Abstract

Metal oxide nanoparticles are commonly used in personal-care formulations as protective agents against exposure to ultraviolet radiation. Although previous research has concluded that nanoparticles do not penetrate healthy skin, it remains contentious whether this conclusion holds under normal conditions of sunscreen use. Humans (n = 20) were exposed to sunscreens containing zinc oxide (ZnO) particles to determine if Zn from the particles was absorbed through skin over five consecutive days under outdoor conditions. Two sunscreens were tested—“nano sunscreen” containing 19-nm nanoparticles and “bulk sunscreen” containing > 100-nm particles. Venous blood and urine samples were collected 8 days before exposure, twice daily during the trial, and 6 days post-exposure. As the first application in nanotechnology studies, stable isotope tracing was used where the ZnO, enriched to > 99% with the stable isotope 68Zn, allowed dermally absorbed zinc to be distinguished from naturally occurring zinc. The overwhelming majority of applied 68Zn was not absorbed, although blood and urine samples from all subjects exhibited small increases in levels of tracer 68Zn. The amount of tracer detected in blood after the 5-day application period was ∼1/1000th that of total Zn in the blood compartment. Tracer levels in blood continued to increase beyond the 5-day application phase in contrast to those in urine. Levels of 68Zn in blood and urine from females receiving the nano sunscreen appeared to be higher than males receiving the same treatment and higher than all subjects receiving the bulk sunscreen. It is not known whether 68Zn has been absorbed as ZnO particles or soluble Zn or both.

The incidence of skin cancer is increasing globally. In Australia, more than 1600 people died from melanoma and nonmelanoma skin cancer in 2005 (Australian Institute of Health and Welfare [AIHW], 2005), and it was predicted that more than 430,000 people would be diagnosed with nonmelanoma skin cancer in 2008 (AIHW, 2008). In the United States, it was estimated that 68,720 men and women would be diagnosed with malignant melanoma in 2009 with the overwhelming proportion among Caucasians (National Cancer Institute, 2009).

The use of sunscreens is advocated for outdoor recreational and occupational activities in order to reduce the risk of skin burn and the development of skin cancer from exposure to ultraviolet (UV) radiation. Historically, organic chemical UV absorbers have been used in sunscreens, and they have their own set of potential health concerns (Environmental Working Group, 2009). Advances in formulation using nanotechnologies have seen the incorporation of the inorganic UV filters titanium dioxide (TiO2) and zinc oxide (ZnO) in nanoparticulate form into commercial sunscreens at amounts ranging from 4 to 30% wt/wt. Such sunscreens coat the skin as a film and work primarily by reflecting and absorbing UV light. One commercial advantage in using nanoparticulate metal oxides as opposed to larger particle sizes is that the film formed on the skin appears transparent in the visible spectrum rather than opaque. The Australian Therapeutic Goods Administration stated that there were almost 400 sunscreen products commercially available in Australia in 2006 (Therapeutic Goods Administration [TGA], Australia, 2006), with many containing nanoparticulate TiO2 and/or ZnO.

Nanoparticles are discrete objects, which have all three Cartesian dimensions less than 100 nm (ISO, 2008). At these sizes, there is potential for nanoparticles to penetrate cell walls and the blood-brain barrier and interact with biomolecules (reviewed in Osmond and McCall, 2010). Furthermore, for the metal oxides typically used in sunscreens, the greater specific surface area and chemical reactivity compared with larger particles potentially result in the generation of higher levels of free radicals and reactive oxygen species per unit mass (Nel et al., 2006). Free radicals resulting from the photoactivity of nanoparticles of TiO2 and ZnO have been reported to damage DNA in human skin cells when exposed to UV light (Dunford et al., 1997; Nakagawa et al., 1997).

The potential for, and consequences of, dermal absorption or penetration of metal oxide nanoparticles from personal-care products has not been determined conclusively. Following the World Health Organization (2006) definition, dermal absorption describes the transport of chemicals from the outer surface of the skin to the systemic circulation, and dermal penetration describes the entry of a substance into a particular layer or structure (e.g., the stratum corneum). Most investigations so far have been in vitro using diffusion cells, with few animal studies and fewer in vivo human studies (reviewed in Monteiro-Riviere and Baroli, 2010 and Sadrieh et al., 2010). In recent trials of sunscreen formulations applied to weanling pigs, Inman et al. (2010) and Monteiro-Riviere et al. (2010) found that TiO2 and ZnO nanoparticles were localized in the stratum corneum; the ZnO particles had a mean size of 140 nm and range 60–200 nm. Related studies to evaluate the dermal penetration of metallic or metal-bearing nanoparticles such as maghemite (Baroli et al., 2007), quantum dots (Ryman-Rasmussen et al., 2006), or silver (Larese et al., 2009) in excised human or porcine skin have shown limited penetration (see additional references in Baroli et al., 2007; Osmond and McCall, 2010; and the review of Monteiro-Riviere and Baroli, 2010).

Several reviews (Monteiro-Riviere and Baroli, 2010; Nohynek et al., 2007, 2010; TGA, 2009) and recent investigations (Cross et al., 2007; Inman et al., 2010; Monteiro-Riviere et al., 2010; Roberts et al., 2008; Sadrieh et al., 2010; Zyvagin et al., 2008) have concluded that metal oxide nanoparticles do not penetrate the stratum corneum, although they can lodge in hair follicles (Lademann et al., 1999, 2006; Nanoderm, 2007), sweat glands, or skin folds.

A major difficulty facing these studies is that highly sensitive methods are required to ensure the detection of very low levels (if any) of dermally absorbed nanoparticles, without altering the properties of the nanoparticles and hence their potential for dermal penetration. A concept proposed by Gulson and Wong (2006) to trace the metal is to make use of stable (nonradioactive) isotopes. Zinc has five stable isotopes, one of which, 68Zn, has a natural abundance of 18.8%. If ZnO particles highly enriched with 68Zn were incorporated into sunscreens and applied to skin, then increases in levels of 68Zn in blood and urine samples, relative to a different naturally occurring stable isotope not included in the sunscreen (in this case, 64Zn whose natural abundance is 48.6%), would indicate dermal penetration of Zn from the sunscreen even if the body were exposed to natural Zn from other sources.

This paper describes the first application of stable isotopes in nanotechnology specifically for tracing absorption or penetration of Zn from ZnO nanoparticles in sunscreen applied to healthy human skin under conditions of normal use.

MATERIALS AND METHODS

68ZnO particles.

ZnO powder enriched to > 99% 68Zn was purchased from Isoflex. Half of the stock was used to make nanoparticles with a final crystallite size of about 19 nm (± 8 nm; minimum 3 nm, maximum 60 nm) using a proprietary method based on high-energy attrition milling (Casey et al., 2006). The rest was used to make larger particles with an average crystallite size of 110 nm (± 46 nm; minimum 25 nm, maximum 284 nm) produced by a modified version of the same method. Primary crystallite size and phase were determined by x-ray diffractometry. Crystallite phase was determined using a Bruker ASXD8 X-Ray Diffractometer with Cu Kα radiation over a 2θ range of 5°–85° with a step size of 0.02°. The crystal structure of all particles was identical and the same as used in commercial sunscreens—single-phase hexagonal wurtzite. Crystallite size was determined by performing a Rietveld refinement of the diffraction data using Siroquant Version 2.5 software and confirmed by transmission electron microscopy (Fig. 1), which indicated that primary particles existed as single crystals, and the polydispersity of particle distributions was smaller at the smaller particle size.

FIG. 1.

Transmission electron microscopy (TEM) images of the two types of ZnO particles used in the sunscreens. For the bulk particles (image on left), the scale bar on the lower right-hand side is 1000 nm, and for the nanoparticles (image on right), the bar is 200 nm.

FIG. 1.

Transmission electron microscopy (TEM) images of the two types of ZnO particles used in the sunscreens. For the bulk particles (image on left), the scale bar on the lower right-hand side is 1000 nm, and for the nanoparticles (image on right), the bar is 200 nm.

Sunscreen formulation.

The uncoated particles were incorporated into an oil-water formulation using a commercial process for preparing sunscreens. Both sunscreens contained ∼20% wt/wt 68ZnO particles. Sun protection factor (SPF) was measured. To determine the character of the ZnO nanoparticles in sunscreen when applied to skin, a tape-stripping method was used, whereby about 12 mg of sunscreen was applied to a marked area of skin on the underside of the forearm over an area ∼6 cm2 (1.5 × 4 cm), equivalent to a dose of 2 mg/cm2 (the recommended dose to determine the SPF factor; Gabard et al., 2000; Lademann et al., 2004). The sunscreen was gently spread across the area by a gloved forefinger (disposable nitrile glove) and gently rubbed until an even coverage was obtained. A piece of tape of matching size was then applied on the marked area of skin. A total weight of 1.2 kg was placed on top of the tape for 30 s, introducing a pressure of 200 g/cm2 prior to removal of the tape from the skin.

The stripped tape was subsequently imaged by scanning electron microscopy (SEM) in back-scattered mode. The SEM analysis showed a relatively even distribution of the nanoparticles on the skin (Fig. 2).

FIG. 2.

Scanning electron microscope (SEM) images illustrating the character of the nanoparticle sunscreen on the skin of two subjects using a skin tape–stripping method. The scale bars are 20 μm.

FIG. 2.

Scanning electron microscope (SEM) images illustrating the character of the nanoparticle sunscreen on the skin of two subjects using a skin tape–stripping method. The scale bars are 20 μm.

Treatment.

The study was conducted over five consecutive days in early March 2009 (late summer in the Southern Hemisphere) using a tracer that was > 99% enriched in 68Zn. Sunscreen containing nanoparticles of 68ZnO was applied to a group consisting of 11 people (the “nano” group), whereas the other sunscreen containing non-nanoparticles of 68ZnO was applied to a group of 9 people (the “bulk” group) (see Fig. 3 for appearance on the skin as applied outdoors). Information on the subjects—age, gender, skin type, and race—is given in Table 1. UV exposure was continuously monitored throughout the trial days with a UV spectrophotometer. The mean UVA and UVB measures averaged over 5 days were 26.7 (± 10.1 SD) and 1.2 (± 0.6) W/m2, respectively. The protocol was developed with the aid of a pilot study conducted earlier with three subjects and using ZnO that was only 51% enriched in 68Zn.

TABLE 1

Subject Information and Key Experimental Results

Subject Particles Δ68Zn% beach Δ68Zn% post Gender Age (years) Skin typea Country Relationship Average dose (mg/cm2
NP 0.18 0.26 Male 60 IV South America  5.1 
NP  0.24 Male 20 II Australia  5.3 
NP 0.20 0.37 Female 23 II Australia Sibling of 4 3.7 
NP 0.22 0.42 Male 20 II/III Australia Sibling of 3 4.7 
Bulk 0.16 0.22 Male 24 I/II Australia  5.1 
NP 0.83 1.31 Female 44 II/III South America  3.8 
Bulk 0.30 0.43 Female 21 III Australia  4.0 
NP 0.27 0.52 Female 60 II/III South America  3.9 
10 Bulk 0.09 0.15 Female 34 UK  3.7 
12 NP 0.08 0.24 Male 66 I/II Australia  4.6 
13 Bulk 0.20 0.41 Male 23 II/III Australia  4.6 
14 Bulk 0.26 0.40 Male 21 II/III Australia/South American parents Brother of 20 3.8 
15 Bulk 0.25 0.42 Female 27 Germany  3.7 
16 NP 0.11 0.23 Male 27 Australia Son of 17 4.3 
17 Bulk 0.06 0.17 Male 59 Australia Father of 16 5.3 
18 NP 0.45 0.69 Female 21 IV South America  3.2 
19 NP  0.80 Female 19 III United States  4.1 
20 Bulk 0.18 0.32 Male 20 IV Australia/South American parents Brother of 14 5.2 
21 Bulk 0.10 0.22 Female 24 IV South America Twin of 22 3.3 
22 NP 0.28 0.58 Female 24 IV South America Twin of 21 4.0 
NP and Bulk 0.35 0.45 Female 64 I/II Australia  Sunscreen applicator 
Subject Particles Δ68Zn% beach Δ68Zn% post Gender Age (years) Skin typea Country Relationship Average dose (mg/cm2
NP 0.18 0.26 Male 60 IV South America  5.1 
NP  0.24 Male 20 II Australia  5.3 
NP 0.20 0.37 Female 23 II Australia Sibling of 4 3.7 
NP 0.22 0.42 Male 20 II/III Australia Sibling of 3 4.7 
Bulk 0.16 0.22 Male 24 I/II Australia  5.1 
NP 0.83 1.31 Female 44 II/III South America  3.8 
Bulk 0.30 0.43 Female 21 III Australia  4.0 
NP 0.27 0.52 Female 60 II/III South America  3.9 
10 Bulk 0.09 0.15 Female 34 UK  3.7 
12 NP 0.08 0.24 Male 66 I/II Australia  4.6 
13 Bulk 0.20 0.41 Male 23 II/III Australia  4.6 
14 Bulk 0.26 0.40 Male 21 II/III Australia/South American parents Brother of 20 3.8 
15 Bulk 0.25 0.42 Female 27 Germany  3.7 
16 NP 0.11 0.23 Male 27 Australia Son of 17 4.3 
17 Bulk 0.06 0.17 Male 59 Australia Father of 16 5.3 
18 NP 0.45 0.69 Female 21 IV South America  3.2 
19 NP  0.80 Female 19 III United States  4.1 
20 Bulk 0.18 0.32 Male 20 IV Australia/South American parents Brother of 14 5.2 
21 Bulk 0.10 0.22 Female 24 IV South America Twin of 22 3.3 
22 NP 0.28 0.58 Female 24 IV South America Twin of 21 4.0 
NP and Bulk 0.35 0.45 Female 64 I/II Australia  Sunscreen applicator 

Note. Subjects 5 and 11 withdrew after the first night.

a

Rated according to the Fitzpatrick (1988) classification and assessed by Gavin Greenoak, director of the Australian Photobiology Testing Facility, Sydney. Type IV is darker than I.

FIG. 3.

Appearance on the skin of the two sunscreen formulations after the sixth application at the beach. The left-hand image is for a female subject who had nano sunscreen applied to her back, whereas the right-hand image is for a male who had sunscreen with larger particles applied. The transparency of the sunscreen on the female indicates that there is little agglomeration of the ZnO nanoparticles.

FIG. 3.

Appearance on the skin of the two sunscreen formulations after the sixth application at the beach. The left-hand image is for a female subject who had nano sunscreen applied to her back, whereas the right-hand image is for a male who had sunscreen with larger particles applied. The transparency of the sunscreen on the female indicates that there is little agglomeration of the ZnO nanoparticles.

The first day of the study was conducted at an aquatic center to refine protocols, with the following 4 days conducted at a Sydney beach. Subjects wore UV protective upper body garments, the backs of which had a specific section cut out. Sunscreen was applied twice daily for a period of 5 days to the skin left uncovered by the garment. All sunscreen applications and removals for all subjects were performed by the same person (“X,” Table 1). After each application, a period of about 30 min was allowed for the formulation to equilibrate with the skin, as recommended by manufacturers, before subjects lay on their stomachs in the sun for a minimum of 30 min.

The tubes containing sunscreen were weighed by the same investigator (subject 12) before and after each application. Mean doses of sunscreen are listed in Table 1. There was no significant difference in the dose of sunscreen (miligrams per square centimeter) for the first and second applications of the day, with an overall mean of 4.3 mg/cm2 and range 2.8–5.8 mg/cm2. This is more than double the usual testing dose of 2 mg/cm2. There was no significant difference between the doses for nanoparticle or bulk sunscreen, but there was a significant difference between doses for males and females with a mean of 4.6 mg/cm2 for males versus 3.7 mg/cm2 for females.

For other body areas not used in the trial, such as the face and legs, subjects were encouraged to apply a commercial sunscreen of similar formulation to the test sunscreen but where the ZnO was replaced by chemical UV absorbers. Our aim was to determine dermal absorption of Zn from sunscreens in humans undergoing normal activities at the beach. Therefore, after a 30-min UV exposure following equilibration of each sunscreen application, the subjects were free for the rest of the day to swim, surf, sunbathe, walk, or spend time sheltered from the sun in a surf life–saving club. No effort was made to control perspiration or movement (resulting in skin flexing).

Blood and urine samples were collected 8 days before the start of the trial, just prior to the first application and after the removal of the last application on each of the 5 days of the trial and also 6 days after the end of the trial. Venous blood was collected by venipuncture into low-metal Vacutainer tubes. Urine samples were collected in precleaned standard containers. Additional urine samples were provided during the day and subsequent to the trial from a subset of subjects. Collection of blood and urine samples prior to any application of these sunscreens meant that each subject acted as his or her own control. This obviated the necessity of having a control group to which we would have applied equivalent sunscreen except with naturally occurring Zn.

Because of the sensitivity of the isotope method, subjects were continuously reminded, verbally and with signage, of the need to minimize contamination from the sunscreen especially during urine collection. Beach towels and a paper towel covering (renewed each day) were supplied to assist in eliminating the potential for contamination. The towels were washed each evening by one of the organizers and UV clothing by another. Sunscreen was removed from the subjects’ backs at the end of each day by using alcohol-lanoline wipes. Subjects kept a diary of UV exposure, other activities, changes of garments, any concerns over inadvertently touching their backs, or other unusual happenings.

Laboratory procedures.

Zn was purified from blood (0.2 ml) and urine (2–6 ml) samples by ion exchange through macroporous resin following digestion with ultraclean nitric acid and hydrogen peroxide. Total Zn levels in blank controls processed by these procedures were routinely less than 3 ng, which is insignificant compared with the total amounts of Zn in the blood and urine samples. Changes in the isotopic abundance of 68Zn of the purified samples were measured by multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS) at the Research School of Earth Sciences, Australian National University, and used to evaluate the dermal absorption of Zn from the sunscreens. A Zn solution with naturally occurring (normal) isotopic ratios was measured several times during each analytical session to obtain an estimate of the precision of the isotopic ratios. These data were complemented by “control” blood and urine samples as well as the base-line (before exposure) samples obtained before any sunscreen containing 68ZnO was applied. To enable comparisons between laboratories, the isotope ratios have been normalized to a 66Zn/64Zn value of 0.596. The total Zn concentrations in all samples analyzed for their Zn isotopic ratios were determined by ICP-MS.

Isotopic measures.

The use of MC-ICP-MS technology enables the precise measurement of isotopic ratios in samples. The key measure used in this study to determine whether Zn from sunscreens is absorbed through the skin is the change, if any, of the ratio of 68Zn to 64Zn with sunscreen exposure. More precisely, the percentage change in 68Zn with sunscreen application (denoted Δ68Zn%) is defined as: 

(1)
graphic
where the 68Zn/64Znexposure refers to the measurement for samples taken at the beach or post-exposure.

The important advantage afforded by this measure is the sensitivity with which it can detect the presence of 68Zn in blood and urine samples absorbed through the sunscreen-exposed skin. Δ68Zn% will be zero if no 68Zn from the sunscreens enters the blood or urine via dermal penetration, even if naturally occurring Zn enters the body from another source. This is because the 68Zn/64Zn ratio of naturally occurring Zn is essentially constant (as vindicated by the preexposure results to be discussed later) and additional amounts of natural Zn that may enter the body during the trial period will not change the (natural) ratio. Any increase, on the other hand, in Δ68Zn% provides an unambiguous indication of 68Zn from 68ZnO in the sunscreens entering the body.

Although this is a sensitive and robust method of detecting the presence of absorbed Zn from sunscreens, for a particular score of Δ68Zn%, further analysis is required before the amount of 68Zn absorbed in absolute terms can be ascertained. Δ68Zn% is not only proportional to the absorbed 68Zn from the sunscreen but also inversely proportional to the amount of natural Zn present in the reservoir measured. For example, if knowledge was required of the absolute amount of absorbed 68Zn from the sunscreen into the blood, in addition to the Δ68Zn% score, it is necessary to know the amount of natural Zn present in the blood of that subject. Two subjects with the same Δ68Zn% score may in fact have different amounts of absorbed 68Zn from sunscreens if the amount of naturally occurring Zn in their blood is different.

Our interest was in comparing absolute amounts of absorbed 68Zn from sunscreen exposure in four different groups, with gender as a factor. Two approaches are available. The first approach is to use fat-free mass as a basis for representing the natural Zn reservoir. To obtain a more reliable estimate of how much Zn an individual absorbed relative to all other subjects in the study, we used the fat-free mass based on the formula of Deurenberg et al. (1991) and the measured blood Zn concentration as representative of the total natural Zn for each subject to adjust the Δ68Zn%; this formula is based on body mass index, gender, and age. This adjustment removed the gender bias inherent in the Δ68Zn% measure, and the adjusted Δ68Zn% score was used in statistical analyses to make comparisons between different groups.

The second approach for estimating the total amount of natural Zn present in the blood of each subject is to use the measured total blood Zn concentration in the before exposure blood sample and multiply this by an estimate of the individual’s blood volume obtained using the formula of Nadler et al. (1962). These estimations provide amounts in micrograms of the 68Zn tracer in blood for individuals and allow comparisons with the known amounts of naturally occurring Zn in blood and other tissues such as liver.

Statistical analyses.

Changes in Δ68Zn% for blood and urine samples were estimated on the last day of the 5-day sunscreen exposure phase and then 6 days later (post-exposure) relative to the initial (before exposure) values using Equation (1) above. The data come from 20 subjects split between the nano group with 6 females and 5 males and the bulk group with 4 females and 5 males.

To evaluate the overall difference between the exposure phase and post-exposure results, the Wilcoxon test (SPSS version 18.0; SPSS Inc., Chicago, IL) was used. It is important to note that comparing exposure and post-exposure results does not involve comparisons across groups but within groups and therefore does not necessitate adjusted Δ68Zn%.

Because the distribution of the dependent variable, Δ68Zn%, was positively skewed (skewness index = 2.1), log10-transformed versions of the data (skewness index = 0.02) were used in analyses. This transformation, as it compresses the scale at the higher end, also reduced the influence of the high value observed for subject 7.

Initial analysis using a mixed model showed a very strong time effect. Further analyses, also exploring male and female differences, were based on postexposure, Deurenberg fat-free mass adjusted (and log10-transformed) Δ68Zn% as described above, and utilized a 2 × 2 independent groups ANOVA. In addition, differences in the amount of 68Zn tracer in blood (micrograms) were also evaluated with a 2 × 2 independent groups ANOVA. Effect sizes were calculated using partial eta-squared (forumla; Olejnik and Algina, 2003).

These studies have been approved by human ethics committees at Macquarie University and CSIRO.

RESULTS

Seventeen of the 20 subjects completed the full trial. Subject 2 had an unforseen commitment on day 5. Subject 7 had an adverse reaction to the sunscreen, and application was discontinued on day 4; however, she continued to provide blood and urine samples. Subject 19 was unable to provide blood samples at the beach but provided a blood sample before exposure and post-exposure and urine samples throughout. Results are presented in Table 1 and in Figures 4–8.

FIG. 4.

(A) and (B) Bar graphs showing the ratio 68Zn/64Zn in blood from subjects on whose backs bulk or nano sunscreen was applied. The before exposure data (red bars) illustrate the uniformity in 68Zn/64Zn ratios prior to sunscreen application, reflecting the isotopic composition of naturally occurring Zn. Increases in the ratio evident in all subjects at end of the beach exposure phase (blue bars) and 6 days post-exposure (pink) are due to skin penetration of 68Zn from the sunscreens. “X” represents the person who applied both sunscreens; her data were not included in the statistical analyses.

FIG. 4.

(A) and (B) Bar graphs showing the ratio 68Zn/64Zn in blood from subjects on whose backs bulk or nano sunscreen was applied. The before exposure data (red bars) illustrate the uniformity in 68Zn/64Zn ratios prior to sunscreen application, reflecting the isotopic composition of naturally occurring Zn. Increases in the ratio evident in all subjects at end of the beach exposure phase (blue bars) and 6 days post-exposure (pink) are due to skin penetration of 68Zn from the sunscreens. “X” represents the person who applied both sunscreens; her data were not included in the statistical analyses.

FIG. 5.

(A) and (B) Box plots for Δ68Zn% and amount of 68Zn tracer (micrograms) in blood samples for subjects, showing significantly higher Δ68Zn% and amounts of 68Zn tracer for females in the nano group (NP) compared with females in the bulk group but no difference between the nano and bulk groups for males. Data for subject 7 (the outlier) are included in the analyses. The data are for the post-exposure sampling.

FIG. 5.

(A) and (B) Box plots for Δ68Zn% and amount of 68Zn tracer (micrograms) in blood samples for subjects, showing significantly higher Δ68Zn% and amounts of 68Zn tracer for females in the nano group (NP) compared with females in the bulk group but no difference between the nano and bulk groups for males. Data for subject 7 (the outlier) are included in the analyses. The data are for the post-exposure sampling.

FIG. 6.

Changes in Δ68Zn% for urine samples. All values peak at the end of the beach exposure phase (day 5) and thereafter show a decrease over time. The red lines are for female subjects administered nano sunscreen. The Δ68Zn% values for five subjects (2, 3, 8, 9, and 16) at day 5 range up to 35, and for subject 21, the value was 330. Urine samples for all 20 subjects were collected up to 6 days post-exposure; samples from six subjects were collected at various times out to 40 days post-exposure.

FIG. 6.

Changes in Δ68Zn% for urine samples. All values peak at the end of the beach exposure phase (day 5) and thereafter show a decrease over time. The red lines are for female subjects administered nano sunscreen. The Δ68Zn% values for five subjects (2, 3, 8, 9, and 16) at day 5 range up to 35, and for subject 21, the value was 330. Urine samples for all 20 subjects were collected up to 6 days post-exposure; samples from six subjects were collected at various times out to 40 days post-exposure.

FIG. 7.

Plot of log(retention ratios) versus log(peak Δ68Zn%) to evaluate possible urine contamination with 68Zn from the sunscreen. Data for samples with minimal contamination (♦) lie on a line with a slope of approximately zero. For urine samples significantly affected by contamination (Δ), the data define a linear function with a negative slope of around −1. The two data points (marked X) in between the above groups do not affect the statistical outcomes.

FIG. 7.

Plot of log(retention ratios) versus log(peak Δ68Zn%) to evaluate possible urine contamination with 68Zn from the sunscreen. Data for samples with minimal contamination (♦) lie on a line with a slope of approximately zero. For urine samples significantly affected by contamination (Δ), the data define a linear function with a negative slope of around −1. The two data points (marked X) in between the above groups do not affect the statistical outcomes.

FIG. 8.

Δ68Zn% results for daily blood sampling collected at about 3 P.M. showing a linear increase with dose over the 5 days of sunscreen application for subjects 4, 8, 9, and 13. Note the nonlinear scale for the x-axis. Individual lines of best fit to the data from day 1 to day 5 have squared Pearson correlation coefficients (R2) of 0.95–0.96. Detection of the 68Zn tracer from the sunscreen was found from day 2.

FIG. 8.

Δ68Zn% results for daily blood sampling collected at about 3 P.M. showing a linear increase with dose over the 5 days of sunscreen application for subjects 4, 8, 9, and 13. Note the nonlinear scale for the x-axis. Individual lines of best fit to the data from day 1 to day 5 have squared Pearson correlation coefficients (R2) of 0.95–0.96. Detection of the 68Zn tracer from the sunscreen was found from day 2.

Blood Samples

Because of constraints on access time to the MC-ICP-MS, Zn isotopic compositions have been measured for only critical samples (before exposure, end of day 5 at the beach, and post-exposure) from all subjects, whereas more complete data have been obtained for four of the subjects. The precision of the analytical method is provided by isotopic data obtained for the before exposure blood and urine samples from all 20 subjects, as well as for the person who applied the sunscreen. The before exposure 68Zn/64Zn ratios of blood data shown by the red bars in Figures 4A and 4B have a mean of 0.41584 ± 0.00002 (± 0.006% [2σ, n = 21]). The variation of ± 0.022% (2σ, n = 21) for the before exposure urine samples is not as good as that of the blood (0.006% [2σ, n = 21]) probably due to lower concentrations of Zn in urine (∼1/10th) compared with blood. The extremely small variation in Zn isotopic composition of the before exposure blood and urine samples is consistent with the few other studies of very limited numbers of biological samples (Cloquet et al., 2008; Steenberg et al., 2005).

In contrast to the uniformity of before exposure 68Zn/64Zn ratios shown in Figures 4A and 4B, this ratio in blood increases for all subjects with sunscreen exposure and it continues to increase post-exposure. Although there are small variations in these data for the male bulk, male nano, and female bulk groups, the increases in 68Zn in blood for the female nano group are greater (Fig. 5A). In the female nano group, subject 7 shows an exceptionally large increase. The sunscreen applicator (X, Table 1, Fig. 4B) was not part of the trial (and not included in statistical analyses), but increases in 68Zn in her blood are comparable with the female nano group, consistent with her handling the nano sunscreen and indicating absorption through the skin of her hand.

The Wilcoxon test of the difference in Δ68Zn% between exposure and post-exposure results is highly significant (p < 0.0001). The mean increase in Δ68Zn% is 0.42 for post-exposure results compared with the mean increase of 0.23 on the last day of the exposure phase, relative to the before exposure mean (Fig. 5A).

The ANOVA results for the fat-free adjusted Δ68Zn% failed to show any significant effects at the 0.05 level. However, the interaction of gender and particle type was marginally significant, F(1, 16) = 4.34, p = 0.053, and because of the relatively large effect size (forumla = 0.21; i.e., the interaction of particle type and gender explained 21% of the variance in the adjusted Δ68Zn%), the simple effects of particle type were examined. Although effect of particle type was clearly insignificant for males, t(16) = 0.22, p = 0.83, forumla = 0.003, it was significant for females, t(16) = 2.71, p = 0.016, forumla = 0.314; i.e., 31% of the variance in adjusted Δ68Zn% for females can be accounted for by particle type.

The equivalent ANOVA on the data from the alternative method (to that of Deurenberg et al. fat-free adjusted Δ68Zn%) for estimating absolute amounts of absorbed Zn from sunscreens, based on measured Zn concentrations and estimated blood volumes, produced essentially identical results. The interaction of gender and particle was marginally significant, F(1, 16) = 4.44, p = 0.051. The effect of particle type was only significant for females (p = 0.012). Figure 5B shows that data for absolute amounts of absorbed 68Zn are consistent with Δ68Zn%.

Before accepting this result, it is necessary to consider whether any individual had an undue influence on it, in light of the small sample. To this end, standardized residuals and Cook’s distance (Cook, 1977) were calculated for each case. None of the values obtained was large: the highest absolute residual was 1.47 and the mean Cook’s distance was 0.20. Particular attention was paid to subject 7, who had the highest value of Δ68Zn%. Partly as a result of the logarithmic transformation, neither the standardized residual (1.21) nor Cook’s distance (0.09) for this subject suggested that she had a misleading influence on the overall results.

In further analyses of the fat-free data, we considered possible confounding effects, which might modify the conclusion that particle type affected the female subjects but not the males. Analyses were carried out in which age, average dose of sunscreen, skin type (treated as a numeric variable), and country (three categories), respectively, were added to the original 2 × 2 ANOVA as covariates. The simple effect of particle type for females was significant with each covariate held constant: with age, t(15) = 2.95, p = 0.010, forumla = 0.366; with skin type, t(15) = 2.35, p = 0.032, forumla = 0.271; and with country, t(14) = 2.43, p = 0.029, forumla = 0.297. For males, the simple effect of particle type remained nonsignificant in each case (p > 0.8).

In summary, in spite of the small numbers of subjects, the simple effect of particle type is substantial, whereby there are larger amounts of the tracer 68Zn in blood of females who received the nanoparticle sunscreen. This result cannot be easily dismissed.

Zinc Blood Levels

The mean total Zn content in whole blood was measured as 3.55 mg/l, ranging from 2.36 to 4.62 mg/l. There were significantly lower amounts of Zn in blood samples taken before exposure for females (mean 3.05 mg/l) compared with males (3.83 mg/l; p = 0.01), and differences were also present in post-exposure samples (3.21 vs. 3.63 mg/l; p = 0.04). Changes over time in Zn content vary for individuals but there do not appear to be any systematic trends and certainly no systematic increases associated with increases in isotopic 68Zn levels.

Urine Samples

Urine samples showed large increases in Δ68Zn% with sunscreen exposure when compared with the blood samples. All subjects showed the same temporal characteristic with Δ68Zn% values peaking around the last sunscreen application and decaying from then onwards (Fig. 6). Positive Δ68Zn% values were still detectable 6 days after the last sunscreen application and for the four subjects tested 25–40 days after application. There was no simple relationship between Δ68Zn% and total Zn concentration in urine.

The large variation in the peak Δ68Zn% values across subjects may appear to make it difficult to detect a relationship between urine and blood findings. Although the bulk of the subjects had peak Δ68Zn% values ranging between 0 and 4, six subjects had peak values ranging from 5 to 330. If the results are taken at face value and subjected to statistical analysis, there are no significant differences between the four groups.

One factor, however, that can account for some of the high peak values is the potential for urine samples taken at the beach to be contaminated with ZnO from sunscreens, especially for females. Strictly speaking, there is no certainty that a particular result is due to contamination, even if it is atypically high. What we can do is suppose that contamination did happen, infer what the effects would be and investigate the results for the effects. Assuming that there is contamination, we would expect to see for samples that are contaminated as outcomes: (1) larger peak Δ68Zn% values than if there were no contamination, given that peak Δ68Zn% values coincide with samples taken at the beach; (2) smaller retention ratios of Zn following the end of sunscreen application (retention ratio being defined as Δ68Zn% on day 11/Δ68Zn% on day 5), and (3) the retention ratio reducing as a function of increasing peak values reached by day 5 (which would produce an asymptotic slope of −1 on a log plot; see Supplementary notes). In contrast, for those samples that are contamination free, retention ratios will be higher and they will not be a function of peak Δ68Zn% values achieved by day 5.

The extent of urine contamination is evaluated in detail in the Supplementary notes. Figure 7 is a scatter log plot of retention ratio versus peak Δ68Zn% for all subjects. The first two expectations complement one another; the six samples with the highest peak Δ68Zn% values (day 5) also have the lowest retention ratios. The total pool of samples separate into two clusters, one (♦) being scattered around a flat straight line suggestive of no correlation between retention rate and peak Δ68Zn% and hence negligible contamination, and the other (Δ) showing a linear relationship with a slope of around −1, indicative of overwhelming contamination (see Supplementary notes).

If the second cluster is omitted and a 2 by 2 factorial ANOVA is performed on the rest of the day 5 samples with particle and gender as factors, there is a significant particle effect, a gender effect, and a second-order gender-particle interaction. Post hoc comparison of the four groups (nano females, nano males, bulk females, and bulk males) corrected for multiplicity using the Bonferroni method show that nano females show higher Δ68Zn% values than the other three groups (with p values of 0.0007, 0.009, and 0.002 for nano males, bulk females, and bulk males, respectively), between which there are no significant differences. This is consistent with the results for the blood data. Furthermore, this finding is robust with respect to the samples that fall between the two clusters (crosses in Fig. 7). The statistical significance of the results does not change whether or not we include either one or both of these samples.

Performing a 2 by 2 factorial ANOVA on the whole set of day 11 samples (without any exclusions) serves the double purpose of testing independently the contamination model for day 5 samples (see Supplementary notes), as well as confirming the statistical finding that nano females show higher Δ68Zn% values than the other three groups. The analysis of day 11 data gives essentially the same result—nano females have higher Δ68Zn% than the other three groups (p values of 0.002, 0.041, and 0.009 for nano males, bulk females, and bulk males, respectively), with no other significant differences.

Tracer Changes with Dose

To evaluate if there was a positive relationship between applied dose and penetration, blood and urine samples collected at the end of each day for one male and one female from each group were analyzed. These four subjects (4, 8, 9, and 13) were selected for more detailed analysis because increases in Δ68Zn% in their blood, as shown in Figures 4A and 4B, were close to the mean increase over all subjects. Figure 8 shows that increases in their Δ68Zn% in blood are linear over 4 days, detectable from day 2 after the fourth application of sunscreen, and continue to increase when sunscreen is no longer applied (tested 6 days after the last application). The profiles for changes in Δ68Zn% over time for most subjects in this trial are similar to those observed in a smaller pilot trial of three subjects conducted earlier to test and refine the design of this larger and more expensive trial.

DISCUSSION

There are four key findings from the present study. The most notable one is that, contrary to the dominant view (Baroli et al., 2007; Inman et al., 2010; Monteiro-Riviere et al., 2010; Ryman-Rasmussen et al., 2006) and for the first time, the study reveals unequivocal evidence that Zn from ZnO particles in sunscreens is absorbed through healthy human skin exposed to sunlight and is detectable in blood and urine. The Zn may have been absorbed by (intercellular or intracellular) diffusion, via hair follicles, skin folds, or sweat glands or a combination of these. The sunscreen formulation may have assisted the absorption. The commercially based formulation for the production of sunscreens used in this study contained isopropyl myristate, a known chemical enhancer of skin penetration (Chan, 2005), and EDTA, a chelating agent which is highly effective for Zn. Molecules from the formulation physically adsorbed onto the ZnO could also affect dissolution, dissolution rates, and/or skin penetration.

The second finding pertains to the absolute amounts of Zn that are absorbed from the sunscreens. The total amounts absorbed as detected in blood and urine were small when compared with the amounts of natural Zn normally present in the human body. For example, for the nano female group, which appeared to show the largest levels of absorption, the average Δ68Zn% in blood post-exposure is 0.71% (Table 1, Fig. 5A) and the amounts of 68Zn tracer from sunscreen circulating in whole blood range from 8.6 to 30.8 μg (mean 15.8 μg; Fig. 5B). These small amounts contrast with an average amount of Zn in whole blood, post-exposure for these females, of about 12 mg. In addition, these amounts are minute when compared with the dietary intake of Zn, recommended daily values for which are 8 mg Zn for females and 11 mg for males (National Institutes of Health, 2009). The amount of 68Zn tracer detected in blood posttrial represents less than 0.001% of the applied dose. Nevertheless, there are grounds which suggest that the amount of Zn absorbed from sunscreens may be somewhat larger than indicated by the blood results. A strong empirical support provided within the confines of this study for the possibility of additional Zn being deposited elsewhere in the body is that the amount of 68Zn detected in blood continued to increase 6 days after the last sunscreen application. 68Zn may have concentrated in particular tissues (e.g., epidermis, liver, muscle) with subsequent slow rerelease into the blood. Though, if the absorbed Zn is in ionic form, it is reasonable to expect that the amounts of 68Zn tracer would once again be overwhelmed by the naturally occurring Zn in these tissues, we cannot make any assumptions if the absorbed Zn is in form of ZnO particles. The binding and retention of residues in the lower epidermis or dermis potentially acting as a long-term chemical reservoir is termed substantivity (Ngo and Maibach, 2010). The continuing increase in blood after the peak of 68Zn in urine is likely to be due to the uptake of Zn by the red blood cells; urine is “fed” from the plasma compartment which turns over rapidly, whereas Zn in the red blood cell turns over slowly (Wastney, written communication, 2010). It should also be kept in mind that sunscreens are recommended for lifetime use. In countries with a sunny climate such as Australia, this translates into many days of actual application, in contrast to the 5-day exposure of our trial.

The third key finding is the interaction between ZnO particle size and gender in determining the levels of absorption (as reflected by blood and urine measures), which may well be an expression of an underlying skin thickness and particle size interaction. Females on average have thinner skin than males. The small sample size notwithstanding, more 68Zn absorption was detected in blood samples for females in the nano group compared with females in the bulk group, whereas there is no significant difference between the male nano and male bulk groups. Similarly, in urine, both exposure and post-exposure data revealed that the female nano group had higher levels of absorbed 68Zn than female bulk, male nano, and male bulk groups with no significant differences within the latter three groups. Other, less obvious, gender-related factors which may contribute to the interaction with particle size are differences in skin pH, lower surface lipid content (e.g., ceramide) in specific areas and, with aging, significant changes in ceramide ratios (reviewed in Dao and Kazin, 2007 and Tagami, 2010).

The fourth finding is that there is a time lag between the first sunscreen application and the first detection of tracer 68Zn in samples. Tracer 68Zn was first detected in blood after the fourth sunscreen application and on the afternoon of the second day of the exposure period. This has implications for the conclusions derived from studies, which involved fewer applications, shorter time periods, or less sensitive methods to detect absorption.

It should be noted both as a limitation and to avoid any ambiguity that Zn detected in the blood and urine from sunscreens is not necessarily in the form of ZnO (nano)particles. Tracer 68Zn may have been absorbed as either ZnO (nano)particles or soluble Zn or both. Nano-sized particles potentially release more ionic Zn due to their larger surface area. The pH of the outer layer of the skin, the stratum corneum, ranges from 5.4 to 5.9 (Schmid-Wendtner and Korting, 2006), and in such an environment, ZnO particles could partially dissolve. For example, Ågren (1990) found a 10-fold increase in Zn extracted from a ZnO dressing at a pH of 5.4 compared with 7.4. Although currently undergoing investigation by confocal microscopy in urine samples from subjects that contained significant 68Zn, ZnO particles may not be detectable because of the limited amount of absorption over the relatively short time of the trial. We have been unable to find any studies investigating dermal penetration of soluble Zn in humans. Three earlier studies of human subjects used larger particles of ZnO in ointments (Ågren, 1990; Derry et al., 1983; Morgan et al., 1980). In contrast, several in vitro studies using diffusion cells have shown that small amounts (0.3–0.4%) of Zn salts such as ZnO and ZnSO4 can be absorbed through human excised skin (e.g., Pirot et al., 1996).

The amount of individual UV exposure appears unrelated to the amount of dermal absorption. Some of our subjects experienced several hours of UV exposure and their results were no different to others who had only the required minimal exposure of 1 h per day. Sweating and increased skin temperature associated with sun exposure and activity may enhance skin permeability (Benech-Kieffer et al., 2003), but these variables are difficult to quantify.

In conclusion, using highly sensitive stable Zn isotopes as tracers, this study has demonstrated that small amounts of Zn from ZnO particles in sunscreens can pass through the protective layers of skin exposed to the sun in a real-life environment and be detected in blood and urine.

SUPPLEMENTARY DATA

Supplementary data are available online at http://toxsci.oxfordjournals.org/.

FUNDING

The CSIRO Flagship Collaboration Scheme and Macquarie University provided partial funding for these trials.

We thank Brent Baxter of Baxter Pharmaceuticals for preparation of the sunscreen formulations for the beach trial; Mary Salter for phlebotomy; the North Curl Curl Surf Life Saving Club and Macquarie University Medical Centre and Aquatic Centre for use of their facilities; Dianne Gulson as the applicator and assistance with logistics; Hong Yin and Lynne Waddington for SEM and TEM analyses and images; Vicki Tutungi for encouragement and CSIRO funding for the trial; and the CSIRO Flagship Collaboration Scheme and Macquarie University for partial funding of these trials.

References

Ågren
MS
Percutaneous absorption of zinc from zinc oxide applied topically to intact skin in man
Dermatologica
 , 
1990
, vol. 
180
 (pg. 
36
-
39
)
Australian Institute of Health and Welfare (AIHW)
Cancer Incidence Projections Australia 2002 to 2011
 , 
2005
 
AIHW Cancer Series no. 30. Canberra, Australia. Available at: www.healthinsite.gov.au/topics/Cancer_Statistics. Accessed July 2009
Australian Institute of Health and Welfare (AIHW)
Non-melanoma Skin Cancer: General Practice Consultations, Hospitalisation and Mortality
 , 
2008
Canberra, Australia
 
AIHW cat. no. CAN 32. Australian Institute of Health and Welfare, Canberra, Australia. Available at: www.healthinsite.gov.au/topics/Cancer_Statistics. Accessed July 2009
Baroli
B
Ennas
MG
Loffredo
F
Isola
M
Pinna
R
Lopez-Qunitela
MA
Penetration of metallic nanoparticles in human full-thickness skin
J. Invest. Dermatol.
 , 
2007
, vol. 
127
 (pg. 
1701
-
1712
)
Benech-Kieffer
F
Muelling
WJ
Leclerc
C
Roza
L
Leclaire
J
Nohynek
G
Percutaneous absorption of Mexoryl SX in human volunteers: comparison with in vitro data
Skin Pharmacol. Appl. Skin Physiol.
 , 
2003
, vol. 
16
 (pg. 
343
-
355
)
Casey
PS
Rossouw
CJ
Boskovic
S
Lawrence
KA
Turney
TW
Incorporation of dopants into the lattice of ZnO nanoparticles to control photoactivity
Superlattices Microstructures
 , 
2006
, vol. 
39
 (pg. 
97
-
106
)
Chan
TCK
Percutaneous penetration enhancers: an update
Proceedings 9th Biennial Conference of Perspectives in Percutaneous Penetration
 , 
2005
 
13 April 2004, La Grande-Motte, France, pp. 18–23
Cloquet
C
Carignan
J
Lehmann
MF
Vanhaecke
F
Variation in the isotopic composition of zinc in the natural environment and the use of zinc isotopes in biogeosciences: a review
Anal. Bioanal. Chem.
 , 
2008
, vol. 
390
 (pg. 
451
-
463
)
Cook
RD
Detection of influential observations in linear regression
Technometrics
 , 
1977
, vol. 
19
 (pg. 
15
-
18
)
Cross
SE
Innes
B
Roberts
MS
Tsuzuki
T
Robertson
TA
McCormick
P
Human skin penetration of sunscreen nanoparticles: in vitro assessment of a novel micronized zinc oxide formulation
Skin Pharmacol. Appl. Skin Physiol.
 , 
2007
, vol. 
20
 (pg. 
148
-
154
)
Dao
H
Kazin
RA
Gender differences in skin: a review of the literature
Gender Med.
 , 
2007
, vol. 
4
 (pg. 
308
-
328
)
Derry
JE
McLean
WM
Pharm
D
Freeman
JB
A study of the percutaneous absorption from topically applied zinc oxide ointment
J. Parenteral. Enteral. Nutr.
 , 
1983
, vol. 
7
 (pg. 
131
-
135
)
Deuranberg
P
Westrate
JA
Seidell
JC
Body mass index as a measure of body fatness: age- and sex-specific prediction formulas
Br. J. Nutr.
 , 
1991
, vol. 
65
 (pg. 
105
-
114
)
Dunford
R
Salinaro
A
Cai
L
Serpone
N
Horikoshi
S
Hidaka
H
Knowland
J
Chemical oxidation and DNA damage catalysed by inorganic sunscreen ingredients
FEBS Lett.
 , 
1997
, vol. 
418
 (pg. 
87
-
90
)
Environmental Working Group
Sunscreen Investigation. Section 4. Nanotechnology & Sunscreens.
 , 
2009
 
Fitzpatrick
TB
The validity and practicality of sun-reactive skin types I through IV
Arch. Dermatol.
 , 
1988
, vol. 
124
 (pg. 
869
-
871
)
Gabard
B
Elsner
P
Surber
C
Treffel
P
Stokes
R
Project sunscreen protection
Dermatopharmacology of Topical Preparations
 , 
2000
Berlin, Germany
Springer–Verlag
(pg. 
359
-
379
)
Gulson
B
Wong
H
Stable isotopic tracing-a way forward for nanotechnology
Environ. Health Perspect.
 , 
2006
, vol. 
114
 (pg. 
1486
-
1488
)
Inman
AO
Landsiedel
R
Wiech
K
Riviere
JE
Schulte
S
Monteiro-Riviere
NA
Assessment of UVB-damaged skin in vivo with sunscreen formulations containing titanium and zinc nanoparticles. Toxicologist abstract 2067
2010
 
p. 439
International Organization for Standardization (ISO)
Nanotechnologies—Terminology and Definitions for Nano-objects—Nanoparticle, Nanofibre and Nanoplate
 , 
2008
 
TC 229
Lademann
J
Richter
H
Schaefer
UF
Blume-Peytavi
U
Teichmann
A
Otberg
N
Sterry
W
Hair follicles—a long term reservoir for drug delivery
Skin Pharmacol. Appl. Skin Physiol.
 , 
2006
, vol. 
19
 (pg. 
232
-
236
)
Lademann
J
Schanzer
S
Richter
H
Pelchrzim
RV
Zastrom
L
Golz
K
Sterry
W
Sunscreen application at the beach
J. Cosm. Dermatol.
 , 
2004
, vol. 
3
 (pg. 
62
-
68
)
Lademann
J
Weigmann
H
Rickmeyer
C
Barthelmes
H
Schaefer
H
Mueller
G
Sterry
W
Penetration of titanium dioxide microparticles in a sunscreen formulation into the horny layer and the follicular orifice
Skin Pharmacol. Appl. Skin Physiol.
 , 
1999
, vol. 
12
 (pg. 
247
-
256
)
Larese
FF
D'Aostin
F
Crosera
M
Adamai
G
Renzi
N
Bovenzi
M
Maina
G
Human skin penetration of silver nanoparticles through intact and damaged skin
Toxicology
 , 
2009
, vol. 
255
 (pg. 
33
-
37
)
Monteiro-Riviere
NA
Baroli
B
Monteiro-Riviere
NA
Nanomaterial penetration
Toxicology of the Skin
 , 
2010
New York, NY
Informa Healthcare
(pg. 
333
-
346
)
Monteiro-Riviere
NA
Wiech
K
Landsiedel
R
Schulte
S
Champ
S
Inman
AO
Riviere
JE
In vitro penetration studies of four sunscreen formulations containing titanium and zinc nanoparticles in UVB- damaged skin. Toxicologist abstract 2068
2010
 
p. 439
Morgan
MEI
Hughes
MA
McMillan
EM
King
I
Mackie
RM
Plasma zinc in psoriatic in-patients treated with local zinc applications
Br. J. Dermatol.
 , 
1980
, vol. 
102
 (pg. 
579
-
583
)
Nadler
SB
Hidalgo
JU
Bloch
T
Prediction of blood volume in normal human adults
Surgery
 , 
1962
, vol. 
51
 (pg. 
224
-
232
)
Nakagawa
Y
Wakuri
S
Sakamoto
K
Tanaka
N
The photogenotoxicity of titanium dioxide particles
Mutat. Res.
 , 
1997
, vol. 
394
 (pg. 
125
-
132
)
Nanoderm
Quality of skin as a barrier to ultra-fine particles
 , 
2007
 
Final report. QLK4-CT-2002–02678
National Cancer Institute
Surveillance Epidemiology and End Results. Stats Fact Sheets. Melanoma of the Skin.
 , 
2009
 
Available at: www.seer.cancer.gov/statfacts/html/melan.html. Accessed December 2009
Nel
A
Xia
T
Mädler
L
Toxic potential of materials at the nanolevel
Science
 , 
2006
, vol. 
311
 (pg. 
622
-
627
)
Ngo
MA
Maibach
HI
Dermatotoxicology: historical perspective and advances
Toxicol. Appl. Pharmacol.
 , 
2010
, vol. 
243
 (pg. 
225
-
238
)
National Institutes of Health
Health Professional Fact Sheet. Office of Dietary Supplements.
 , 
2009
 
Available at: www.Od.nih.gov/FactSheets/Zinc.asp. Accessed December 2009
Nohynek
GJ
Antignac
E
Re
T
Toutian
H
Safety assessment of personal care products/cosmetics and their ingredients
Toxicol. Appl. Pharmacol.
 , 
2010
, vol. 
243
 (pg. 
239
-
259
)
Nohynek
GJ
Lademan
J
Ribaud
C
Roberts
MS
Grey Goo on the skin? Nanotechonology, cosmetic and sunscreen safety
Crit. Rev. Toxicol.
 , 
2007
, vol. 
37
 (pg. 
251
-
277
)
Olejnik
S
Algina
J
Generalized eta and omega squared statistics: measures of effect size for common research designs
Psychol. Methods.
 , 
2003
, vol. 
8
 (pg. 
434
-
447
)
Osmond
MJ
McCall
MJ
Zinc oxide nanoparticles in modern sunscreens: an analysis of potential exposure and hazard
Nanotoxicology
 , 
2010
, vol. 
4
 (pg. 
15
-
41
)
Pirot
F
Millet
J
Kalia
YN
Humbert
Ph
In vitro study of percutaneous absorption, cutaneous bioavailbility and bioequivalence of zinc and copper from five topical formulations
Skin Pharmacol.
 , 
1996
, vol. 
9
 (pg. 
259
-
269
)
Roberts
MS
Roberts
MJ
Robertson
TA
Sanchez
W
Thörling
C
Zou
U
Zhao
X
Becker
W
Zvyagin
A
In vitro and in vivo imaging of xenobiotic transport in human skin and in the rat liver
J. Biophotonics.
 , 
2008
, vol. 
1
 (pg. 
478
-
493
)
Ryman-Rasmussen
JP
Riviere
JE
Monteiro-Riviere
NA
Penetration of intact skin by quantum dots with diverse physicochemical properties
Toxicol. Sci.
 , 
2006
, vol. 
91
 (pg. 
159
-
165
)
Sadrieh
N
Wokovich
AM
Gopee
NV
Zheng
J
Haines
D
Parmiter
D
Siitonen
PH
Cozart
CR
Patri
AK
McNeill
SE
, et al.  . 
Lack of significant penetration of titanium dioxide from sunscreen formulations containing nano- and submicron-size TiO2 particles
Toxicol. Sci.
 , 
2010
, vol. 
115
 (pg. 
156
-
166
)
Schmid-Wendtner
MH
Korting
HC
The pH of the skin surface and its impact on the barrie function
Skin Pharmacol. Appl. Skin Physiol.
 , 
2006
, vol. 
19
 (pg. 
296
-
302
)
Steenberg
A
Malinovsky
D
Öhlander
B
Andren
H
Forsling
W
Engström
L-M
Wahlin
A
Engström
E
Rodushkin
I
Baxter
DC
Measurement of iron and zinc isotopes in human whole blood: preliminary application to the study of HFE genotypes
J. Trace Elem. Med. Biol.
 , 
2005
, vol. 
19
 (pg. 
55
-
60
)
Tagami
H
Monteiro-Riviere
NA
The stratum corneum in aged and photoaged skin
Toxicology of the Skin
 , 
2010
New York, NY
Informa Healthcare
(pg. 
153
-
166
)
Therapeutic Goods Administration, Australia
Safety of Sunscreens Containing Nanoparticles of Zinc Oxide or Titanium Dioxide.
 , 
2006
 
Available at: www.tga.gov.au/npmeds/sunscreen-zotd.htm. Accessed July 2006
Therapeutic Goods Administration, Australia
A Review of the Scientific Literature on the Safety of Nanoparticulate Titanium Dioxide or Zinc Oxide in Sunscreens.
 , 
2009
 
Available at: www.tga.gov.au/npmeds/sunscreen-zotd.htm. Accessed July 2009
World Health Organization
Dermal Absorption
 , 
2006
 
EHC 235. WHO Press, World Health Organization, Geneva, Switzerland
Zvyagin
AV
Zhao
X
Gierden
A
Sanchez
W
Ross
JA
Roberts
MS
Imaging of zinc oxide nanoparticle penetration in human skin in vitro and in vivo
J. Biomed. Opt.
 , 
2008
, vol. 
13
 pg. 
064031
 

Author notes

The authors certify that all research involving human subjects was done under full compliance with all government policies and the Helsinki Declaration.