5.10 THIS IS HOW WE DO IT: Does sunscreen use reduce skin cancer risk?

Wear sunscreen.

It’s a simple piece of advice, and it’s endorsed by the American Medical Association. Exposure to the sun is implicated in many skin cancers. And the dramatic effect of sunscreen in making it possible to spend time in the sun without getting burned is obvious the first time you use it.

So it might come as a surprise that, according to many studies, sunscreen does not protect people from skin cancer. And even more surprising is that many studies have shown that sunscreen use is even associated with an increased risk of the most dangerous of all skin cancers, melanoma.

Let’s look at what happens when you’re exposed to ultraviolet radiation from the sun.

When skin cells are exposed to a type of ultraviolet radiation called UVB radiation, the cells and their contents—including RNA and DNA molecules—can be damaged. This signals nearby cells to produce chemicals that cause inflammation—which you know as sunburn—as part of the process by which the damaged cells are cleared away.

The relationship between sun exposure and skin cancer has been noted and studied for more than 70 years:

• People with lighter-colored skin are much more sensitive to the sun than those with darker skin, burn more easily, and have more than twice the rate of melanoma When melanoma occurs in people with darker skin, it primarily affects areas with lighter skin color, such as the palms and soles.
• There is a strong relationship between the amount of time a person spends in the sun and the risk of melanoma.

Sunscreens contain chemicals that absorb or reflect some of the UV radiation. In doing so, they protect against many of the harmful effects of the sun, including sunburn. Nearly everyone assumed that the same wavelengths of sunlight that caused sunburn were also responsible for causing skin cancers. And so it followed that sunscreen would provide protection from skin cancer as well as sunburn.

Sunscreens first became widely available in the 1960s and 1970s. Since then, there has been an increase in the incidence of melanoma. Strangely, in countries where sunscreen is most recommended and most often used—Canada, the United States, Australia, and Scandinavia—the increase in melanoma has been greatest. Deaths from melanoma in the United States more than doubled between 1950 and 1990.

In a study in Sweden, researchers compared sunscreen use between two groups: 571 patients diagnosed with melanoma and 913 healthy controls. Contrary to the researchers’ expectations, sunscreen users had a significantly higher risk for developing melanoma than non-users—almost twice as high! (This was just one of eight studies published between 1979 and 1995 that found a positive association between sunscreen use and melanoma.)

People using sunscreen may have a false sense of security. This may lead them to spend more time in the sun and to be less likely to wear hats or other clothing that offers protection. Further complicating matters, until the mid-1990s, sunscreens with full UVA and UVB protection (called “broad spectrum” sunscreens) were not available. And even in 2010, only one-third of sunscreens available in the United States offered protection from all the wavelengths of UV radiation in sunlight. For these reasons, because the sunscreens reduced sunburn while allowing more time in the sun, they may have caused people to unknowingly overexpose themselves to those wavelengths for which the sunscreen gave no protection.

Before throwing away your sunscreen, though, it’s important to be sure that it really is of no benefit. This is an important element of scientific thinking and the guiding principle taken by researchers in a more recently published study about melanoma and sunscreen.

All of the studies purporting to show a positive relationship between sunscreen use and melanoma risk were “case-controlled” studies. This is a study design in which individuals with a particular outcome (such as developing melanoma) are identified and then compared with a group of individuals who do not have that outcome (do not have melanoma). The groups are analyzed to see whether they differ in some significant way—such as sunscreen use. Such a study design has some limitations. It can identify factors influencing the outcome of interest, but the validity of any associations depends on how similar the two groups actually are. In this case, for example, an assumption was made that the healthy subjects and the subjects with melanoma experienced similar sun exposure.

The results from such studies are undermined when the groups are heterogeneous—for example, if those in the melanoma group had more sun exposure than those in the healthy group. In fact, subsequent investigations suggested that the case-controlled studies were hindered by “confounding variables” that caused comparison groups to differ in significant ways. (Can you think of ways the two groups might differ?)

An alternative—and usually more powerful—study design is a randomized controlled study. In 2011, researchers reported on just such a study in Queensland, Australia, which randomly assigned 1,621 adults to one of two groups: regular sunscreen use and discretionary sunscreen use. (Would it have been a better study if subjects in the second group had not been allowed to use sunscreen at all? Why do you think the researchers didn’t do that?)

For 5 years, those in the sunscreen group received unlimited sunscreen and were asked to apply it every morning (and to reapply it after sweating, bathing, or long sun exposure). The discretionary sunscreen users were allowed to use sunscreen at their usual frequency (which included no use at all for some people). At the end of the 5-year treatment period, and continuing for 10 additional years (until 2006), the researchers noted the incidence of melanomas. (This was made possible because in Queensland, all melanomas must be reported to the Queensland Cancer Registry.) They also obtained, from questionnaires filled out by study participants, information about time spent outdoors and sunscreen usage.

The results were dramatic. Among the 812 subjects in the sunscreen-use group, during the 10-year follow-up period, 11 new melanomas were identified. Among the 809 subjects in the discretionary-use group, twice as many new melanomas were identified. Invasive melanomas, the most severe type—occurred almost four times more frequently among people in the discretionary group. Based on the questionnaires, there were no differences between the groups for any known risk factors, or for the amount of time they had spent in the sun during the trial.

Because this is the first randomized controlled study of sunscreen use and melanoma, cancer experts have described it as a potential “game changer.” They view the results as a clear indication that melanoma-prevention strategies should incorporate efforts at increasing regular use of sunscreen.

These studies reveal just how important well-controlled experimental design is to demonstrating causal relationships. The case-controlled studies were important, but their inherent limitations led to results that may not hold up under closer scrutiny. Randomized controlled trials, though, are difficult to conduct and relatively expensive—and it can be difficult or even impossible to ensure full compliance with treatments (or with non-treatment by individuals in control groups). But we have much greater confidence in the evidence from such studies.