1

Introduction

In the United States, skin cancer is the most commonly diagnosed form of cancer (CDC, 2021; Guy et al., 2015). Its occurrence is strongly associated with ultraviolet (UV) radiation resulting from sun exposure and other sources of UV light, particularly for people with fair skin (e.g., Ci y ska et al., 2021; Savoye et al., 2018; Wu et al., 2014). Sunscreens have been available since the 1930s to help mitigate these harms to human skin from the sun (Ma and Yoo, 2021). The active ingredients in sunscreen formulations—the UV filters—change the way the sun’s UV radiation interacts with the body by absorbing, reflecting, and/or scattering its rays, thereby reducing their ability to reach the skin. When used as directed, sunscreen products may reduce sunburn and skin cancer risk and slow the pace of skin aging (e.g., Stern et al., 1986; Young et al., 2017). As understanding of the linkages between UV radiation exposure, sunburns, skin cancer, and other health implications gradually evolved throughout the twentieth century, so did the availability of multiple, new, photoprotective ingredients and formulations to the sunscreen product market (Stiefel and Schwack, 2015). Box 1.1 distinguishes between the terms used to describe sunscreen and its composition and establishes the terms used by the committee in this report.

UV filters are found in a variety of hair and skin care products, cosmetics, insect repellents, household items, and commercial and industrial products, in addition to a broad selection of sunscreen formulations (Stiefel and Schwack, 2015). As a result of their multiple product uses, UV filters have been detected in aquatic environments and biota (e.g., see reviews by Cadena-Aizaga et al., 2020; Gago-Ferrero et al., 2012). The application of sunscreens for aquatic recreation provides a direct avenue for potential entry of UV filters into the aquatic environment. Their presence in the environment, while itself not indicative of exposure, bioavailability, or adverse effects, has led to a rapid increase in research in recent years on their potential environmental impacts (e.g., see reviews by Brausch and Rand, 2011; Carve et al., 2021b; Kwon and Choi, 2021; Ma et al., 2013a; Minetto et al., 2014; Rainieri et al., 2017; Figure 1.1). Concern about potential environmental impacts has led to the ban of some active ingredients in sunscreen products, including a legislated ban on the sale and distribution of sun protection products containing the UV filters oxybenzone or octinoxate in the state of Hawaii without a prescription (Hawaii SB 2571, Act 104). Similar bans exist in the U.S. Virgin Islands, Palau, Bonaire, Aruba, and some parts of Mexico.

Variability in UV filter physico-chemical parameters, potential sources and inputs, spatial and temporal concentrations, aquatic environment characteristics (e.g., varying salinities, hydrology, or co-stressors), species present, and species sensitivities makes it challenging to generalize about UV filters and their effects. Additionally, there is very little if any environmental exposure or hazard data on many UV filters, and there is not widespread agreement that available research sufficiently supports the conclusion that individual UV filters have or do not have

negative effects on aquatic organisms (e.g., Mitchelmore et al., 2021; Watkins and Sallach, 2021, for reviews of coral studies). The existing bans on certain UV filters may be considered precautionary in principle, in that they protect the environment against a potential threat now rather than wait for more data. This approach has raised questions, though, about the potential human health implications resulting from reduced availability of sunscreens with some widely used UV filter ingredients, and messaging about sunscreens overall.

This report reflects the twofold nature of the sunscreen challenge: Sunscreen is used as a critical tool for the prevention of UV skin damage and skin cancer; however, specific UV filter ingredients may also impact the health of aquatic environments, resident species, or ecosystem services.

STUDY TASK AND APPROACH

H.R.1865 - Further Consolidated Appropriations Act of 2020, which became Public Law 116-94, included a request that the U.S. Environmental Protection Agency (EPA) fund a study by the National Academies of Sciences, Engineering, and Medicine on the environmental impact of currently marketed sunscreens in the United States. This study would review the scientific literature about the potential risks to the aquatic environment from currently marketed sunscreen active ingredients as well as the potential public health implications associated with reduced use of currently marketed sunscreen ingredients for protection against ultraviolet radiation.

The task for the study, developed by EPA and the National Academies in response to the Congressional request, is found in Box 1.2. A committee of experts is tasked with outlining the information available about sources and inputs, exposure (i.e., exposure analysis), fates, and environmental effects (i.e., effects analysis) from UV filters in a manner that informs ecological risk assessments. It is not within the committee’s scope to conduct the final step of a risk assessment: the risk characterization. Therefore, the report does not contain estimates of risk posed to aquatic organisms from any UV filter. Instead, the committee provides the state of science informative to the

issue, including high-level assessments of the quality, reliability, and uncertainties in the data currently available, and identification of knowledge gaps to fill to improve risk assessments. The committee is also tasked with outlining what is known about human use of sunscreens and how potential changes in usage may affect human health. Knowledge gaps are highlighted in this case as well.

The chapters of the report are structured to inform an eventual ecological risk assessment (ERA) and also to examine implications of potential changes in sunscreen usage on human health. Chapter 2 describes the characteristics of sunscreens and the UV filters that are within the scope of the report. Throughout the report, the committee also describes instances where information is available about degradation products of UV filters; however, this information is sparse. Chapter 3 characterizes the problem explored by the ERA process by reviewing information about sources, inputs, settings, and ecological receptors of UV filters from sunscreens in the environment. Chapter 4 describes the fate processes of UV filters in the environment and the resulting measured concentrations in water and sediment. Chapter 5 describes the state of knowledge on exposure and accumulation of UV filters in aquatic biota. Chapter 6 reviews potential effects to aquatic organisms and ecosystems and summarizes the toxicity data of utility for an ERA. Section 2 of the statement of task—a review of health benefits of sunscreens, potential changes in sunscreen usage based on environmental concerns or changes in availability of certain ingredients, and potential resulting health effects—is included in Chapter 7. The task did not call for a review of sunscreen’s safety for humans, so this topic is included as a high-level description as part of the discussion of its health benefits, but not given a comprehensive review. Each chapter includes an assessment of priority knowledge gaps to fill. Last, Chapter 8 provides a narrative that ties the findings from each chapter together and summarizes the committee’s advice. This report is informed by a review of the literature as well as consultation with researchers, human health providers, private sector representatives, and environmental managers during public information gathering meetings. Research on sunscreens is a rapidly expanding field. The committee comprehensively reviewed information that was accessible to them through February 2022, and reviewed additional information as possible, recognizing that more studies will likely be available even before this report is final.

The scope of the study is limited to the United States. The information in the report is meant to be applicable to the ecosystems and species found within U.S. jurisdictions. However, informative research from outside the United States is included, especially when domestic data were scarce. The scope of the study is also limited to UV filters currently marketed in the United States. Currently, 16 active ingredients are allowed by the U.S. Food and Drug Administration (FDA) for use in sunscreens sold in the United States, plus an additional proprietary

UV filter, ecamsule, approved for use in limited products (Table 1.1). These chemicals and consideration of their degradates, where feasible, constitute the committee’s scope. It is beyond the scope of the study to review any UV filters not listed in Table 1.1, including UV filters not currently available in the United States even though they could be found in U.S. waters through use by tourists from other countries.

While it is the committee’s task to focus on contributions from sunscreens specifically, these ingredients are also found in other products and may make their way into the environment from these sources as well. The report includes the information available specific to sunscreens, but refers to where records of occurrence may be confounded by potential contributions from other products. The report also does not address the inactive ingredients

TABLE 1.1 UV Filters Currently Marketed in the United States

^a As identified in the EPA CompTox Chemicals Dashboard.

^b As compiled by Mitchelmore et al. (2021).

^c May be found in nano or bulk form.

that constitute the remainder of sunscreen formulations. However, it is possible that there are situations where the influence of the inactive ingredients is an important consideration for understanding the physico-chemical characteristics of the UV filters and their fate and potential environmental effects.

ECOLOGICAL RISK ASSESSMENT

The challenge for understanding the risks from UV filters to aquatic environments is in determining whether and under what conditions individual UV filters or mixtures of UV filters are a risk to organisms and ecosystems—either alone or in combination with other environmental stressors—and where these conditions might occur. Aquatic ecosystems vary greatly in physical processes that influence exposures as well as in the spatial distribution of ecological stressors. Understanding the influence of UV filters introduced into the environment is further complicated by the fact that freshwater, estuarine, and coastal ecosystems are under stress from a variety of factors including climate change, pollution, physical damage, water abstraction and regulation, invasive species, and disease.

ERA is an approach that considers both the possible effects as well as the likelihood that organisms may be exposed to a stressor. ERAs for chemicals involve integrating information about exposures in the environment (exposure assessment) with information about adverse effects associated with exposure (dose-response, effects, or hazard assessments). This integration is accomplished using methods that range from simple comparisons of one value to another (e.g., calculating a Hazard Quotient) to population-level and probabilistic risk assessments. The choice of method depends on regulatory frameworks and the risk-related questions that need to be answered to inform management decisions established during problem formulation. This quantitative step provides risk values that can be judged against risk-related criteria and guidelines. Within an ERA, toxicity tests are used to identify relationships between UV filter concentration and lethal or various sublethal biological endpoints following exposures that may be short-term (e.g., hours to days), variable (e.g., pulses or diurnal variations), and/or longer-term (e.g., to substantial portions of a lifetime or a critical life stage). Combinations of acute and chronic toxicity tests are typically used to match the nature of the exposure to the potential effects of concern; designs can also be employed that simulate the variable nature of the exposure regime. Thus, an ERA would identify particular exposure settings in which sunscreens could be the cause of ecological impacts—such as the concentrations of the UV filters, the physical and chemical properties that affect their partitioning, bioavailability, bioaccumulation and persistence in the environment, and hydrodynamic properties that influence their residence time in a particular location—that may be expected to occur. Additionally, standardized toxicological test methods have been developed to ensure comparability and appropriateness of data for use in ERAs, especially in a regulatory setting, though nonstandard toxicity tests also provide ERA-relevant data (e.g., additional nonstandard test species or biological effect endpoints) and/or insights into modes of action, and additional species or ecosystem endpoints of concern.

Both exposure and effects estimates will be accompanied by a degree of uncertainty driven by gaps in the research. Figure 1.2 illustrates how increasing amounts of information can lead from lower-tier screening analyses to higher-tier ERAs that have lesser degrees of uncertainty. While an ERA can be conducted using only quantitative structure–activity relationship (QSAR) information and models of exposure, it will result in a risk characterization with a larger degree of uncertainty than an ERA that includes acute and chronic toxicity tests across a range of taxa and model ecosystems. In some cases, screening-level analyses are sufficient to make conclusions about risks in relation to informing management; in other cases, more information gathered at higher tiers is needed to provide the technical support for management decisions. A tiered approach also helps guide allocation of resources to refining estimates of exposure and effects.

Uncertainty in an ERA is also considered by making adjustments in the calculations or interpretations. This could involve applying uncertainty or adjustment factors to the effects data, thereby increasing the risk magnitude or by comparing quantitative risk estimates to Levels of Concern. In some cases, probabilistic estimates of risk may be derived that incorporate distributions of exposures and effects information. Finally, ERAs may also rely on what is referred to as “weight-of-evidence” when there are many lines of evidence on effects and exposure that are being integrated into the assessment.

These degrees of uncertainty explain why risk characterization in an ERA is rarely as simple as comparing one quantitative estimate to another and why the committee does not make direct comparisons between exposure

and effects estimates. The committee did not undertake any of the approaches for considering uncertainty and was asked not to conduct a risk characterization as part of their task, but it is important to note that this will be a necessary next step for undertaking an ERA with the data available on UV filters. Filling the knowledge gaps highlighted throughout the report can serve to move to increasing tiers of complexity in an ERA with reduced degrees of uncertainty.

REGULATION OF SUNSCREENS IN THE UNITED STATES

Currently Marketed Sunscreens

The scope of ingredients under review in this report is driven by the regulatory setting for sunscreens in the United States, where they are regulated by FDA as over-the-counter (OTC) drugs. In Europe and other parts of the world, sunscreens are regulated as cosmetics, a possible explanation for the wider range of available ingredients (Reisch, 2015). OTC drug categories, such as sunscreens, are reviewed as part of the FDA OTC monograph process. A monograph establishes the conditions, routes of administration, labeling, and testing requirements under which an OTC drug is considered “generally recognized as safe and effective” (GRASE). The GRASE designation refers to human safety and does not consider potential effects on the environment. Information that supports a GRASE determination is requested from sunscreen formulators, though data can also come from published scientific literature, if available. Sunscreens that are not covered by the monograph may be approved via the new drug application process, which approves specific uses (in this case, sunscreen formulations) for an ingredient but does not constitute broad approval for the ingredient itself (i.e., a specific UV filter). In the case of sunscreens,

the French personal care product company L’Oréal has received approval to market certain sunscreen products with the UV filter ecamsule; however, the ingredient is not available for use in any other products without further approval via the new drug application process.

A 1999 monograph established the conditions for GRASE and recognized 16 UV filters as meeting requirements to be considered within this classification. The rule was stayed shortly after it was finalized to address UVA and broad-spectrum requirements. However, the Coronavirus Aid, Relief, and Economic Security (CARES) Act in 2020, which reformed the FDA monograph process overall, established the 1999 stayed rule as the deemed final order. The CARES Act also required FDA to propose an administrative order revising the deemed final order within 18 months of enactment, which was done in September 2021 (86 FR 53322; Final Administrative Order OTC000006). The new order has not been finalized as of July 2022.

Prior to the CARES Act, a new sunscreen monograph containing information about active ingredients, dosage forms, SPF and broad-spectrum requirements, labeling, and final formulation testing and record keeping had been proposed—but had not been finalized—in 2019 (84 FR 6204). The requirements in the new proposed order are “substantively the same as those that the [FDA] described in the 2019 Proposed Rule” (86 FR 53322). The proposed 2019 monograph identified zinc oxide and titanium dioxide (in any form) as GRASE, aminobenzoic acid and trolamine salicylate as not GRASE, and the remaining 12 UV filters covered by the monograph as not having sufficient information to make a determination as to whether they are GRASE or not. FDA does not typically conduct clinical trials and has requested relevant information from the sunscreen industry. New information was invited as part of the public comment period for the proposed order. Aminobenzoic acid and trolamine salicylate have already been phased out of use due to known adverse reactions in humans. They are still included in the report to the extent that information is available because, until the new order is finalized, they may theoretically still be included in sunscreen products.

Managing Sunscreen Impacts to the Environment

UV filters are not currently regulated as environmental pollutants by any U.S. federal statutes. In May of 2021, FDA announced a Notice of Intent to prepare an Environmental Impact Statement (EIS) to assess environmental effects that may result from the proposed monograph order (86 FR 26224). The National Environmental Policy Act (NEPA) requires an evaluation of all major federal actions that may have an adverse effect on the environment and of any possible alternatives. The EIS will be limited to potential effects from oxybenzone and octinoxate because, as explained in the notice, these are the ingredients for which the most concern had been raised in comments on the 2019 proposed monograph. NEPA cannot grant an agency the authority to act outside its existing authority, and the GRASE determination, for which FDA does have authority, does not consider environmental effects. Sunscreen labeling rules defined by FDA (or any other agency) also do not specify claims about environmental fates or effects (e.g., “reef safe” or “biodegradable”).

The ERA process described earlier in this chapter is informative to EPA in determining whether a chemical needs regulatory attention. ERAs are specific to an individual chemical, so in the case of sunscreens, ERA would be conducted for individual UV filters. Environmental laws that could be applied to UV filters include the Clean Water Act and the Endangered Species Act. Under the Clean Water Act, EPA recommends national ambient water quality criteria for pollutants for adoption by states and tribes based on the highest concentration of a pollutant in water so that it is not expected to pose a risk to the majority of aquatic species or to humans. The Endangered Species Act requires that no federal actions, such as the determination of water quality criteria, jeopardize species listed as threatened or endangered or destroy or adversely modify their habitat. The U.S. Fish and Wildlife Service (USFWS) has jurisdiction over freshwater and land animals while the National Oceanic and Atmospheric Administration (NOAA) has jurisdiction over marine and anadromous species. Listing determinations take into consideration the range of threats to a species, and recovery plans describe the conditions that will be required for a species to recover. USFWS and NOAA consult with agencies like EPA to determine if changes need to be made to federal actions to prevent jeopardy to listed species. Chapter 6 of this report includes recognition of listed aquatic species that could be vulnerable to impacts from UV filters.

In the absence of federal regulation, other management measures at the state and local level have been implemented or attempted based on perceived or potential environmental impacts. As noted earlier, legislation has been

passed in the state of Hawaii to ban the sale of sunscreen products containing oxybenzone and octinoxate without a prescription. A ban was passed at the local level in Key West, Florida, but SB 172, the Florida Drug and Cosmetic Act, subsequently preempted any local ordinances pertaining to OTC drugs and cosmetics. Non-regulatory control measures have included education campaigns¹ and free sunscreen dispensers at beaches for formulations assumed to have minimal or no environmental impact.²

DATA QUALITY ASSESSMENT IN DECISION MAKING

As part of the review of the literature, the committee considered not only the availability but also the utility of the information available for the purposes of conducting ERAs and assessments of the potential impact of changes in sunscreen usage on human health.

Considerations for utility for use in an ERA fall into two primary categories: reliability (an assessment of the study’s repeatability based on strength of documentation) and relevance (the ability to directly or indirectly use the study in an assessment with the purpose of addressing specific protection goals and ultimately regulatory decision making). Studies may be reliable, but irrelevant (e.g., a thorough and objective study which measured levels or responses in a medium/organism to which the chemical will never be found or organism exposed to) or not reliable, but relevant (a toxicity study that is conducted outside of regulatory guidelines but on a very important ecological or commercial species). Other studies will be variations within those categories (i.e., reliable with reservations) and criteria are used to further assess their utility and guide ERAs (e.g., Klimisch et al., 1997; Moermond et al., 2017). Martin et al. (2019) provided a broad set of goals for data sharing in the peer- and non-peer-reviewed literature. They emphasized the importance of transparency, completeness, descriptive clarity, and objectivity in regard to information sharing for the purpose of regulatory decision making. Some areas of science have well-known principles, but these are also not universally adopted for all endpoints, exposure media, (eco)toxicity, and physical chemistry. Thus, assessments of data quality generally vary greatly by discipline, and individual judgments (study-by-study, endpoint-by-endpoint) for acceptability are manifest in this document.

This report covers a broad range of scientific disciplines, with various degrees of standardization of methodologies at the national and international level. As a result, approaches for reviewing methodologies are described within each chapter. Physical chemistry is known to drive many environmental attributes and as such, these have been standardized in addition to guidance for toxicity testing variables under the auspices of both regulatory (e.g., EPA) and non-regulatory (e.g., OECD, ISO, ASTM) standardization organizations. Collectively, physical chemistry, environmental test methods (fate and effects), and human safety test methods are generally requested to be of a standard method execution when put forward for regulatory purposes. This often leaves nonstandard (including tests on taxa without standard methods), novel, and investigatory methods (and their associated observations) on the “outside looking in” (Martin et al., 2019). Such studies may provide useful insight into the behavior of chemicals in the environment and how they may interact/impact exposed organisms (i.e., identifying mechanisms of action including novel pathways). They may also provide improved weight of evidence to derive conclusions regarding organism responses that standard studies may not. From the human health side, population health and behavioral health/social science are not governed by national or international standards, though there are best practices that can be applied.

The committee considered studies based on generally accepted concepts of study reliability and relevance when developing overviews of each of the subject matter areas. In cases where guidelines exist, these guidelines were used as considerations to evaluate the relevance and reliability to studies for use for ERAs. Additionally, nonstandard studies or cases where standards do not exist are described accompanied by a description of their utility. There are also studies that are exploratory in nature and designed to understand potential toxicological or physiological processes. While these studies may not be directly useful for an ERA, they can point toward areas where further knowledge is needed and have thus been included in the present report.

___________________

¹ See https://www.nps.gov/subjects/oceans/reeffriendlycampaigngraphics.htm.

² See https://kohalacenter.org/reef-friendly-sunscreen-dispensers-launched-at-kahaluu-beach-park.

This page intentionally left blank.