skin compatibility by transepidermal water loss

j. Soc. Cosmet. Chem., 40, 33-39 (January/February 1989) Evaluation of hand cleansers: Assessment of composition, skin compatibility by transepidermal water loss measurements, and cleansing power RON A. TUPKER, JALIYA PINNAGODA, PIETER-JAN COENRAADS, HENK KERSTHOLT, and JOHAN P. NATER, Occupational Dermatology Unit, State University Hospital (R.A.T., J.P., P.-J.C., J.P.N.), and Inspection Service of Foods and Commodities (H. K. ), Groningen, The Netherlands. Received October 1 O, 1988. Synopsis The purpose of this study was to investigate the relationship among the composition, skin compatibility, and cleansing effectiveness of seven hand cleansers. The skin compatibility was assessed by a three-week multiple exposure model and evaluated by transepidermal water loss (TEWL) measurements and visual scoring. Three cleansers contained anionic detergents and no solvents. These products were more irritating to the skin in comparison with the other four cleansers which contained aliphatic solvents and no anionic detergents. In these four cleansers, a combination of good skin tolerance and effective washing action was found. Thus, it is concluded that cleansing effectiveness is not dependent on the presence of anionic detergents and does not necessarily exclude good skin compatibility. INTRODUCTION In many occupations, the hands are soiled by grime, oil, or other fat-soluble substances, which compels the use of powerful hand cleansers. The majority of these cleansers are generally composed of a surface active agent combined with a solvent and/or an abrasive (1). Hand cleansers are available in a liquid or gel-like form. Ideally, a hand cleanser should be both effective in cleansing and noninjurious to the skin. However, it is difficult to unify these two characteristics; skin cleansing products are a major factor in occupational irritant contact dermatitis (2). Several authors (3-6) have investigated skin cleansing products and their effects on human skin using a fiveday repeated exposure model and the visual scoring system according to Frosch et al. (7). Attempts have been made to relate the skin tolerance to the washing power of a cleanser (8). The purpose of this study was to test and compare seven commercially available hand cleansers on the basis of their composition, skin compatibility, and cleansing effectiveness. In order to better reflect the conditions in practice, the skin compatibility test in 33

34 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS this study is based on a three-week multiple-exposure model (9). To evaluate skin effects, transepidermal water loss (TEWL) measurements and visual scoring were used. MATERIALS AND METHODS COMPOSITION Of the many hand cleansers available on the European market, a selection was made of seven frequently used products. According to the manufacturers, all of the tested products belonged to the group of "heavy duty" cleansers, except brands A and C (see Table I). The consistence of the cleansers was gel-like, again with the exception of brands A and C which were liquid. Brand F was the only product which contained an abrasive. Table I shows the composition of each product. The percentages (g/100 g) of the following compounds were determined: soap [by means of the methodology prescribed by the Dutch law for food and commodities (10)], anionic detergents other than soap [by the methylene blue method (11)], nonionic detergents [qualitative: by gas chromatographic analysis (12); quantitative: by ion exchange analysis (13)], and water (gravimetric analysis). Aliphatic and/or aromatic hydrocarbons were determined by means of gas chromatography (14). Furthermore, the ph of the 5% solutions of the products was determined. SKIN COMPATIBILITY A group of 25 healthy volunteers (8 women and 17 men) without skin ailments, ranging in age from 19 to 41 years, participated in the study. The experiments were performed between February and April, 1988. The hand cleansers were diluted in water to a 5% solution. Each solution (0.3 ml) was applied to a disc of absorbent Whatman paper of 20-mm diameter and 1.0-mm thickness (no. 2589C, Schleicher & Schuell, The Netherlands). Each paper disc was fitted into a round, flat, plastic chamber (Agfa Bayer, Leverkusen, Germany) with an inside diameter of 22 mm. Four of these chambers were attached to a metal strip next to each other, and the strip was fixed to the skin of the volar side of each forearm with nylon clip bandages. The distance between the most distal chamber and the wrist joint was kept at 75 mm. Two applications lasting for one hour each were made on each working day during a period of three weeks, except the last day (a total of 28 applications). The mean time interval between Table I Hand Cleansers: Composition and ph Brand % Soap % Anionic % Nonionic % Solvent % Water ph A 0.2 12.4 2.2 0 78.7 7.5 B 6.8 0 7.5 26.7 56.0 8.6 C 0.4 4.5 2.2 0 87.5 6.6 D 3.8 O 21.0 41.2 32.1 8.8 E 5.8 0 4.3 43.4 43.3 8.4 F* 1.4 6.4 0 0 62.6 7.0 G 12.8 0 8.8 23.0 57.3 8.6 * Contains an a brasive.

EVALUATION OF HAND CLEANSERS 35 these two applications was three hours. After their removal, the skin was cleaned with running water and gently dried with a cotton towel. Every three weeks fresh solutions were made. The effects of exposure were evaluated by TEWL measurements and by a visual scoring system every Monday and Friday before the first application. TEWL measurements were performed with the ServoMed Evaporimeter EP1 (ServoMed AB, Vailingby, Stockholm, Sweden). The operating principle of this instrument is described in detail by Nilsson (15). To avoid errors caused by air currents, the measurements were performed in a box into which the subject had placed the forearm. These measurements were performed in an air-conditioned room at a constant temperature of 20øC and a relative humidity of 35 --- 5%. After an acclimatization period of five to ten minutes, the probe was applied perpendicular to the skin surface. For further details concerning the exact methodology of the measuring technique, see Pinnagoda et al. (16). Clinical changes at the exposure sites were graded according to Frosch et al. (7): i) erythema: slight redness, spotty or diffuse (1 + ); moderate redness (2 + ); intense redness (3 + ); fiery red with oedema (4+); ii) scaling: fine (1 +); moderate (2 +); severe with large flakes (3+); iii) fissures: fine cracks (1+); single or multiple broader fissures (2 +); wide cracks with hemorrhage or exudation (3 + ). CLEANSING EFFECTIVENESS According to the manufacturer's directions, brands A, B, C, D, and F have to be applied to the soiled hands and subsequently washed with running water, whereas the products E and G have to be wiped off with an absorbent paper or a cloth towel; thereafter, the hands are to be rinsed with water. Therefore, it was decided to perform two different kinds of cleansing tests for all products. Healthy volunteers (a total of 15, six women and nine men aged from 19 to 37 years) cooperated in both tests. Cleansing test 1 mimics the conditions prescribed by the manufacturers of brands A, B, C, D, and F and is a modification of the method developed by Kresken (17). The hands were soiled by greasing them for 30 seconds with 1.5 g of standard dirt of which the composition closely resembles occupational dirts (ModeHschmutz B, Stockhausen GmbH, Krefeld, Germany). The soiled hands were allowed to dry for 90 seconds. Subsequently, 0.5 g of the pure, undiluted hand cleanser was applied and rubbed in for 30 seconds. After this, the hands were washed with 1 ml water for 30 seconds, after which the hands were moistened again with 1 ml water and washed for 30 seconds. Thereupon, the hands were cleaned with running water and subsequently judged for cleansing effectiveness. The evaluation was based on the quantity of rest soil (17), graded from 0 to 5: 0, = clean;! = slightly stained; 2 = moderately stained; 3 = heavily stained; 4 = very heavily srained; 5 = no cleansing effect. Cleansing test 2 is based on the direcrions of use for brands E and G. After rubbing the soiled hands with the cleansers, the hands were wiped two times with an absorbent paper (Kleenex, Kimberly-Clark, Veenendaal, The Netherlands), after which they were rinsed by running water. For the evaluation, the same grading system was used as in cleansing test 1. STATI, STICAL ANALYSI;S The Student-tes,t for paired o,bservatio,ns was used to compare th, e means of the TEWL

36 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS values for the cleansers on the last day of the experiment. To compare the means of the rest soil scores, the Wilcoxon signed rank sum test for paired data was applied. RESULTS COMPOSITION It is striking that only three products (A, C, and F) contained anionic detergents (i.e., other than soap) and that the same products did not contain solvents, in contrast with the other cleansers, which did not contain anionic detergents but contained solvents (see Table I). The solvents in the products tested were all aliphatic hydrocarbons; aromatic hydrocarbons could not be detected. SKIN COMPATIBILITY In Table II the products are ranked in decreasing order of injurious influence on the skin, determined by the mean TEWL value on the last day of the experiment. On this day, the TEWL values for all products reached the highest score. Of all products, A and C caused the highest TEWL values. The results of the visual scoring system on the last day showed positive scores for two cleansers only, the same products which had high TEWL values (data not shown). In one person brand A caused a total visual score of 1 +, in three persons 2 +, and in one person a total score of 3 +; brand C caused in one person a total score of 1 +. CLEANSING EFFECTIVENESS Cleansing test 1: Apart from brands C and E, of which the mean rest soil scores were significantly higher than the scores of the other cleansers, the cleansing effectiveness of the products can be considered to be moderate to good (see Table III). Cleansing test 2: All seven cleansers caused clean hands (score 0) (not shown). Table Comparison of Irritancies as Evaluated by Transepidermal Water Loss (TEWL)(g/m 2 h) on Day 19 II Brand TEWL (g/m 2 h) A C F G E D B UNX* 1. A 15.8 s s s s s s s 2. C 13.1 ns s s s s s 3. F 12.2 s s s s s 4. G 9.6 ns ns s s 5. E 9.4 s s s 6. D 8.9 ns s 7. B 8.8 ns 8. UNX 8.! Paired t-test; s = significant (p < =0.05); ns = not significant. The hand cleansers are ranked in decreasing order of mean TEWL values. * UNX = unexposed skin.

EVALUATION OF HAND CLEANSERS 37 DISCUSSION SKIN COMPATIBILITY In a previous study, when studied separately, a soap, a nonionic detergent, and an aliphatic solvent appeared to be safer to the skin than most of the anionic detergents (9). Several other authors have also identified by visual scoring a striking difference in irritating potency between soap and "mild" anionic detergents on the one side, and the more widely used "irritating" anionics on the other side (3-6). However, Sauermann et al. observed a higher irritancy of soap compared with a (mild) anionic detergent using a 24-hour application (8). The results of the present study are in accord with those of the investigations mentioned above (3-6,9): the presence of anionic detergents, in combination with a soap, resulted in a stronger irritating influence, in contrast with nonionics in combination with soap. The aliphatic solvents in the cleansers tested apparently did not exert a damaging action either. This finding contradicts the results of Dobson (18): of the three hand cleansers tested, the one not containing a solvent had the least irritating effect on the skin. It was thus concluded that the presence of solvents renders the cleanser more injurious to the skin. However, he did not investigate the presence of the other agents in the products, nor did he determine the type of solvents. The injurious effect on the skin of a solvent is dependent on many characteristics: the type of solvent (aromatic or aliphatic hydrocarbon, alcohol, ester, ketone, etc.), the volatility, the boiling range, the viscosity, etc. (19). The aliphatic hydrocarbons are more friendly to the skin than the aromatics (2,19). For hand cleansers, use is made of petroleum-derived hydrocarbons, such as mineral oil or kerosene, themselves mixtures of different types of solvents (1,2). The solvents used in hand cleansers are mostly refined hydrocarbons, i.e., the percentage of aromatics has been lowered. Nevertheless, the degree of refinement is variable among the different solvent suppliers (19). The hand cleansers tested in this study all contained highly refined aliphatic hydrocarbons as solvents. Undiluted, kerosene is a potent skin irritant (19-21). In our model, the test concentration of the solvents was about 2 g/100 g or less, which may further explain the low irritating potency of the solvents tested. In multicomponent mixtures like hand cleansers, interactive forces between the constituents may be evoked (22), so it is difficult to demonstrate the contribution of a particular agent to the overall effect on the skin. However, for four hand cleansers, the combination of a soap, a nonionic detergent, and a solvent, each of which proved to be mild in a previous study (9), appeared Table III Hand Cleansers Ranked in Decreasing Order of Mean Rest Soil Score Rest soil Brand score E C A B G D F 1. E* 3.1 s s s s s s 2. C 2.1 s s s s s 3. A 1.5 ns ns s s 4. B 1.5 ns s s 5. G* 1.5 s s 6. D 0.8 s 7. F 0.7 Wilcoxon test for paired data; s = significant (p < =0.05); ns = not significant. A low rest soil score indicates high cleansing efficacy. * Rest soil score = 0 tested in cleansing efficacy test according to use directions (see Results).

38 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS to be equally harmless when tested together. Since this observation is only applicable to the products tested in the present study, it must be emphasized that this kind of observation may not be valid for other products. The difference in skin tolerance between these four cleansers and brands A, C, and F might as well be due to the presence of the so-called "skin protective agents," determination of which was not performed in this study. The fact that brand F was the only cleanser which contained an abrasive does not explain that it was more irritating than other products (Table II); only mechanical movement may induce skin damage, by a grating action. Comparison of skin compatibility of cleansers by means of occlusive patch tests may lead to misinterpretations. Particularly in occlusive conditions, in which the water cannot evaporate, it may act as a penetration enhancer for the detergents. Dahl noticed a more irritating effect of sodium lauryl sulphate in water applied under occlusion than in an open application (23). In order to avoid the above-mentioned sources of errors, one might propose to perform a test in which the skin is exposed to the products in the same manner as in daily practice, like the elbow wash test. Several investigators observed a high correlation between the elbow wash test and the occlusive patch test (4-6). The disadvantages of tests in which the hands or elbows are washed are the fact that only two products can be compared in one test and the fact that such a test method might also lead to false conclusions since it involves problems with standardization (7). CLEANSING EFFECTIVENESS For all hand cleansers, wiping the mixture of dirt and cleanser off with absorbent paper was more effective in removing the dirt than rinsing with water. Since the directions of use of five products recommend running water to remove the mixture of dirt and cleanser, the results of the cleansing test based on the method according to Kresken (17) (Table Ill) are still valid for these five products. The high cleansing power of brand F may be due to the fact that it contained an abrasive. Cleansers B, D, E, and G do not contain an anionic detergent. This finding contradicts the common opinion that an effective skin cleanser by preference should contain an anionic detergent (1). For cleansing bars, classical soap showed a higher cleansing power than an anionic detergent (8). Thus, also for cleansing effectiveness, the combination of a soap, a nonionic detergent, and a solvent offers a good alternative to compositions with anionic detergents. SKIN COMPATIBILITY AND CLEANSING EFFECTIVENESS When the results of both the skin compatibility test and the cleansing power test are combined, it becomes evident that both characteristics do not necessarily counteract each o,ther. In brands B, D, E, and G, it appears that there is coexistence of good skin tolerance and effective washing action. In the prevention of irritant contact dermatitis it is of great importance to choose a safe skin cleanser. This is particularly true for persons with a sensitive skin. The type of hand cleansers tested is of great practical use for mechanics, painters, road construction workers, etc., especially when conventional cleansing techniques with soap and water are difficult to. perfo,rm. It is strongly recommended to use hand cl, eansers instead o,f pure solvents to remove soi o.r remnants of paint (2'0). Finally, it should always be kept

EVALUATION OF HAND CLEANSERS 39 in mind that hand dermatitis may also be caused by contact allergens. Hand cleansers may contain several potential contact allergensuch as perfumes, pigments, and preser- vatives. ACKNOWLEDGMENTS This study was funded by the Directorate General of Labour (Ministry of Social Affairs), The Hague, and the Foundation Arbouw, Amsterdam, The Netherlands. REFERENCES (1) L. Chalmers, Chemical Specialties. Domestic and Industrial, 2nd ed. (Unwin Brothers Ltd., Old Woking, Surrey, England, 1979), Vol. 2, pp. 288-289. (2) R. M. Adams, Occupational Skin Disease (Grune& Stratton, Inc., New York, 1983), pp. 192-203. (3) P. G. M. v. d. Valk, Skin irritancy of commercially available soaps and detergents measured by water vapour loss, Dermatosen in Berufund Urnwelt, 32, 3:87-90 (1984). (4) P. T. Frosch, "Irritancy of Soaps and Detergent Bars," in Principles of Cosmetics for the Dermatologist, P. Frost and S. N. Horwitz, Eds. (C. V. Mosby Company, St. Louis, 1982), pp. 5-12. (5) M. Puschmann and J. Meyer-Rohn, Hautvertraeglichkeitsnachweis neuartiger Syndetpraeparate, Aertzliche Kosmetologie, 13, 225-234 (1983). (6) E. M. Koch and A.M. Kligman, Klinisch-experimentelle Untersuchungen zur Characteristik yon Seifen und Syndets, Pharmazeutische Zeitung, 128, 963-968 (1983). (7) P. J. Frosch and A.M. Kligman, The soap chamber test,j. Am. Acad. Dermatol., 1, 35-41 (1979). (8) G. Sauermann, A. Doerschner, U. Hoppe, and P. Wittern, Comparative study of skin care efficacy and in-use properties of soap and surfactant bars, J. Soc. Cosmet. Chem., 37, 309-327 (1986). (9) R. A. Tupker, J. Pinnagoda, P. J. Coenraads, and J. P. Nater, The influence of repeated exposure to surfactants on the human skin as determined by transepidermal water loss and visual scoring, Contact Dermatitis, in press. (10) Wasmiddelenbesluit, Warenwet, (1977). (11) G. F. Longman, The Analysis of Detergents and Detergent Products (John Wiley-Interscience, London, 1975), pp. 237-240. (12) B. M. Milwirdsky and D. M. Gabriels, Detergent Analysis (George Godwin, London, 1981), pp. 261-264. (13) B. M. Milwirdsky and D. M. Gabriels, Detergent Analysis (George Godwin, London, 1981), pp. 270-272. (14) H. Kerstholt and B. Raske, Determination of aliphatic and aromatic hydrocarbons in household chemical, De Ware(n) Chemicus, 17, 102-112 (1982). (15) G. E. Nilsson, Measurement oœ water exchange through the skin, Med. Biol. Eng. Cornput., 15, 209-218 (1977). (16) J. Pinnagoda, R. A. Tupker, P. J. Coenraads, andj. P. Nater, Comparability and reproducibility of the results of water loss measurements: A s.tudy of four evaporimeters, Contact Dermatitis, in press. (17) J. Kresken, Personal communication. (18) R. L. Dobson, Evaluation of hand cleansers, Contact Dermatitis, 5, 305-307 (1979). (19) J. V. Klauder and F. A. Brill, Correlation of boiling ranges of some petroleum solvents with irritant action on the skin, Arch. Dermatol., 56, 197-215 (1947). (20) L. Schwartz, L. Tulipan, and D. J. Birmingham, Occupational Diseases.of the Skin (Lea& Febiger, Philadelphia, 1957), pp. 611-624. (21) A. P. Lupolescu and D. J. Birmingham, Effect of protective agents against lipid solvent-induced damage, Arch. Environ. Health, 31, 33-36 (1976). (22) K. Miyazawa, M. Ogawa, and T. Mitsui, The physico-chemical propertl=es and protein denaturation potential of surfactant mixtures, Int.. Cosm. ScL, 6, 33-46 (1984). (23) M. V. Dahl and M. J. Roering, Sodium lau.ryl sulkfate irritant patch tests. ILl. Evaporation of aqueous vehicle influences inflammatory response.. J. Am. Acad. DermatoL, ll, 477-479 (1984).