The FDA New Proposed Final Rule On Sunscreens: Considering Safety



It should be noted that I am no lover of bad news. Yet, sometimes bad things happen to good people. Sometimes good intentions have bad consequences.

On February 26th, 2019, the Food And Drug Administration published a proposed final rule on Sunscreen Drug Products for Over-the-Counter Human Use. I’ve previously summarized rule’s proposed the changes to sunscreen regulation, which can be read here.

As I set forth in my summary, the proposed rule does a lot of work. I am, on the whole, a big fan of the proposals set forth in the FDA’s proposed final rule on sunscreen drug products for over-the-counter human use. However, the framing of the issue of safety in regards to sunscreen active ingredients is, for me, bad news. Today, I attempt to explain why.

One piece of legislation which I find needs robust rebuttal is the FDA’s definition of safety in regards to sunscreens. In this post, I seek to explain and contextualize what the FDA defines as safe in regards to sunscreens, and provide my commentary on their logic when I find it problematic.

I will begin by outlining all of the safety testing detailed in the proposed final rule. I will conclude with my comments.

Why is more data on sunscreen ingredients necessary?

  • In an effort to comply with the Sunscreen Innovation Act of 2014, the FDA has provided a framework for determining the safety of active ingredients in sunscreens.
  • The following data are required to indicate the safety of sunscreen filters which are not currently on the Sunscreen Monograph (like Tinosorb S, Uvinul T 150, on and on), as well as the sunscreen ingredients currently on the Monograph which the FDA has given the designation of Category III. These filters are:
    • Avobenzone
    • Cinoxate
    • Dioxybenzone
    • Ensulizole
    • Homosalate
    • Meradimate
    • Octinoxate
    • Octisalate
    • Octocrylene
    • Oxybenzone
    • Padimate O
    • Sulisobenzone
  • The FDA requests more data on these filters because they claim to lack research that fulfills the testing outlined below.
  • It is clear to me that these filters have been deemed suspicious due to the lobbying arm of notable “consumer advocacy” groups.
  • This is a disappointing turn of events. These filters are deemed safe in virtually every country in the world, including regions with much stricter rules regarding the safety of cosmetic ingredients. There has never been evidence of a causal link between the use of these sunscreen filters, which have been marketed since the 1970s, and any human health harm. In fact, they’ve served the public well in the prevention of non-melanoma skin cancers.
  • Calling their safety into question is not unfounded, as it is important to understand the ramifications of these filters, if any, on human health. Yet, the question remains: why look at their safety now, after 40 years of marketing, when there is no apparent risk?

The Required Data to Determine Safety

  • Since sunscreens are regulated as over-the-counter drugs, the FDA has determined that it is necessary to prove that all sunscreen active ingredients are safe to use as directed.
  • The FDA writes that its recommendations for safety are often linked to animal studies, in which animals are exposed to certain concentrations of a substance being studied. The highest animal no observed adverse effect level (aka, the highest dose where nothing bad happens, or NOAEL) is effectively ‘rounded up’ to humans.
  • Since the human sensitivity to the drug is often unknown, and animal models cannot predict human sensitivity with complete accuracy, the FDA allowance for human use is generally well below the concentration in which adverse effects were found in animal models.
  • The FDA seeks both clinical and non-clinical data in regards to sunscreen safety.
  • The Clinical data should address:
    • Irritation (does it irritate the skin)
    • Sensitization (does it sensitize the skin)
    • Photosafety (does it cause problems when interacting with sunlight)
    • Bioavailability (does it absorb through the skin, and if so, to what quantity)
    • Any other adverse effects
  • The non-clinical data will use animal models to address:
    • Carcinogenicity (does it cause cancer)
    • Reproductive Toxicity (does it cause problems for people who are pregnant or seek to become pregnant)
    • Other effects (the FDA cites ‘hormonal disruption’ the claim frequently tossed around by critics of organic sunscreens)
  • The FDA wants studies done on all sunscreen filters in isolation, which is to say, they want one filter tested at a time. If that filter is safe on its own, the FDA will assume that it is safe to use with other filters determined safe, unless there is some data suggesting that there are filters which may act differently in conjunction, at which time that combination of filters would be subject to the same assemblage of testing to determine safety.

What the studies need to look like

Dermal Irritation & Sensitivity

  • The tests should consist of the highest concentration of the active ingredient for which a safe determination is sought (ie, if L’Oréal wants approval of Ecamsule at 10%, these studies should be done with 10% Ecamsule).
  • The FDA suggests that patch testing on the back is fine (which is pretty standard)
  • These studies should be conducted in three phases:
    • The induction phase (3 weekly applications for 3 weeks)
    • The rest phase (no product application for 10 to 14 days)
    • The challenge phase (patch applications to new sites for 48 hours)
  • These studies will be evaluated in terms of erythema (redness), edema (swelling), papular response (pimples or other bumps), or skin erosion (self explanatory, albeit grossly put).
  • The writing here seems to suggest that the FDA isn’t particularly concerned with dermal irritation, and believe to have fine data on dermal irritation based on the marketing history of sunscreen products. It seems as though the FDA wants confirmatory data on file, as it were.

Photosafety Studies

  • The active ingredients should be tested to see if there are photoallergy or phototoxic reactions to UV and Visible light.
  • Photoallergy tests can be patch tests, conducted as above.
  • Phototoxicity tests include some specific parameters for testing, meant to see if the filter, after exposure to UV light, causes skin irritation.
  • Again, the writing here seems to indicate that the FDA has little concern that the sunscreen ingredients currently in use will pass these tests.

Human Absorption

  • The FDA wants to know if the ingredients in sunscreen make it through the skin, and if so, to what extent.
  • In order to determine this, the FDA requests that a maximul usage trial (MUsT) be conducted for each sunscreen filter.
  • The MUsT should be conducted in a variety of sunscreen formulations, as different formulations (spray, cream, gel, etc) are likely to have different absorption profiles.
  • The formulations should have the sunscreen filter being evaluated as the only active ingredient, at the highest concentration being sought for approval.
  • The formulation should be applied to 75% of the surface area of healthy, in tact human skin.
  • The FDA writes that “An important consideration for designing a MUsT is that it should include testing for a duration that allows for the attainment of steady state levels to ensure that maximum penetration of the ingredient has taken place and to optimize the chances of the ingredient being detected” (2614). This is to say, the FDA wants the formulation reapplied multiple times, per the instructions for using a sunscreen.
  • This is important: if the ingredient is absorbed at a concentration of 0.5 nanograms per milliliter of “receptor fluid” (which, according to the first set of tests, is blood plasma), the ingredient should be subject to non-clinical safety testing
  • This is important: the FDA writes “The threshold value of 0.5 ng/mL is based on the assessment that the level would approximate the highest plasma level below which the carcinogenic risk of any unknown compound would be less than 1 in 100,000 after a single dose” (6215).
  • So, if the sunscreen fails the MUsT, the FDA finds that it is in the realm of possibility that the sunscreen could be as carcinogenic as the most carcinogenic known things.
  • A MUsT was conducted on avobenzone, oxybenzone, octocrylene, and ecamsule, and they all failed. This is what all the newspapers wrote about in a fear-monger kind of way. My commentary on that mess will follow, eventually.

Carcinogenicity Studies

  • The FDA says that in order to be deemed “safe,” sunscreen active ingredients need to be tested on rats or mice for two years to determine tumor development, if any.
  • If the sunscreen filter fails the MUsT testing, the FDA will also require a second carcinogenicity study, to be conducted on a different species than the topical study, with a method of application that forces systemic exposure (so the ingredient would need to be administered orally or via injection). If this fails to exhibit any changes to carcinogenicity signals, the ingredient will be deemed “safe.”
  • There seems, to me, to be a broad jump between systemic absorption at a level which is possibly carcinogenic and the need to feed a given sunscreen ingredient to rats. If systemic exposure through the skin is the concern, I do wish to understand why topical application of the ingredient on hairless rats wouldn’t be more accurate way of determining a filter’s carcinogenicity in real-life conditions.

Developmental and Reproductive Toxicity Studies

  • If the filters fail MUsT testing, the ingredients should be administered orally to pregnant rats / rabbits, and they will be evaluated, as will their offspring, in a variety of ways to determine which harm was caused, if any.


I acknowledge and appreciate the serious rigor that the FDA lays out to test the safety of sunscreen actives. Indeed, if a filter passes these tests, there can be absolutely no debate regarding their potential safety. However, I find this method of determining “safety” flatly insane.

I supply the following three arguments as evidence to my claim:

  1. The maximum use trials bear no resemblence to the way that most people use sunscreen, because the adequate dose of sunscreen is poorly communicated, which undermines the findings of any research completed.
  2. The testing procedures outlined above fail to provide evidence for the impact of the systemic exposure to sunscreens in actual use.
  3. The public relations impact of the research executed to evince the claims above will only raise public fear of sunscreens, thereby lowering likelihood of compliance with a set of behaviors that have the ability to reduce the likelihood of skin cancer.

The maximum use trials bear no resemblence to the way that most people use sunscreen

The testing set forth in the guidelines for a MUsT are not without merit, and the net effects of sunscreen use as directed (2mg / square cm of skin, reapplied every two hours) must, of course, be known. Yet, it must be stated that the dosage in the MUsT is not communicated adequately to the public. Indeed, the Drug Facts on sunscreen products read “Apply liberally 15 minutes before sun exposure.” “Liberally,” in this context, fails to give an adequate measure of the amount of product to be applied. It is as if Advil directions read “take an adequate amount of medication every 4-6 hours as long as symptoms persist.” If this were the case, the consumer would be likely to use an incorrect amount of Advil, and that would be the FDA’s fault, not the consumer’s.

This is to say, if the FDA seeks to understand what the ramifications of an adequate dose of sunscreen are, they must first communicate the adequate dosage of sunscreen is.

I find that this failure to indicate the correct amount of sunscreen to be used has led to consumers using an inadequate amount of sun protection broadly. If a consumer is using half of the recommended dose of sunscreen, they will (1) receive inadequate protection and (2) fail to use the product in a way that makes meaningful the findings of any research done. As such, this whole exercise fails to bear actual significance, and that is a terrible shame.

The testing procedures outlined above fail to provide evidence for the impact of the systemic exposure to sunscreens in actual use

While I accept the FDA’s given threshold of 0.5 ng sunscreen ingredient / 1 mL fluid, it must be stated that this is the minimum concentration likely to have any carcinogenic effect at all. With 50 years of marketing data on all of the sunscreen ingredients currently on the Monograph, I think we have a good understanding of the fact that many of these sunscreen ingredients are absorbed systemically, and that there has never been a clear link between any of them and any actual human health hazard, beyond an occasional allergic contact dermatitis. Jumping from “absorbed at a level to which the strongest known carcinogen is carcinogenic” to “let’s feed oxybenzone to rats” seems to fail to address this concern adequately. Both perspectives seem, to me, to measure the potential to cause harm in the most dramatic ways possible. I find that framing the issue of safety in this way fails to address actual concerns of sunscreen users, and that is really disappointing.

The net public relations impact of the research executed to evince the claims above will only raise public fear of sunscreens, thereby lowering likelihood of compliance with a set of behaviors that have the ability to reduce the likelihood of skin cancer.

It is my firm belief that this research will, without question, lead to a public distrust of sunscreening products broadly. This is terrible news.

Speaking of terrible news, I feel like I can’t blink my eyes without seeing a misinformed beat reporter taking the word of the lobbing arm of the Environmental Working Group over that of the medical community, and take the FDA’s request for more data out of context. It’s really horrible, and it’s scaring people.

It is incumbent upon the FDA, the medical community, and industry to frame this request for further research in the context of the clear and abundant ramifications of sunscreen use. In my estimation, this has been done poorly. No one is saying “using sunscreens reduces the risk of skin cancer, and one shouldn’t avoid using the sunscreen filters for which more testing is required for fear of some potential harm about which we don’t yet know.”


In Conclusion

Sunscreens are wonderful things. I am so grateful to live in a world where we can protect our skin from complete carcinogenicity of UV radiation. I cannot believe that America – the land of the free – the country in which I proudly claim citizenship – has more restrictive rules on sunscreen products than the People’s Republic of China. It is my firm belief that this draconian aspect of what is otherwise good and necessary legislation will negatively impact the lives of millions of Americans.

I think the worst part of all of this is the fact that there is no clear reason why these sunscreen ingredients are being reconsidered. It isn’t as if people are suddenly dropping flat dead after 40 years of sunscreen use. The FDA says they are doing this to comply with the Sunscreen Innovation Act, but that legislation clearly sought to make the United States get its shit together in regards to approving the fantastic sunscreen ingredients that are available in every other developed country in the world (including Russia, where they don’t even have sun for half of the year). It’s a shame that the FDA sought to respond to that legislation with more bureaucratic red tape.

Of course, the clear counterargument to all of this is that zinc and titanium based sunscreening products can be used, and that they are widely regarded as both safe and effective. I don’t accept this as the end of the sunscreen discussion. This is a topic that deserves its own post, but I’ll briefly say: mineral sunscreens don’t work for everyone, and I don’t think people should need to choose between looking like Casper the Ghost and having skin cancer.

I find myself to be quite depressed by this aspect of this proposed final rule. I am depressed about it because I actually am very grateful for the regulations set forth in the rest of the rule. However, they will fail to yield any positive impact if a consumer cannot choose a sunscreen product that is acceptable.

I may be misinformed, and I may be being overly dramatic. I don’t know. I am just reading the proposed rule and thinking about it a lot. If anyone can offer any assurances which negate my concerns raised above, I’d welcome the good news. I’d also welcome advisory regarding whether or not there is something I can do to lobby for any of this (Hello, L’Oréal / BASF / et al, HMU, I am willing to try and help with something!!!)

Yours in Broad Spectrum Protection, D