SKIN CONDITION MEASURED BY SONIC VELOCITY 9 calculated elastic modulus changes are a lower limit of the actual decrease in outer layer skin elasticity the actual change in the modulus is greater than that calculated. Finally, lower limit elastic toodull obtained in this study may be compared to those determined by other workers using different techniques (18,19) as set out in Figure 5. In 12 10 8 Elasticity (Newton/m 2 X 10 ?) 6 4 2 0 _ ,._ .._ .._ .._ ,•,,• "' "' "' ' •.,.•,--, - 18-35 36-50 50 Age (Yrs.) Figure 5. Young's modulus of elasticity as a function of age and body site. A. Torso skin from Rollhauser (1950) B. Forearm skin from Grahame (1%9) C. Hand skin (this study). this figure, the effects of both age and body site are examined. The moduli calculated from this work for dorsal hand skin are consistent with the other investigators' results in that an age-related increase in elastic modulus (vide infra.) was observed as well as a body site effect. Comparing the three sets of observations, a lower elastic modulus was found in the more flexible skin areas (i.e., torso skin modulus forearm skin modu- lus hand skin modulus). Direct intercomparison of skin elasticity measured by different techniques must be viewed with caution, however. SONIC VELOCITY MEASUREMENTS IN THE CLINICAL ENVIRONMENT Having established the technique for making sonic velocity measurements and the probable physical basis for the changes in velocity, two examples of the use of the method in skin clinical studies will be detailed. These studies are examined, not to document the performance of competitive products (which will not be named), but rather to demonstrate the ability of sonic velocity measures to discriminate between treatments establish the sensitivity of the technique examine the effects of the clinical regimen and base size on sonic velocity results and explore the relationship between observed sonic velocity trends, the dermatologist's evaluations, and consumer percep- tion of treatments/products. A small base, acute product treatment forearm study and a large base, ad. lib. usage hand study will be considered. The data from these investigations will establish that sonic velocity maps differences between product treatments at a statistically significant level before visual differences among the treatments occur. It will also be shown that the early detection of elasticity changes calculated utilizing this technique are consistent with the final, longer term visual effects of products and the consumer

10 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS rankings of the treatments. Finally, evidence will be examined that establishes that visual dermatological grades and sonic velocity-derived elasticity measures are nearly orthogonal representations of skin condition and thus each measures unique attributes of the skin state. 1. FOREARM ACUTE PRODUCT TREATMENT STUDY This forearm test was designed to observe the time evolution of treatment effects from a single application of five marketed skin care products. Eight sites on the inner aspect of forearms of nine individuals were examined. Each forearm was abused three times daily with surfactant insults for one week prior to the test. At the end of the abuse period, each site was graded visually by a single expert grader using a 0 to 5 scale (0 representing perfect skin condition and 5 representing poor condition) and the sonic velocity measured in the standard manner (vide supra). Immediately after this initial grading and measurement, five products were applied to five sites per individual and three sites were left untreated as controls. Each site was then graded visually and instrumentally after five minutes and then again at thirty minute intervals up to three hours after product application. The results of this investigation are set out in Table I. In this table the product/sites have been assigned letter codes and untreated control sites are marked with an asterisk. Both the visual grades and sonic velocities are listed as a function of time after treatment, with bracketed values not being significantly different (oz .05). Inspection of the sonic velocity data at the top of the table reveals that, relative to the untreated control sites, all products caused a statistically significant drop in sonic velocity five minutes after application. This statistically significant difference from the control sites remained throughout the 180-minute test period for products F, B, and H. From the visual grading data at the bottom of the table, at five minutes after application, all products produced a statistically significant improvement relative to the untreated controls. This visual improvement then gradually diminished until, at 180 minutes after treatment, only product B was still significantly better than control. On comparison of the bracketed results in this table, statistically significant differences between active products were detectable instrumentally at five minutes after applica- tion and lasted throughout the test. Visually, however, active product differences were not detected until 150 minutes after application. The results after 180 minutes are consistent between the two evaluation methods, although the actual order of the product grades do differ both techniques indicate product B is significantly different from the control at this time. The sonic measures also identify products F and H as different from control at the conclusion of the test. Analysis of variance implemented utilizing the Statistical Analysis System ANOVA procedure (36) was applied to the observations of this study in order to quantitate differences between the visual and instrumental techniques. Changes in the calculated elastic modulus of each site from the pretreatment calculated modulus, and the change in the visual grades, were hypothesized to depend on which treatment was applied, the time after treatment, the subject, and second order interactions between time and treatment. This model was found to account for 52% of the variance in the calculated elasticity changes and 72% of the variance in the visual grades. The results of this analysis are set out in Table II.