May 26, 2015
Sunscreens
Well Memorial Day Weekend has come and gone and with the weather heating up and kids going out, it’s time to finally write up the results of reading a way too-big stack of papers on sunscreen.
Of all the products we frequently use, I'm most afraid of sunscreen. The reason is simple: sunscreen contains huge quantities of chemicals known to penetrate skin, affect biology (especially during development) and be secreted into milk. These chemicals act like hormones, alter reproductive and thyroid function, they can cause severe allergies and they're associated with endometriosis. Great thing to smear on your kids right? And guess what? Studies show that using sunscreen doesn't affect if you get cancer! (I'll get to that.)
Results in summary:
Figure 1. How often each brand was mentioned in EWG's Hall of Shame for sunscreens. Sorry Coppertone, you're scary!
1) Organic sunscreen chemicals really worry me, to the point our family won’t use them. [See 2 below; there are alternatives! Note as well that "organic" does NOT mean "grown with no pesticides" but rather "chemicals made from carbon." Dry cleaners intentionally confuse people with that all the time.] Many of these chemicals are known to cause health problems like allergic reactions, many are absorbed in significant quantity through the skin and many are known to be able to disrupt our bodies’ endocrine systems. Most importantly, their long-term health effects—especially on developing minds and bodies—aren’t known or, for that matter, studied. (This is an exceedingly difficult thing to study btw; see below.)
2) Mineral sunscreens (titanium dioxide and zinc oxide) don’t scare me at all. (Incidentally, the Environmental Working Group, EWG, came to basically the same conclusion. Their report will scare you, and it should.) They’re actually the only ones our family will use. We buy them by the gallon.
3) Since probably very few people want to memorize all the good/bad sunscreens, a few easy rules of thumb: if it has an active ingredient other than titanium dioxide or zinc oxide, just don’t buy it. EWG has a cell phone app you can download but I actually can’t figure out why one would want to use anything but mineral sunscreens. Finally, I put together a histogram by brand of EWG’s Hall of Shame for sunscreens (Fig. 1). I don’t know who’s doing formulations at Coppertone (that was the brand that caused my wife to have some huge body-wide case of almost hives…) but I’m not touching anything from them or CVS. And Neutrogena, I thought you were a good brand??? What gives? In any case, just stick to the mineral sunscreens.
4) Sunscreens work in controlled settings (like in labs and on mice).
5) This doesn’t always hold true in the “real world”: people use like, 1/2, 1/3 of 1/4 of the recommended amount, andwe also don’t reapply often enough.
6) Figuring out the relationship between skin cancers and sunscreen is annoying and highly complex. In particular, the people who are most likely to get cancer are often exactly the ones using sunscreen and people who use sunscreen go out in the sun more. These facts (and others) make it really hard to tease cause and effect apart.
7) Partially because of the above, the results on cancer and sunscreen use are literally all over the map: some studies say sunscreen use is associated with an increased risk of cancer and other studies say the reverse. The end conclusion if you look at all the studies? In terms of cancer, there’s no statistically significant difference with sunscreen use.
8) The above non-significant result really comes down to human behavior (see #4), and the rule is really simple: the more time you spend in the sun, the more damage your skin will accumulate and if you get enough damage and get unlucky, you’ll get cancer. If you always, ALWAYS protect yourself with generous amounts of sunscreen, never miss any spot and always reapply it all the time so you never ever get red or even tan, then yes, I’m pretty sure you can rely on sunscreen to sun-induced cancer. And how many of you are going to try to get a tan this weekend?
More info
Why the organic chemical sunscreens worry me
In all the below, we have to remember there are a bunch of different active organic chemical UV screens in sunscreens (plus a bunch of non-active ones) and some are far worse than others. That said, take a look at EWG’s guide and you’ll see why I’m not breaking this down by chemical. Obviously, these are generalizations and the complete lack of these same concerns with mineral sunscreens is why my family only uses them:
1) Many of these chemicals are efficiently absorbed straight through the skin. For example, from 1% to 9% of the infamous oxybenzone (benzophenone-3) applied to skin is absorbed through it. 2% of dermally applied doses ends up in urine and breast milk.
2) Many of these chemicals mimic hormones (see below) and have particularly tricky, interactive, long-term and hard to study biological effects. Consequently, I’m not sure we’ll spot changes these chemicals cause. Namely, hormones often have so-called “U-shaped” activity curves. This means that at high concentrations, the activity is high (e.g. causes cancer), at medium concentrations the activity is less (sometimes much less) and then at low concentrations, the activity goes up again! [Note that the U can also be upside down so that only intermediate concentrations have significant effect.] Actually, in the lab (i.e. controlled settings), we’re pretty good at figuring out these sorts of activity curves. However, the effects are dependent on a whole host of other factors, age (see below) and gender being two really important ones, but also genetics (how sensitive you are) and environment (what else are you’re being exposed to). We’re not collecting enough data in enough detail to come anywhere close to deciphering these parameters.
If you use these chemicals (and probably you are), it might be that you get “lucky,” and by chance end up in some region where you’re not causing your kids hormone disruption. And of course, probably you’ve got a couch and pajamas loaded with fire retardant, and maybe you don’t end up that lucky. Unfortunately, the effects of hormone disruption might not be seen for years, or, because it’s an uncontrolled experiment, we may not even know they happened. In either case, changes especially during development may be irreversible (see below).
We know these effects happen, we know that they’re complex and we know our environment is loaded with a bunch of similar hormone mimics. Because these effects are so difficult to research, and because I have no desire to perform uncontrolled developmental biology experiments on our family, my “risk sensors” light up bright red. For our family, it’s simple: we just won’t go near non-mineral sunscreens.
3) Sunscreen contains stonkingly high levels of these chemicals and we absorb enough to probably dysregulate our biology. The maximum allowable concentration of active ingredient in sunscreen depends on the exact chemical and country. It’s usually between 2% and 15% and you have to keep in mind that the limits apply on a per-chemical basis (so, for example, you might have one chemical at 5% and another at 10% for a total of 15% possible hormone mimic).
That paper in the above link suggests an average application of 18 g and for purposes of understanding how much chemical is in sunscreen, let’s assume it’s a 10% formulation. Thus, of the 18 g, 1.8 g (or 1800 mg) is active chemical: like a big spoonful. Let’s take the 2% dermal absorption rate from above and we get 36 mg of chemical absorbed into the body. Now that falls right in line with the range of most medications (say a 5 mg – 500 mg dose) or, for the “average” 60 kg person, we’d get 36 mg / 60 kg = 0.6 mg/kg.
However, we have think about this as a hormone. Because they’re signaling molecules that regulate our bodies, the concentrations of hormones are low, like really low, like usually in the microgram to nanogram per kg range. By example, the doses in estrogen replacement therapy are in or below the single-digit mg range. In other words, those 36 mg is not something we can immediately dismiss.
We shouldn’t forget that different compounds have different activities, and true estrogen derivatives usually have far stronger effects than estrogen mimics. That said, when frequently used UVA and UVB screens were tested for their ability to make breast cancer cells proliferate (grow), the median effective concentrations needed to cause maximal growth fell between 1.56 and 3.73 microM. That absorbed 36 mg in the “average” 60 kg person will come out to be 2.63 uM. Hm, could certainly be influencing biology (especially when combined with other estrogen mimics around, like plasticizers)… The same paper found that dermal application of 2.5%, 5% and 7.5% sunscreen caused uterine growth in female mice. [Note they did go a bit “overboard” in that they dunked mice in oil with sunscreen and then painted more them more after. Remember though, the researchers were looking for an effect in 6 days—not say, 6 years of development of our kids—and effects are often cumulative. Anyone going to volunteer to have their daughters’ uteri measured over time as a function of sunscreen use? Fat chance.]
Note in all of this, I’m not talking about toxicity: the levels are insufficient to cause acute toxicity. Rather, my concern is the very real chance for these chemicals to dysregulate normal biology. (Note as well that we haven’t even considered chap-stick, yet another rout of administration (you eat it). I view sunscreen chap-stick as genuine poison and immediately chuck it in the trash.) My take on this is above: we can avoid all these possibilities, so why ever take a chance with our families???
4) The effects of chemicals can be vastly different at different stages of development (ages) and we’re not monitoring effects on development. In addition to the whole “I’m not even sure we’re measuring what’s changing” of #2 above, I’m absolutely certain we don’t have the data of “concentration in maternal blood or milk over time as a function of embryo/baby age” for larges population. The effects of chemicals can be extremely different during different stages of development and, as with the rest of science, it’s a case of “if you don’t look, you won’t see it.” We often learn this the hard way. Take thalidomide: for adults, it’s actually a pretty good drug (low toxicity, doesn’t cause cancer, works well for certain things) but if a pregnant women took it three to eight weeks after her last menstruation, her child often ended up with birth defects like missing limbs and heart defects. Outside that developmental window, things often ended up just fine.
Unsurprisingly, scientists do research in systems they can use, namely cells and animals, not on, say, the brains of developing kids. Contrast that understandable approach to how we’re exposing our families: depending on your activities and where you live, you might use these chemicals seasonally, semi-regularly or regularly from the time your kids are what, say maybe 2 until they leave the house? And nursing moms pass on significant amounts to their babies through breast milk. This is a fully uncontrolled experiment! Until we know exactly what during development might be affected (which yields “when”), we’re simply going to miss those effects. Even if we do know which system is altered, those changes might be really subtle, thus requiring massive studies to observe and making study yet more difficult.
I have an idea to (sort of) test the activity of organic sunscreens on “brain tissue” using “neural tissue” derived from stem cells (i.e. they’re sort of mostly neurons growing in a petri dish, which is what we can do; conveniently my wife is a Stanford PhD in neurodevelopmental biology who studies brains J) but those are expensive, long experiments, I’m not sure who’d want to fund them and…we have good alternatives; why even spend the time and money? In sum, we know these chemicals have biological activity in various systems, we know that the biological activity of compounds can vary depending on development (sometimes becoming really important like “remove baby limbs”) and we know that historically we’ve missed those effects until it’s too late. I’m just not going to take that chance.
A parting, perfect example of “oops” as a function of the age of an organism: you know how all our bees are dying off? Guess what pesticides (etc.) have been tested on? Yup, adult bees. And guess what? Bees have babies too and just like for humans, the effect of chemicals vary depending on age. Those pesticides are far more toxic to the bee-baby larvae and kill them. Apparently we’re pretty good at spotting effects of chemicals too late in both bee- and human babies.
5) The toxicity of these chemicals isn’t limited to people. Apparently they’re pretty good at killing coral reefs too. Great. Wonder what they do to bees…
6) Finally, two rules of thumb: 1) If you see octinoxate (aka octylmethoxycinnamate) or especially if you see oxybenzone, just don’t buy that product (actually, if you have them, just throw it out; I disposed of ours with household chemical waste). These two compounds are really bad with oxybenzone absolutely taking the cake. EWG’s description of it: “Detected in nearly every American; found in mother’s milk; 1-to-9% skin penetration in lab studies. Acts like estrogen in the body; alters sperm production in animals; associated with endometriosis in women. Relatively high rates of skin allergy.” Hopefully that does (and should) freak you out. 2) If you can’t find a mineral sunscreen and must use a sunscreen, find one that’s only got avobenzone in it (it’s got the lowest toxicity so far). Unfortunately, there too I’m just waiting for some new of “Oh, we found this new effect.”
Why the minerals in sunscreen don’t worry me at all
In contrast to the whole list of nasty worries above, titanium and zinc oxides (TiO2 and ZnO respectively) are about as innocuous as you can find. TiO2 is used as a control in a bunch of science fields where they use particles and the studies I’ve seen have shown it has the same activity as adding nothing. Titanium and zinc oxides don’t have any hormone mimicking effect, don’t go through the skin at all (or some vanishingly small concentration of zinc oxide does, which also doesn’t worry me) and they won’t screw up the development of your kids or their beautiful brains. They don’t cause cancer or promote its growth. They’re in diaper cream (that’s zinc oxide), toothpaste and about everything else that’s white (candy, sprinkles, frosting…). In other words, you eat titanium dioxide probably frequently. There is a genuine concern if you’re dealing with their powders (dusts are always bad) so don’t use powders; I’ve never had modern sunscreen turn dusty. These statements are true as well with nano formulations, but because the particles are embedded in a cream, I’m still not worried about those. (I prefer non-nano to nano if I’ve the choice however.) If you have damaged skin, as a general rule I wouldn’t apply sunscreen and that applies here too.
If you dig deep enough, you’ll find concerns about mineral sunscreens catalyzing free radicals in the presence of sunlight. I view that as a non-starter because 1) it’s unlikely to happen (for chemical reasons) and 2) it’s on the very exterior of your skin. I’m not worried about the outer layers of our skin (which gets oxidized etc. every day from our atmosphere), I’m worried about, say, brain cells and cancer. Long story short, these compounds are about as safe as chemicals come.
