Sunscreen is one of my favorite health topics. The advice to wear sunscreen (nearly) every day is cheap and easy to follow and the science behind it is quite robust. In this post, I’ll summarize the randomized controlled trials on long-term daily use of sunscreen and then discuss a few practical considerations for choosing which sunscreen to wear.
Much of what we know for sure about the long-term effects of wearing sunscreen every day comes from the lab of Adele Green. Of all the different topics I’ve looked into for my How-to-115 bibliography, there is no topic so thoroughly researched by a single lab as sunscreen is by the Green lab. This post is more or less my quick summary of her painstaking research.
As background, there are two basic kinds of skin cancer: melanoma and (basal cell and squamous cell) carcinoma a.k.a. non-melanoma skin cancer. Non-melanoma skin cancer is much more common and much less fatal (3.3 million people per year diagnosed, 2k deaths in the US) than melanoma (91k cases, 9k deaths in the US in 2018). Because the less-terrible kind happens much more often, any experiment that can eventually tell you if sunscreen prevents the terrible kind will first tell you if it prevents the less-terrible kind.
I. The Nambour Skin Cancer Prevention Trial
i. Skin Cancer
The RCT (free abstract, then paywall) started in 1992 with ~1,600 people in Nambour, Australia randomly assigned to wear sunscreen or not. People in the sunscreen group were told to rub water-resistant SPF 15 on their head, neck, arm and hands every morning. People in the non-sunscreen group were told to wear sunscreen “at their usual discretionary rate.” To ensure that people followed the instructions, the experimenters gave them the sunscreen, then took the bottles back every 3 months and weighed them to see how much sunscreen people had actually used.
After 4.5 years, the only statistically significant result was a 39% drop in the number of squamous cell carcinoma tumors for people in the sunscreen group. The effect was too small to be statistically significant when counting the number of people instead of the number of tumors. There was also no effect on basal cell carcinoma.
The experimenters continued to follow their subjects for 8 more years, although they were no longer pestered to wear sunscreen daily or given free sunscreen. In the follow-up period there was again a statistically significant 40% reduction in the number of SCC tumors. The combined data from the 12.5-year period finally reached statistical significance for a 35% reduction in the number of people diagnosed with SCC. There continued to be no effect on BCC.
- (I love the brief acknowledgments section of this paper, which makes no reference to funding and only thanks, “the residents of Nambour who have contributed to this
study for so many years.”)
Melanoma is so rare, the experiments didn’t bother to comment on it for the first 12 years of the student. After 10 years of follow up, bringing the total observation time to 14.5 years, they reported a 50% reduction in melanoma for the daily sunscreen group versus the control group, with a p-value of 0.051, right at the (totally arbitrary) boundary of statistical significance. A little under half of the melanomas were classified as “invasive,” with a ~70% reduction for the daily sunscreen group.
As a 20-something, I am fundamentally incapable of considering my own death. To that end, my favorite part of this long saga is the part about not getting wrinkly. Dr. Green and her lab divide skin aging into two very technical sources: “photoaging and growing old.” For this part of the trial they just looked at people in their trial below 55, because those people have apparently not yet been diagnosed with “growing old.” People described as “experienced assessors” (dermatologists? cosmetologists?) who were blind to sunscreen group assignment rated each participant’s skin on a scale of 1 (baby’s bottom) to 6 (worn out leather jacket). Over 4.5 years, people in the daily sunscreen group were 24% less likely to advance a step on this methuselah scale.
iii. Cost Effectiveness
The Green lab went on to measure the cost-effectiveness daily sunscreen use. Based on their previous skin-cancer RCT (and excluding the benefits of not getting wrinkly) they conclude you can buy 1 quality-adjusted life-year (a standard measure in public health research) for AU$40,890 ~ US$30,310 (they stop at $10 precision because anything more would just be silly).
iv. Anti-oxident supplements
This trial also randomized people to taking anti-oxidant supplements (30 mg betacarotene daily) and found no effect on any of the measures discussed above.
II. Practical Considerations
i. Any Sunscreen is Safe for Humans
There are two basic kinds of sunscreen: chemical and physical. Both are perfectly effective. Some groups, in particular the Environmental Working Group, argue that chemical sunscreens are unhealthy for humans and bad for the environment. If they were otherwise identical you may as well default to physical sunscreen. The problems are 1) that physical sunscreen is harder to rub in, so you might fight it harder to get into the habit of using it every day and 2) that physical sunscreens often don’t match their claimed SPF ratings.
In general, I do not consider the EWG to be a reliable source. Although they are a non-profit, they seem to get their funding (and salaries) by scaring the crap out of people without reason and telling them to buy needlessly expensive products. Note the giant trademark logos all over their pseudo-scientific pages. They are literally selling you something.
More importantly, they are bad at science. For every chemical in a consumer product there exists some study that says bad things happened to a mouse or rat who ate it. Whenever you see a scary headline like that, beware! The purpose of these studies is not to find out if the chemical is toxic. All chemicals, including water, are toxic at sufficiently high dose. The question is how high that dose is compared to what you would be exposed to by some product.
As far as chemical sunscreen is concerned, the EWG’s objection seems to revolve around a study that forced rats to eat copious amount of oxybenzone. This comment from a bunch of neutral dermatologists says it would take hundreds of years of applying oxybenzone-based sunscreen to get the same dose the mice got in four days. They point to studies of actual humans who have used chemical sunscreens for quite a long time without any of the problems associated with the rats who gorged themselves on it. In short, do not eat oxybenzone-based sunscreen for breakfast, lunch and dinner. Feel free to eat all the zinc-based sunscreen you want, though.
The EWG’s arguments are bad enough that I believe they are not acting in good faith. Their sunscreen scaremongering is their most famous, but when I see the EWG and poor well-meaning souls who have fallen prey to its marketing warning against any and all chemicals I immediately seek out other more careful reporting, and rarely find good science behind the claims (here, here, here, here).
ii. Oxybenzone Not Safe for Coral Reefs
Although oxybenzone is perfectly safe for humans, it’s not safe for coral reefs. If you’re swimming near them or even if you’re on land near them, maybe stick to physical sunscreens. But if you’re well clear of coral reefs, just use whatever sunscreen you like best.
iii. Sunscreen Recommendations
Absurdly thorough product reviews from The Wirecutter are, to me, the consumerist equivalent of RCT’s. Here are their sunscreen recommendations.