SPF TESTING METHODOLOGIES 195 Here in, we report data on 2,503 observations. In the data reported herein, the SPF of P2 (15.6 ± 2.5) decreases by 1.55 units for every 10 mJ/cm2 increase in MED (dashed line in Figure 3). This data suggests that to maximize the SPF value of a sunscreen, one should use subjects with very low MEDs. These data also suggest that people with higher MEDs will have less protection from a sunscreen compared with people with lower MEDs. This negative correlation between unprotected MED and SPF of a sunscreen might ac- count for some of the variability found in data from different sunscreen testing laborato- ries. This correlation also suggests that sunscreen testing should include subjects with a range of unprotected MED values, rather than Fitzpatrick Skin Phototype or ITA°. For example, perhaps not more than three subjects below an unprotected MED of 15 mJ/cm2 and at least three subjects above an unprotected MED of 40 mJ/cm2. The purpose of such a requirement would be identical to the reasons for a variety of Fitzpatrick Skin Photo- type in the 2011 FDA-Final Rule. CONCLU SIONS No cli nically signifi cant difference or statistically signifi cant difference was found be- tween the average SPF of P2 using the 2011 FDA-Final Rule methodology versus that using ISO 24444 methodology at this laboratory. Furthermore, the average SPF of P2 is independent of the type of solar simulator (multiport versus single-port), age of subject, gender of subject, or Fitzpatrick Skin Phototype of subject. The data clearly show a sta- tistically signifi cant negative correlation between a subject’s SPF of P2 and the subject’s unprotected MED. REFERE NCES (1) D e partment of Health and Human Services, Food and Drug Administration, Sunscreen drug products for over-the-counter human use fi nal monograph, Fed. Regist., 4, 27666–27693 (1999). (2) D e partment of Health and Human Services, Food and Drug Administration, Sunscreen drug products for over-the-counter human use proposed amendment of fi nal monograph proposed rule, Fed. Regist., 72, 49070–49122 (2007). (3) The Australian/New Zealand Standard™ (1998), Sunscreen Products - Evaluation and classifi cation, (Standard # AS/NZS 2604). (4) COLIPA, International Sun Protection Factor (SPF) Test Method, February 2003 (Joint Confe rence on Harmo- nization, Colipa, JCIA, and CTFA SA). (5) COLIPA, International Sun Protection Factor (SPF) Test Method, May 2006 (Joint Conference on Harmoniza- tion, Colipa, JCIA, CTFA SA and CTFA). (6) Department of Health and Human Services, Food and Drug Administration, Labeling and effe c tiveness testing sunscreen drug products for over-the-counter human use, Fed. Regist., 76, 35620–35665 (2011). (7) Cosmetics–Sun Protection Test Methods–In Vivo Determination of the Sun Protection Factor (SPF), ISO 24444:2010(E), accessed is November 15, 2010, https://www.iso.org/standard/46523.html. (8) K. Garzarella, M. Caswell, Disparate SPF testing methodologies generate similar SPFs, J. Cosmet. Sci., 64, 297–307 (2013). (9) D. L. Damian, G. M. Halliday, R. S. Barneston, Sun protection factor measurement of suns c reens is dependent on minimal erythemal dose, Br. J. Dermatol, 141, 502–507 (1999). (10) World Medical Association Declaration of Helsinki–Ethical Principles for Medical Resear c h Involving Human Subjects, accessed January 27, 2019, www.wma.net/Policy/Current Policies/WMA Declaration of Helsinki–Ethical Principles for Medical Research Involving Human Subjects. (11) The Belmont Report–Ethical Principles and Guidelines for the Protection of Human Subject s of Research, 1979, accessed January 27, 2019, www.hhs.gov/OHRP/Reg0075lations&Policy/TheBelmontReport.

JOURNAL OF COSMETIC SCIENCE 196 (12) D. S. Berger, Specifi cation and design of solar ultraviolet simulators, J. Inves tig. Der matol., 53, 192–199 (1969). (13) M. Caswell, C. Wood, G. Roberts, A. Martinez, No Difference Detected in SPF Det ermined U sing a 300 Watt Solar Simulator Versus a 150 Watt Solar Simulator, 19th Annual Photomedicine Society Meeting, accessed March 3, 2010. (14) A. Martinez, G. Roberts, K. Garzarella, M. Lutz, M. Caswell, Interoperability of 300 wat t and 150 watt xenon arc solar simulators in sun protection factor and in UVA protection factor clinical testing, Photo- dermatol. Photoimmunol. Photomed., 29, 78–83 (2013). (15) T. B. Fitzpatrick, Soleil et peau, J. Aesthetic Med., 2, 33–34 (1975). (16) T. B. Fitz p atrick, The validity and practicality of sun-reactive skin types I through V I , Arch. Dermatol., 124, 869–871 (1988). (17) S. Nakagawa, H. Schielzeth, A general and simple method for obtaining R2 from generaliz e d linear mixed-effects models, Methods Ecol. Evol., 4, 133–142 (2013). (18) R Core Team, R: A Language and Environment for Statistical Computing (R Foundation for St a tistical Com- puting, Vienna, Austria, accessed on January 27, 2019, www.R-project.or). (19) A. Kuznetsova, P. B. Brockhoff, R. H. B. Christensen, ImerTest package: tests in linear m i xed effects models, J. Stat. Softw., 82, 1–26 (2017). (20) H. Wickham, ggplot2: Elegant Graphics for Data Analysis (Springer-Verlag, New York, 2009) . (21) E. D. Baron, R. S. Stern, C. R. Taylor, Correlating skin type and minimum erythema dose, A rch. Der- matol., 135, 1278–1279 (1999). (22) B. K. Ho, J. K. Robinson, Color bar tool for skin type identifi cation: a cross-sectional s tudy, J. Am. Acad. Dermatol., 73, 312–313 (2015). (23) A. Chardon, I. Cretois, C. Hourseau, Skin colour typology and suntanning pathways, Int. J . Cosmet. Sci., 13, 191–208 (1991). (24) R. M. Sayre, D. L. Desrochers, C. J. Wilson, E. Marlowe, Skin type, minimal erythema dose (MED), and sunlight acclimatization, J. Am. Acad. Dermatol., 5, 439–443 (1981).