- ture and humidity, the horny layer develops a dimensional shrinkage-swelling gradient therefore, as it is flexed, the resultant shearing stresses can crack the stratum corneum. The above studies suggest that the ability of the stratum corneum to absorb water may be involved in determining the ultimate condition of the skin and in mechanisms of ingredient effects on skin. We previously showed that for a homologous alkyl sulfate series (C 8 to C•6) , maximal swelling is produced by the C12 homologue (2). That the C•2 homologue of alkyl sulfates causes maximal swelling is similar to known in vivo irritancy of alkyl sulfates (the C•2 alkyl sulfate is also the most irritating) (5). This suggests swelling of stratum corneum may be an important part of the mechanism of surfactant-induced irritation. In the present study, we examine in more detail the relationship of stratum corneum swelling to the structure of various surfactants. Additionally, we explore the effects of mixing surfactants on the swelling response. From the findings, we then discuss the most likely mechanism(s) of surfactant-induced swelling. MATERIALS AND METHODS Surfactants used were reagent grade or were in some cases specially prepared to yield the highest purity available. The surfactants used are displayed in Table I. Human abdom- inal stratum comeurn was isolated using the method of Kligman and Christophers (6). Vertical sections of male and female (25-80 years of age) autopsied, abdominal skin samples were used directly and subjected to heat treatment by immersion in water at 55øC for 30 sec. The stratum corneum was rubbed off with a teflon spatula. The mem- brane was floated over water, spread out, picked up on a wire mesh screen, and air dried. The membranes were sectioned and swelling was measured in the crosswise length, all as described previously (2). The membranes were incubated at room temperature over- night in distilled water and the hydrated lengths determined. The membrane pieces were randomized with respect to location within the abdominal strip (taken from one individual). Six membranes were placed together in one treatment solution at 40øC and their lengths determined at the various incubation times shown in the Results. Up to eight treatments could be accommodated in one experiment. All surfactant solutions (except the competitive bar soap and tallow soap solution) were adjusted to pH 7 using hydrochloric acid or sodium hydroxide prior to incubation. Swelling values are ex- pressed as percentage increase above the hydrated length (length of the membrane after soaking overnight in distilled water). Unless otherwise specified, concentrations are on a weight/weight basis. Statistical analysis of the data was done using Analyses of Covariance (7) with the covariate measurement being the hydration response prior to treatments. This was done to increase the precision of these experiments because biological variation is prevalent using stratum corneum. When the F value in the Analyses of Covariance was signifi- cant, least significance testing was employed to identify those means that differ signifi- cantly. A 2% difference in the treatment means was usually significant at the p • 0.05 level.

SURFACTANT SWELLING OF STRATUM CORNEUM 127 Table I List of Surfactants and Suppliers Surfactant Description and/ or Abbreviations Suppliers Alkylbenzene Sulfonates *Linear Alkyl (C•2 average) Benzene Sulfonate Alkyl Sulfates Alkyl Trimethylammonium bromides *Alkyl (C12-C14 average) Ethoxy Sulfates Paraffin Sulfonates Alpha Olefin Sulfonates Homologous series based on alkyl chain length LAS Homologous series based on alkyl chain length Homologous series based on alkyl chain length Homologous series based on average number of moles of ethylene oxide present, eg AEOS-3EO, AEOS-6EO, etc.* Homologous series based on an average alkyl chain length* Homologous series based on an average alkyl chain length* Coco Amidopropyl Betaine Betaine Lauryl Dimethyl Amino Oxide LDAO Magnesium Lauryl Sulfate - Triethanolammonium TEA lauryl sulfate Lauryl Sulfate Sodium Lauryl Sulfate SLS Dodecyl Trimethylammonium DTAB Bromide Tallow Soap - Tween 20 Nonionic Alfonic 1012 - 6 Nonionic Tergitol 25-L-7 Nonionic Conoco (Vista), Saddlebrook, NJ and Monsanto, St. Louis, MO Conoco (Vista), Saddlebrook, NJ Eastman Kodak, Rochester, NY and Sigma Chem. Co., St. Louis, MO Eastman Kodak, Rochester, NY and Sigma Chem. Co., St. Louis, MO Conoco (Vista), Saddlebrook, NJ Hoechst, Somerville, NJ Stepan, Northfield, IL Lonza, Fairlawn, NJ Onyx Chem. Co., Jersey City, NJ Henkel, Maplewood, NJ Henkel, Maplewood, NJ Sigma Chem. Co., St. Louis, MO Eastman Kodak, Rochester, NY Colgate-Palmolive, Piscataway NJ ICI, Wilmington, DE Conoco (Vista), Saddlebrook, NJ Union Carbide Co., Danbury, CT * Average means that in addition to the homologue specified, some species of the higher and lower homo- logues are present. RESULTS CONCENTRATION DEPENDENCE AND REVERSIBILITY OF STRATUM CORNEUM SWELLING Crosswise swelling of stratum corneum by SLS was assessed in response to concentration (Figure 1). The swelling response was concentration-dependent and appeared to saturate