HAIR DAMAGE 43 Table I Effect of Post-Shampoo Combing on Protein Loss From Oriental Hair Turbidity (Optical density at 600 nm) Number of combing strokes Protein loss (mg/gm Hair*) Number of combing strokes Treatment 100 200 25 50 100 200 Shampoo A (2-in-1) Tress 1 0.178 0.315 0.243 0.376 0.824 1.614 Tress 2 0.189 0.371 0.219 0.464 0.949 2.133 Tress 3 0.199 0.442 0.215 0.461 1.058 2.149 Av. 0.189 0.376 0.226 + .01 0.433 + .03 0.944 + .07 1.965 + .17 Shampoo B (2-in-1) Tress 1 0.042 0.123 0.115 0.269 0.511 1.096 Tress 2 0.067 0.181 0.101 0.269 0.502 1.238 Tress 3 0.093 0.202 0.139 0.333 0.623 1.514 Av. 0.067 0.169 0.118 + .01 0.290 +- .02 0.545 -+ .04 1.283 -+ .12 * Three-gm tresses, prepared by using a commercial Oriental hair sample -- 10 inches long, were prewashed with 1 cc of 20% SLES, followed by shampooing with the test product @ 1 cc/tress. Each tress was then combed for a total of 200 combing strokes, and the abraded cuticular material was collected in a glass beaker containing 50 ml of distilled water. The turbid water solution was analyzed for protein concentration at the end of 25, 50, 100, and 200 combing strokes (see Materials and Methods section for details). The observed differences in protein loss for shampoo A vs shampoo B are significant at the p = 0.05 level for all combing strokes. shampoo B did also provide less protein loss. This is usually, but not always, in agreement with easier combing and less protein loss. However, these comparisons are not part of this paper for proprietary reasons. To further validate the test methodology, another experiment was conducted using Asian hair shampooed with three different shampoos identified as shampoos A, B, and C shampoos A and B were the same as tested above. These studies were conducted by a_different operator. As shown in Table II, protein loss from hair shampooed with shampoo B is significantly less as compared to hair shampooed with shampoo A or shampoo C. Additional studies showed that the results are reproducible from day to day, although the level of measured damage, that is absolute quantitative values for protein loss, may vary slightly from operator to operator. The above studies clearly support our hypothesis that protein loss measurements can provide meaningful results in studying hair damage during combing. PROTEIN LOSS FROM HAIR DURING COMBING AFTER TREATMENT WITH DIFFERENT SHAMPOOS AND CONDITIONERS To explore whether we could detect significant differences in protein loss due to comb- ing from hair shampooed with various shampoos, hair tresses (European hair) were shampooed with five different shampoos, identified as shampoos 1 through 5. Among the five shampoos tested, the first three were leading-brand commercial products,

44 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Table II Effect of Post-Shampoo Combing on Protein Loss From Asian Hair Turbidity (O.D. at 600 nm) Protein loss (mg/gm Hair @ 50 combing strokes*) Shampoo A (2-in-1) Tress 1 0.097 0.341 Tress 2 0.067 0.281 Tress 3 0.071 0.337 Av. 0.078 0.319 -+ .02 Shampoo B (2-in-1) Tress 1 0.007 0. 108 Tress 1 0.006 0.104 Tress 3 0.003 0.108 Av. 0.013 0.107 -4-_ .01 Shampoo C (2-in-1) Tress 1 0.054 0. 184 Tress 2 0.053 0. 191 Tress 3 0.062 0.191 Av. 0.056 0. 189 -4-_ .01 * For experimental details, see footnote in Table I. Observed differences in protein loss among the three shampoos tested are significant at the p = 0.05 level. whereas the last two were prototype formulations prepared in our laboratory. Shampoo 1 was a non-conditioning shampoo primarily containing a mixture of ALS/ALES (19% AI). It has been used for most of our studies as a standard shampoo to often compare results from study to study. All other shampoos tested were 2-in-1-type conditioning shampoos containing the following conditioning ingredients: Shampoo 1 None Shampoo 2 Dimethicone Shampoo 3 Dimethicone, guar hydroxypropyltrimonium chloride Shampoo 4 Dimethicone, guar hydroxypropyltrimonium chloride, and polyquater- nium 10 Shampoo 5 Dimethicone, guar hydroxypropyltrimonium chloride, polyquaternium 10, polyquaternium 6, and polyquaternium 16 After shampooing, each tress was combed for a total of 100 strokes, and the water solution was analyzed for protein concentration. Results of these experiments are sum- marized in Figure 1. As shown, total protein loss was the highest from hair shampooed with the non-conditioning shampoo (shampoo 1). It is not surprising to see that, among the five shampoos tested, hair shampooed with the non-conditioning shampoo, namely shampoo 1, experienced greater protein loss during combing compared to hair treated with the conditioning shampoos. Even among the various conditioning shampoos tested, significant differences (p = 0.05) in hair damage could be detected using this method, illustrating the sensitivity and the utility of the method. This work was further extended to study the effect of hair conditioners on protein loss during combing. In this experiment the protection efficacy of three leading-brand U.S. conditioners was compared to that of the conditioning shampoo 5, previously found to