j. Soc. Cosmet. Chem., 33, 179-191 (July, 1982) Statistical evaluation of vehicle effect on anti-perspirant activity with a limited number of subjects GEORGE E. OSBORNE and JOAN M. LAUSIER, Dept. of Pharmacy, WILLIAM D. LAWING, Dept. of Computer Sciences and Experimental Statistics, University of Rhode Island, Kingston, RI 02881,' MELANIE SMITH, Chesebrough-Pond's, Inc., Trumbull, CT 06611. Received November, 1979. Synopsis The effect of formulation factors (e.g., vehicle) on the efficacy of the antiperspirant agent, aluminum chlorhydrate, was studied. Aluminum chlorhydrate was solubilized in a cream base, an aqueous lotion, and a hydroalcoholic base, and was suspended in a solid stick. Aluminum chlorhydroxy allantoinate was incorporated into all vehicles for its therapeutic and cosmetic properties. A statistically useful experimental design to evaluate antiperspirant efficacy data with a limited number of subjects was developed. Four vehicles containing aluminum chlorhydrate were compared for efficacy by a well-recognized gravimetric procedure during a normal work day in each subject's normal work environment. A pretest sweat collection was used as a blocking factor in designing the 4 x 4, split plot, Latin Square. Post-test sweat collection data were evaluated statistically using the geometric mean of treatment results. All ratio treatment means were converted to percent sweat reduction to determine antiperspirant efficiency. Statistical analysis indicated that the side treated may be selected randomly without compromising the results. The carrier vehicle does not affect antiperspirant efficacy. However, efficacy differences were observed as a function of solution versus suspension. Treatment response in terms of percent sweat reduction was: aqueous lotion, 38.0% hydroalcoholic solution, 32.4% cream base, 31.8%, and solid stick, -6.2%. INTRODUCTION Most of the articles in the cosmetic literature are concerned with the efficacy of the antiperspirant agent or the methods of testing antiperspirant effectiveness (1-4). Studies of topically applied drugs have shown that drug efficiency is dependent on the carrier vehicle in which the drug is placed (5). Kligman (6) tested dimethyl sulfoxide (DMSO) as a carrier vehicle for a variety of topical drugs and observed that certain concentrations of DMSO potentiated the antiperspirant effect of strong aluminum chloride solutions. 179

180 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS O'Malley and Christian (2) used a continuous recording method to evaluate perspira- tion properties of six commercial antiperspirant agents and three commercial bases in the forms of a stick, a cream, and a lotion all were compared with a 24% w/w aluminum sulfate solution. Their tests indicated that aluminum methionate was the most effective agent, and that vehicle did not affect antiperspirant efficacy. Jungerman (1), using a gravimetric procedure, compared results of antiperspirant efficacy under normal activity and in a controlled thermal environment. He evaluated antiperspirant agents in various vehicles for antiperspirant and deodorant efficacy, safety, staining potential, and cosmetic acceptability. He found no significant difference among the active ingredients' effectiveness in the same formulation, but did notice a difference in activity among vehicles and concentrations. Recently, Bretschneider, et al. (7) tested varying aluminum to halide ratios, multi- ingredient formulations, varying concentrations of single-ingredient formulations, and aqueous versus anhydrous dose forms for efficacy. Their definitive results showed that aqueous vehicles increased effectiveness and that optimum concentration efficacy reached a maximum rather than a plateau. In an effort to determine how vehicular excipients and the active ingredient's state (e.g., suspension versus solution) in an antiperspirant formulation affect the efficacy of a product, a single, proven antiperspirant entity in different vehicles, an aqueous lotion, a cream base, a solid stick, and a hydroalcoholic base were formulated and evaluated gravimetrically for the influence of vehicle on antiperspirant efficiency. Because the aim of an antiperspirant is to control human axillary perspiration in all types of environments and aid situations, the gravimetric procedure's antiperspirant efficacy results resemble most closely consumer use of an antiperspirant product. Further, a gravimetric technique is recognized officially by the FDA-OTC Advisory Review Panel on Antiperspirant Drug Products as the method of measuring antiperspi- rant efficacy. This simple and inexpensive method provides quantitative results suitable for statistical analysis. The accepted antiperspirant effectiveness qualification test is a modification of Fredell and Read's (12) gravimetric procedure introduced in 1951. It has been adapted by various workers to suit varying conditions and to clarify calculation and interpretation of antiperspirant efficacy data. There is considerable controversy concerning the effect of side treated and the method of calculation of efficacy data. To control asymmetry in axillary sweating, Majors and Wild (13) adjusted posttest sweat ratios with pretest sweat ratios. They evaluated efficacy results using the arithmetic mean of the ratio method. Wooding and Finklestein (14) used the geometric mean in a Sides Subjects Effects Model. Their model produced antiperspirant efficacy results similar to those of the ratio method and eliminated pretest collection periods. The Review Panel accepts the efficacy results of both methods, and even suggests a simple binomial test to evaluate the antiperspirant effectiveness levels using a minimum of 20 subjects with or without pretest sweat ratios. To eliminate side treated differences and to compare treatment differences, we developed a statistical experimental design for a smaller number of subjects, viz., eight.