High-SPF sunscreens (SPF $ 70) may provide ultraviolet protection above minimal recommended levels by adequately compensating for

lower sunscreen user application amounts

Hao Ou-Yang, PhD,a Joseph Stanfield, MS,b Curtis Cole, PhD,c Yohini Appa, PhD,a and Darrell Rigel, MDd Los Angeles, California; Winston Salem, North Carolina; Skillman, New Jersey; and New York, New York
Background: The manner in which consumers apply sunscreens is often inadequate for ultraviolet protection according to the labeled sun protection factor (SPF). Although sunscreen SPFs are labeled by testing at an application density of 2 mg/cm2, the actual protection received is often substantially less because of consumer application densities ranging from 0.5 to 1 mg/cm2. High-SPF sunscreens may provide more adequate protection even when applied by consumers at inadequate amounts. Objective: We sought to measure the actual SPF values of various sunscreens (labeled SPF 30-100) applied in amounts typical of those used by consumers. Methods: Actual SPF values were measured on human volunteers for 6 sunscreen products with labeled SPF values ranging from 30 to 100, applied at 0.5, 1.0, 1.5, and 2.0 mg/cm2. Results: There was a linear relationship between application density and the actual SPF; sunscreens with labeled SPF values of 70 and above provided signicant protection, even at the low application densities typically applied by consumers. Sunscreens labeled SPF 70 and 100 applied at 0.5 mg/cm2 provided an actual SPF value of, respectively, 19 and 27. Limitations: The study was conducted in a laboratory setting under standardized conditions and results are extrapolated to actual in-use situations. Conclusion: Sunscreens with SPF 70 and above add additional clinical benets when applied by consumers at typically used amounts, by delivering an actual SPF that meets the minimum SPF levels recommended for skin cancer and photodamage prevention. In contrast, sunscreens with SPF 30 or 50 may not produce sufcient protection at actual consumer usage levels. ( J Am Acad Dermatol 2012;67:1220-7.) Key words: minimal erythema dose; sun protection; sunburn; sunburn protection factor; sunscreens; ultraviolet protection.

T
not

he harmful effects of ultraviolet (UV) radiation on the skin have been known for quite some time. Exposure to UV radiation leads only to skin photodamage1 but, more

importantly, to immunosuppression2 and skin cancers3 such as squamous cell carcinoma,4 basal cell carcinoma,5 and melanoma.6,7 To reduce the incidence of UV-related skin cancers, the American
conducting this study, and he is currently a consultant to Galderma Laboratories LP. Dr Rigel is a consultant for Neutrogena Corp, Johnson and Johnson Consumer Companies, Beiersdorf, and P&G. Previously presented as a poster at the 69th American Academy of Dermatology Annual Meeting, New Orleans, Feb. 4-8, 2011. Accepted for publication February 17, 2012. Reprint requests: Hao Ou-Yang, PhD, Neutrogena Corp, 5760 W 96 St, Los Angeles, CA 90045. E-mail: houyan4@its.jnj.com. Published online April 1, 2012. 0190-9622/$36.00 2012 by the American Academy of Dermatology, Inc. doi:10.1016/j.jaad.2012.02.029

From Neutrogena Corp, Los Angelesa; Suncare Research Laboratories LLC, Winston Salemb; Johnson and Johnson Consumer Companies, Skillmanc; and New York University Medical Center.d Funded in full by Neutrogena Corp; the preparation of this manuscript was sponsored in full by Johnson and Johnson Consumer Companies. Disclosure: Drs Ou-Yang and Appa are employees of Neutrogena Corp, the manufacturer of two of the sunscreens tested. Dr Cole is an employee of Johnson and Johnson Consumer Companies, a sister company of Neutrogena Corp. Mr Stanfield is an employee of Suncare Research Laboratories, the independent testing laboratory that received compensation for

1220

J AM ACAD DERMATOL
VOLUME 67, NUMBER 6

Ou-Yang et al 1221

Academy of Dermatology recommends, among users, to determine if high-SPF products applied per other measures, the generous application of a consumer behavior would still yield an SPF adequate broad-spectrum, water-resistant sunscreen with for skin cancer prevention,26 compared with lower SPF products. sun protection factor (SPF) of at least 30 to all exposed skin.8 Many studies have shown that sunscreens can METHODS protect from UV-induced damage and skin cancer.9-12 This study evaluated the effects of various application densities on the actual The determination of their SPF values of 6 sunscreen SPF value is regulated by CAPSULE SUMMARY products, with labeled SPFs the Food and Drug ranging from 30 to 100. The Administration (FDA) and Consumers typically apply sunscreens investigation was designed other internationally recogunevenly and in amounts lower than with a 2-step approach. An nized testing standards, recommended, reducing the actual initial study (study A) was which all dictate that sunsunburn protection factor. conducted on 251 volunteers screens be tested at an appliSunscreens with sun protection factor of to determine the SPF value of cation density of 2 mg/ 70 and above yield significant protection 6 sunscreens (labeled SPF cm2.13,14 However, several ineven when underapplied in typical vestigations have shown that from 30-100) at 4 different consumer use situations. consumers typically apply application densities (0.5, less than 2 mg/cm2,15-20 thus 1.0, 1.5, and 2.0 mg/cm2). Sunscreens with sun protection factor of reducing the actual protecThe second study (study B), 70 and above may deliver the minimal tion of the sunscreen.21-23 In conducted on 76 volunteers SPF levels for skin cancer and an open study conducted at after the results of study A photodamage prevention, even when the beach on 42 volunteers, were reviewed, was carried misused or under-applied. Bech-Thomsen and Wulf15 out to compare the SPF calculated that the average values of the 6 sunscreens sunscreen density applied by sunbathers was 0.5 at low application densities (0.5 and 1 mg/cm2) side by side. mg/cm2. Another group assessed the amount of sunscreen applied by 10 women with photosensitive Both studies were single center, controlled, ranskin conditions, as they would use it during a sunny domized, and evaluator blinded. Subjects were day; they found that the median application dose was male/female healthy volunteers with Fitzpatrick 0.5 mg/cm2, with site variations from 0 to 1.2 mg/ skin types I, II, and III.27 To facilitate estimation of 2 16 17 cm . Grosick and Tanner reported a study in the minimal erythema dose (MED), colorimetric which 700 women were given free access to various measurements of each subjects skin were obtained sunscreens SPF 15 and higher to be used over 3 to 6 according to the international SPF method, and the weeks; the study found that the SPF achieved in actual individual typology angle value was computed.14 Both studies were conducted in accordance to the use was below that of the labeled SPF. Investigations FDA final monograph on sunscreen products13,28 on sunscreen application by beachgoers have further and to the Declaration of Helsinki principles; protoconfirmed that consumers rarely apply sunscreens cols and informed consent were approved by an uniformly and usually do not reapply them often independent institutional review board. enough, particularly after swimming.18,19 In 2011, the FDA proposed to cap the labeled SPF at 501, stating that there is insufcient evidence Test sunscreen formulations and test sites that SPF values above 50 produce additional clinical Six US-marketed sunscreen formulations were benets.24 However, a consumer in-use study contested (4 lotions and two sprays). The 4 lotions ducted to compare the efficacy of a sunscreen with were selected based on similar vehicles and types of SPF 50 versus one with SPF 85 demonstrated that the active ingredients. The two sprays also contained higher SPF sunscreen provided better sunburn prosimilar types of active ingredients. The test products tection.25 In addition, it is widely presumed that were: high-SPF sunscreens may deliver an SPF adequate (A) SPF 30 lotion sunscreen (Coppertone Sport 30, for skin cancer and photodamage prevention even SPF 30, Schering-Plough HealthCare Products when underapplied by consumers. Inc, Memphis, TN). The purpose of this study was to test the actual (B) SPF 100 lotion sunscreen (Neutrogena Ultra SPFs of sunscreens with labeled SPFs 30 to 100 under Sheer Lotion SPF 100, Neutrogena Corp, Los the reduced densities that are typically applied by Angeles, CA).
d d d

1222 Ou-Yang et al

J AM ACAD DERMATOL

DECEMBER 2012

Abbreviations used: FDA: MED: SPF: UV: Food and Drug Administration minimal erythema dose sunburn protection factor ultraviolet

sources were calibrated using spectroradiometric measurements (model OL756 spectroradiometer, Optronic Laboratories, Orlando, FL) and were found to meet the specications for UV spectral distribution outlined in the FDA proposed amendment to the sunscreen monograph13 and the international SPF method.14 Procedures A total of 251 volunteers were enrolled in study A and 76 volunteers were enrolled in study B. Study B was conducted after reviewing the results from study A. Both studies followed the same procedures below, except where indicated. Day 1eMED determination. On day 1, 5 UV doses, progressively increasing by 25% increments, were applied to predetermined unprotected sites of the mid aspect of the back, as indicated in the FDA sunscreen monograph.28 Within 22 to 24 hours after irradiation, subjects returned to the testing laboratory for evaluation of UV-induced skin responses. A blinded trained evaluator graded the UV exposed sites for erythema using a warm fluorescent lamp emitting 450 to 550 lux: the MED was the smallest UV dose in the series to produce the first mild erythema with clearly defined borders. Day 2eapplication of products and UV doses for static SPF determination. Study A. On day 2, two to four 50-cm2 rectangles were marked on the subjects back between the belt line and shoulder blades. The number of sites depended on the expected SPF: only two high-SPF products could be tested together at the higher dose levels, because of the long exposure times. A technician applied the randomized density (0.5, 1.0, 1.5, or 2.0 mg/cm2) of one to three test products and the homosalate standard sunscreen (2.0 mg/cm2) to the designated rectangles on the subjects back. Test product allocation was determined by randomization. After at least 15 minutes from application, 7 progressive UV doses were applied to determine the MED for sunscreen-treated skin sites. Dose series were calculated by multiplying the expected SPF of each formulation, the initial subjects unprotected MED, and the following number: 0.76, 0.87, 0.93, 1.00, 1.07, 1.15, and 1.32.28 Expected SPF values were based on in vitro measurements of each product at the 4 application densities: they were anticipated to be approximately 75%, 50%, and 25% of the labeled SPF for application densities of 1.5, 1, and 0.5 mg/cm2, respectively. Study B. On day 2, four 50-cm2 rectangles were marked on the subjects back between the belt line and shoulder blades. A technician applied the randomized density (0.5 or 1.0 mg/cm2) of the 4 selected

(C) SPF 100 spray sunscreen (Neutrogena Ultimate Sport Spray SPF 100, Neutrogena Corp, Los Angeles, CA). (D) SPF 50 lotion sunscreen (Coppertone Sport 501, SPF 501, Schering-Plough HealthCare Products Inc, Memphis, TN). (E) SPF 50 spray sunscreen (Coppertone Sport 50 Continuous Spray, SPF 501, Schering-Plough HealthCare Products Inc, Memphis, TN). (F) SPF 70 lotion sunscreen (Coppertone Sport 701, SPF 701, Schering-Plough HealthCare Products Inc, Memphis, TN). The formulations were applied to the subjects back using a pipette and delivered in small droplets within pre-outlined skin sites; the droplets were spread as quickly and uniformly as possible using a nger cot. For spray-on products, formulations were rst sprayed into a beaker and swirled until no bubbles remained before being transferred to skin by pipette. In study A, the 6 test products at the 4 application densities were randomized among skin sites and subjects. Each subject received a standard sunscreen (homosalate 8%)28 and up to 3 test products, applied at one of the 4 application densities (0.5, 1.0, 1.5, and 2.0 mg/cm2) so that each test product was evaluated at all 4 amounts on at least 20 subjects. The 8% homosalate control was obtained commercially and had a labeled SPF of 4. In study B, the SPF determination of the 6 sunscreen formulations applied at 0.5 and 1.0 mg/ cm2 was conducted in 3 parallel panels of at least 20 subjects each. Each subject received 4 test products/ densities tested side by side to allow within-subject comparisons of similar types of products (Table I); each panel received a fixed product/density combination: one panel compared the two sprays at both application densities; a second panel compared the 4 lotions applied at 0.5 mg/cm2; and a third panel compared the 4 lotions applied at 1.0 mg/cm2. Test products and/or densities were randomized within each panel. UV source Single-port xenon arc solar simulators (model 16S, Solar Light Co, Glenside, PA) were used to administer UV doses. Output spectra of the light

J AM ACAD DERMATOL
VOLUME 67, NUMBER 6

Ou-Yang et al 1223

Table I. Test product allocation

Test products

Panel 1 (20 subjects) Panel 2 (20 subjects) Panel 3 (20 subjects)

SPF, Sunburn protection factor.

(E) SPF 50 spray at 0.5 mg/cm (E) SPF 50 spray at 1.0 mg/cm2 (A) SPF 30 lotion at 0.5 mg/cm2 (F) SPF 70 lotion at 0.5 mg/cm2 (A) SPF 30 lotion at 1.0 mg/cm2 (F) SPF 70 lotion at 1.0 mg/cm2

(C) SPF 100 spray at 0.5 mg/cm2 (C) SPF 100 spray at 1.0 mg/cm2 (D) SPF 50 lotion at 0.5 mg/cm2 (B) SPF 100 lotion at 0.5 mg/cm2 (D) SPF 50 lotion at 1.0 mg/cm2 (B) SPF 100 lotion at 1.0 mg/cm2

test products in their designated rectangles. The test products to be applied were determined by a randomization code balanced by panel, application density, and site location, to the maximum extent possible. At least 15 minutes after application, 7 progressive UV doses were applied to determine the MED for sunscreen-treated skin sites. Dose series were calculated as in study A. Expected SPF values were based on the in vitro measurements of each product at the two application densities (study A): they were anticipated to be approximately 50% and 25% of the labeled SPF for application densities of 1.0 and 0.5 mg/cm2, respectively. Day 2erepeated MED determination. On day 2, 5 UV doses, progressively increasing by 25% increments, were also readministered to an unprotected area of the mid aspect of the back. The 5-dose series was calculated based on multiples of the original MED, as follows: 0.64, 0.80, 1.00 (original MED), 1.25, and 1.56. Day 3eevaluation of responses for SPF calculation and repeated MED. Skin responses were evaluated 22 to 24 hours after administration of UV doses (day 3), using the 7-point erythema grading scale. Grading was conducted by a trained evaluator, who did not participate in product applications or administration of UV doses. SPF computation After determination of the repeated MED as detailed above, SPF values were calculated for each subject as follows: Individual SPF = MED of protected skin/repeated MED of unprotected skin. Mean SPF values and statistical information for each test product and application density were then calculated.28 Statistical analysis A one-way analysis of variance was performed within samples to test for differences in SPF by dose, and orthogonal contrasts were constructed to test linearity of the dose response. Within each dose

group, between-sample comparisons were analyzed by a one-way analysis of variance. Panel 1 data were analyzed by a mixed linear model, with dependent variable SPF values, and random model term for dose (0.5, 1.0) and xed term for sample (C, E), and the interaction of dose and sample. Panel 2 (at dose = 0.5) and panel 3 (at dose = 1.0) data analysis used a repeated measures analysis of variance to test for SPF differences among samples. Summary statistics for these statistical analyses include the least-squares mean difference between sample pairs and 95% condence intervals, using Tukey-Kramer adjustment for multiple comparisons. Statistical analysis was conducted using SAS software (SAS Institute Inc, Cary, NC) Version 9.2 on an XP Pro platform.

RESULTS
Study A A total of 237 subjects completed the study and 233 were included in SPF calculations. Data for 4 subjects were excluded because of protocol violations. Mean SPF values and statistical information for each test product and application density are presented in Table II. For all 6 test sunscreens, the mean SPF value was progressively lower with lower application densities. However, reduced application densities yielded proportionately higher mean SPF values for products with higher labeled SPFs. SPF values for the 2.0-mg/cm2 application were slightly higher than labeled values. Measured SPF values were statistically different for sunscreens with different labeled SPF when tested at the same application density and were statistically different for the same product when tested at different application densities. Correlation coefficients (R2) for linear curve fits were all greater than 0.99, suggesting a linear correlation between application densities of 0.5 to 2 mg/ cm2 and SPF (Fig 1). Study B A total of 65 subjects completed the study and all were included in SPF calculations. The results are

1224 Ou-Yang et al

J AM ACAD DERMATOL

DECEMBER 2012

Table II. Study A: sunburn protection factor results

AeLabel SPF: 30 DeLabel SPF: 501 FeLabel SPF: 701 BeLabel SPF: 100 EeLabel SPF: 50 spray CeLabel SPF: 100 spray

2 mg/cm Mean SPF SD n 95% CI 1.5 mg/cm2 Mean SPF SD n 95% CI 1 mg/cm2 Mean SPF SD n 95% CI 0.5 mg/cm2 Mean SPF SD n 95% CI

31.0 4.2 20 1.9 21.8 2.4 20 1.1 16.0 1.3 20 0.6 8.8 1.1 20 0.5

52.8 6.4 21 2.7 41.3 4.2 20 1.8 26.0 2.5 20 1.1 13.9 1.5 20 0.7

70.6 8.7 20 3.8 54.4 5.0 20 2.2 37.1 4.5 20 2.0 19.3 1.4 20 0.6

104.6 8.1 20 3.6 79.3 7.4 20 3.2 55.9 5.0 20 2.2 27.1 1.6 20 0.7

50.7 5.0 20 2.2 38.6 4.3 20 1.9 25.7 2.8 20 1.2 12.6 1.5 20 0.7

105.3 11.1 20 4.9 75.0 8.0 20 3.5 50.1 5.2 20 2.3 22.4 3.1 20 1.4

CI, Confidence interval; SPF, sunburn protection factor.

mg/cm2, the average SPF for labeled SPF 30 was 9.11, for labeled 501 was 14.34, for labeled 701 was 19.27, and for labeled 100 was 27.85. In this side-by-side test, the sunscreens with the higher labeled SPF values consistently provided statistically better protection than the sunscreens with the lowerlabeled SPF values at the application densities relevant to consumers.

DISCUSSION
In recent years, advancements in sunscreen formulations and access to photostabilized UV lter technology have allowed companies to manufacture products with SPF value up to 100. The marketing of very higheSPF products (SPF [50) has spurred an ongoing debate questioning their added clinical benets. Indeed, comparative investigations on very higheSPF sunscreens are scant. Russak et al25 conducted a split-face, double-blind, randomized study comparing an SPF 50 with an SPF 85 product. A total of 56 skier/snowboarders applied each sunscreen once to the assigned side of the face during a sunny day at high altitude and subjects were exposed to the sun during the snow activity for an average of 5 hours. The results showed that the product with SPF 85 provided significantly more protection against sunburn than the one with SPF 50. Many studies have demonstrated that consumers tend to use signicantly less product than recommended (0.5-1.0 mg/cm2) and in uneven applications,15-20 with a decrease in actual sunburn

Fig 1. Correlation for study A. Correlation coefcients (R2) for linear curve fits were all greater than 0.99, suggesting linear relationship between application amount and sunburn protection factor (SPF ) for all test products.

presented in Table III. The results confirmed the linear relationship between amount of sunscreen applied and static SPF value for the high-SPF sunscreens tested in study A, with the mean SPF value for 0.5-mg/cm2 applications resulting in approximately 25% of the labeled SPF. At applications of 0.5

J AM ACAD DERMATOL
VOLUME 67, NUMBER 6

Ou-Yang et al 1225

Table III. Sunburn protection factor results for study B

AeLabel SPF: 30 DeLabel SPF: 501 BeLabel SPF: 100 FeLabel SPF: 701 CeLabel SPF: 100 spray EeLabel SPF: 50 spray

0.5 mg/cm Mean SPF SD n 95% CI 1 mg/cm2 Mean SPF SD n 95% CI

9.11 1.30 21 0.6 15.15 1.63 21 0.7

14.34 2.16 21 0.9 26.69 3.55 21 1.5

27.85 2.84 21 1.2 51.78 5.08 20* 2.2

19.27 2.75 20* 1.2 34.80 4.82 20* 2.1

25.82 3.44 22 1.4 49.67 7.45 20y 3.3

13.28 2.22 22 0.9 25.4 3.6 21 1.6

CI, Confidence interval; SPF, sunburn protection factor. *Test failure on one subject. y Test failure on two subjects.

protection.21-23 In fact, a study evaluating two sunscreens (SPF 15 and 30) at 4 applications densities (0.5, 1, 1.5, and 2 mg/cm2) on 40 volunteers, found that the actual SPF decreased from 27 to 7 (SPF 30 at 2 and 0.5 mg/cm2) and from 16 to 4 (SPF 15 at 2 and 0.5 mg/cm2), with a more linear trend than that suggested by the Beer-Lambert law.22 Another study, testing two sunscreens (SPF 30 and 35) at the same 4 application densities on 15 Asian volunteers, found a similar decrease in sun protection.21 Bimczok et al,23 using products with SPF between 20 and 25 and 3 application densities (0.5, 1, and 2 mg/cm2), showed a linear correlation between application density and SPF value, with reduced amounts of sunscreens producing a decrease in their actual SPF. In the current study, using much higher SPF values, there was also a linear relationship between application densities (from 0.5-2 mg/cm2) and labeled SPF (from 30-100), with the mean actual SPF of the 0.5-mg/cm2 dose being approximately 25% of the labeled SPF for all 6 test products. Herzog29 also concluded, using in-silico methods, that sunscreen products with UVA/ UVB absorbance ratios between 0.4 and 0.8, such as the ones used in the current study, show a linear relationship between application density and SPF. It is not surprising that sunscreen protection on human skin may not obey an exponential Beer-Lambert relationship, because the irregular surface of the skin can lead to uneven film distribution within the testing site.30 It should be noted that during actual consumer use, high-SPF formulations (with lower water content) may have different spreadability when compared with lower SPF products (with higher water content), and, therefore, laboratory results conducted under standard application conditions may be somewhat different from actual in-use SPF. Emulsion type and other ingredients in the emulsion may also

play a role in nal lm uniformity. In this regard, the authors selected formulations with very similar emulsion bases (among lotions), so that comparison between products was more meaningful. In addition, SPF testing at low application densities was also conducted in a side-by-side manner to conrm the validity of the differences between formulations with differently labeled SPF. The current study did not nd an increase of variation as observed by Bimczok et al23 when the application density was moved from 2 to 0.5 mg/cm2 in this single-center test. However, the authors are in agreement with the FDA and Bimczok et al23 in continuing to support the application density of 2.0 mg/cm2 as a standard for laboratory SPF testing to maintain reproducibility across laboratories. SPF values should be considered to be a relative indicator of protection when comparing two different products, and not as absolute values that guarantee sunburn-free protection based on individual sunburn burn time and the current sun UV intensity.31 Consumers have learned to choose appropriate SPF levels that work for them under their usual sunexposure habits, and complaint rates for sunscreen failure to protect are in fact low. The results of this study are particularly important in view of the 2007 FDA proposed amendment to the 1999 sunscreen nal monograph13 and of the latest FDA proposed rule on over-the-counter sunscreens released in June 2011,24 both advocating a maximum SPF label of 501. According to the 2011 proposed rule, the FDA continues to question whether there is any additional clinical benefit in using sunscreens with SPF above 50 and is thus proposing to limit the maximum labeled SPF to 501 until further clinical proof is submitted. The authors believe that this study supports a more appropriate level of clinical protection obtained from sunscreens of labeled SPF

1226 Ou-Yang et al

J AM ACAD DERMATOL

DECEMBER 2012

values above 50 when used at doses more consistent with consumer behavior. In fact, sunscreens with SPF 70 and above, when broad spectrum, can provide a margin of safety for consumers by yielding at least an SPF 19 in this study even at concentrations as low as 0.5 mg/cm2, exceeding the minimal requirement of SPF 15 and thereby decreasing the subsequent risk of skin cancer and photodamage, based on current FDA recommendations,26 whereas sunscreens at or below SPF 50 may not produce enough protection at the same dose (actual SPF 13-14 in this study). It is noteworthy that the SPF 30 sunscreen delivered actual mean SPFs 15.15 and 9.11 when applied, respectively, at 1 and 0.5 mg/cm2, thus providing inadequate effectiveness at 0.5 mg/cm2 based on current recommendations for decreasing skin cancer risks. In summary, sunscreens with SPF 70 and above add additional clinical benets when applied by consumers at typically used concentrations by delivering an actual SPF that meets the minimum SPF levels recommended for skin cancer and photodamage prevention. In contrast, sunscreens with SPF 30 or 50 may not produce sufcient protection at actual consumer use levels.
The authors would like to thank Dr Alessandra Pagnoni, of Pagnoni Consulting LLC, for her writing and editorial assistance in the preparation of this manuscript. Dr Pagnoni is a consultant for Johnson and Johnson Consumer Companies and Neutrogena Corp. The authors would also like to thank Mark Van Buskirk, of Data Reduction, for statistical analysis of the study data. Mr Van Buskirk is a consultant for Johnson and Johnson Consumer Companies.
REFERENCES 1. Gilchrest BA, Szabo G, Flynn E, Goldwyn RM. Chronologic and actinically induced aging in human facial skin. J Invest Dermatol 1983;80:81S-5S. 2. Aubin F. Mechanisms involved in ultraviolet light-induced immunosuppression. Eur J Dermatol 2003;13:515-23. 3. Rigel DS. Cutaneous ultraviolet exposure and its relationship to the development of skin cancer. J Am Acad Dermatol 2008; 58:S129-32. 4. English DR, Armstrong BK, Kricker A, Winter MG, Heenan PJ, Randell PL. Case-control study of sun exposure and squamous cell carcinoma of the skin. Int J Cancer 1998;77:347-53. 5. van Dam RM, Huang Z, Rimm EB, Weinstock MA, Spiegelman D, Colditz GA, et al. Risk factors for basal cell carcinoma of the skin in men: results from the health professionals follow-up study. Am J Epidemiol 1999;150:459-68. 6. Ortonne JP. Photobiology and genetics of malignant melanoma. Br J Dermatol 2002;146:S11-6. 7. Markovic SN, Erickson LA, Rao RD, Weenig RH, Pockaj BA, Bardia A, et al. Melanoma Study Group of the Mayo Clinic Cancer Center. Malignant melanoma in the 21st century, part 1: epidemiology, risk factors, screening, prevention, and diagnosis. Mayo Clin Proc 2007;82:364-80.

8. American Academy of Dermatology. Melanoma. Available from: http://www.aad.org/media-resources/stats-and-facts/ conditions/melanoma/melanoma. Accessed December 8, 2011. 9. Kligman LH, Akin FJ, Kligman AM. Sunscreens prevent ultraviolet photocarcinogenesis. J Am Acad Dermatol 1980;3:30-5. 10. Roberts LK, Beasley DG. Sunscreen lotions prevent ultraviolet radiation-induced suppression of antitumor immune responses. Int J Cancer 1997;71:94-102. 11. Sambandan DR, Ratner D. Sunscreens: an overview and update. J Am Acad Dermatol 2011;64:748-58. 12. Green AC, Williams GM, Logan V, Strutton GM. Reduced melanoma after regular sunscreen use: randomized trial follow-up. J Clin Oncol 2011;29:257-63. 13. Sunscreen drug products for over-the-counter human use. proposed amendment of final monograph proposed rule for over-the-counter human use. Federal Register 21 CFR Parts 347 and 352. Sunscreen drug products 2007;72:49070-122. 14. COLIPA, CTFA-SA, JCIA, CTFA. COLIPA guidelines; international sun protection factor (SPF) test method, 2006. Available from: http://www.colipa.eu/publications-colipa-the-europeancosmetic-cosmetics-association/guidelines.html?view=item&id= 21. Accessed December 8, 2011. 15. Bech-Thomsen N, Wulf HC. Sunbathers application of sunscreen is probably inadequate to obtain the sun protection factor assigned to the preparation. Photodermatol Photoimmunol Photomed 1992-1993;9:242-4. 16. Azurdia RM, Pagliaro JA, Diffey BL, Rhodes LE. Sunscreen application by photosensitive patients is inadequate for protection. Br J Dermatol 1999;140:255-8. 17. Grosick T, Tanner P. Efficacy as used, not as tested, is true measure of sunscreen performance. J Am Acad Dermatol 2004;50:P31. 18. Lademann J, Schanzer S, Richter H, Pelchrzim RV, Zastrow L, Golz K, et al. Sunscreen application at the beach. J Cosmet Dermatol 2004;3:62-8. 19. Wright MA, Wright ST, Wagner RF. Mechanisms of sunscreen failure. J Am Acad Dermatol 2001;44:781-4. 20. Reich A, Harupa M, Bury M, Chrzaszcz J, Starczewska A. Application of sunscreen preparations: a need to change the regulations. Photodermatol Photoimmunol Photomed 2009; 25:242-4. 21. Kim SM, Oh BH, Lee YW, Choe YB, Ahn KJ. The relation between the amount of sunscreen applied and the sun protection factor in Asian skin. J Am Acad Dermatol 2010;62: 218-22. 22. Schalka S, dos Reis VM, Cuc e LC. The influence of the amount of sunscreen applied and its sun protection factor (SPF): evaluation of two sunscreens including the same ingredients at different concentrations. Photodermatol Photoimmunol Photomed 2009;25:175-80. 23. Bimczok R, Gers-Barlag H, Mundt C, Klette E, Bielfeldt S, Rudolph T, et al. Influence of applied quantity of sunscreen products on sun protection factorea multicenter study organized by the DGK task force sun protection. Skin Pharmacol Physiol 2007;20:57-64. 24. Revised effectiveness determination; sunscreen drug products for over-the-counter human use; proposed rule. Federal Register 21 CFR Part 201. Revised 2011;76:35672-8. 25. Russak J, Chen T, Appa Y, Rigel DS. A comparison of sunburn protection of high-sun protection (SPF) sunscreens: SPF 85 sunscreen is significantly more protective than SPF 50. J Am Acad Dermatol 2010;62:348-9. 26. Labeling and effectiveness testing; sunscreen drug products for over-the-counter human use; final rule. Federal

J AM ACAD DERMATOL
VOLUME 67, NUMBER 6

Ou-Yang et al 1227

Register 21 CFR parts 201 and 310. Labeling 2011;76: 35620-65. 27. Fitzpatrick TB. The validity and practicality of sun-reactive skin types I through VI. Arch Dermatol 1988;124:869-71. 28. Sunscreen drug products for over-the-counter human use; final monograph. Federal Register 21 CFR parts 310, 352, 700 and 740. Sunscreen 1999;64:27666-93.

29. Herzog B. Understanding sunscreeneinfluence of amount of application on performance determined by in vivo and in silico experiments. J Am Acad Dermatol 2007;56:AB40. 30. ONeil JJ. Effect of film irregularities on sunscreen efficacy. J Pharm Sci 1984;73:888-91. 31. Diffey B. Has the sun protection factor had its day? BMJ 2000; 320:176-7.