High UV Exposure Tied to Elevated Melanoma Risk in Large NIH-AARP Study: Samer Wahood, AB

In this interview, Samer Wahood, a fourth-year MD candidate at the Warren Alpert Medical School of Brown University, shares the background, key findings, and future directions of his abstract, "Association of UV Radiation Exposure With Risk of Malignant Melanoma and Melanoma in Situ in the NIH-AARP Diet and Health Study," presented last week at the American Association for Cancer Research Annual Meeting 2025 in Chicago.

This transcript has been lightly edited; captions were auto-generated.

Transcript

What inspired you to investigate the association between UV radiation and both malignant melanoma and melanoma in situ? What gap in the existing literature did you hope to address?

This really started with research with my mentor, Dr Eunyoung Cho. She explored a link between melanoma and fish intake. From there, that research kind of showed that there might be something interesting going on with UV exposure, particularly this type of UV exposure called July erythemal UV radiation, and invasive melanoma and melanoma in situ.

To just recap a little bit, July erythemal UV exposure is looking at the amount of UV radiation in the month of July based on some data from NASA. It's really based on the residential level, so it's joules per meter squared in particular areas.

The particular study that we're using is the NIH-AARP [Diet and Health] study, which is one of the largest prospective studies in the United States. Our cohort, after pruning it to the variables that we're looking at, was a little over 490,000 participants between [the] ages [of] 50 to 71. This was an opportunity to look at some really robust data and see the association between melanoma in situ and invasive melanoma.

Particularly, the gaps here that we're looking at are looking at this association from a prospective study. There are a number of prospective studies out there, like the Nurses’ Health Study back in the '70s and '80s, but this is a particularly robust study given that its population is over 490,000.

The other part is [that] we're able to get down to not just melanoma in general, total melanoma, but break it down by melanoma in situ and invasive melanoma and see [what] that risk difference looks like.

Can you summarize the methods you used to explore these associations and share the key findings? Did any of the results surprise you?

We were looking at the July erythemal UV exposure, and we were relating it to melanoma, invasive melanoma, and melanoma in situ. Based off of different quartiles here, so anywhere from less than 180 joules per meter squared to over 236 joules per meter squared, and whatever is in between that, how much does someone's risk for malignant melanoma or melanoma in situ differ?

We noticed that, in the first couple of quartiles, under 236 joules per meter squared, [the hazard ratio] was pretty much around 1:1. So, someone who isn't around a lot of sun exposure and someone who is has about the same risk for either type of melanoma.

Once you get above 236 [joules per meter squared], that's when someone's risk really jumps. The increase in risk appears to be about 32% for invasive melanoma and 37% for melanoma in situ.

How does UV radiation exposure compare with other known risk factors for skin cancer, such as family history or lifestyle choices?

UV radiation is known as the most well-known modifiable risk factor for invasive melanoma and melanoma in situ, compared with family history or history of blistering sunburns.

Comparing this now to some other interesting findings that we've had, including someone's educational level. As someone's educational level increases, we're noticing that they have higher rates of UV radiation exposure. That brings into question their residence, their lifestyle factors, and how that could relate.

Eventually, we're going to do some stratified analyses to look at socioeconomic index and, also, poverty level data to see if there's any association there, as well.

How can health care providers use your findings to improve prevention, education, and early detection efforts for malignant melanoma and melanoma in situ, particularly in regions with higher UV radiation exposure?

While our study was an epidemiological study, so we were only looking at associations and not causation here, it's still fair to go ahead and make some recommendations to policymakers and health care providers. Things that we could think about are really localizing our efforts and making sure that we're targeting these areas that have high UV exposure.

The NIH-AARP study was looking at Florida, California, North Carolina, Detroit, and Atlanta. The areas that particularly have more sun exposure, I think, can benefit from localized prevention strategies.

Other things that we can consider are integration with the electronic medical record and really looking at patients who might be coming from some of these more sun-exposed areas and offering targeted interventions for them.

Based on your findings, are there any unanswered questions or areas for future research you're particularly interested in exploring next?

A few things that we're interested in are the different types of melanoma. Invasive melanoma and melanoma in situ, these are subtypes, but it'd be interesting to look at acral lentiginous melanoma, which is more common in darker skin tones.

Unfortunately, the dataset is pretty homogenous and does not reflect the diversity that we have in the US, so it's majority White, middle-aged, and non-Hispanic [patients]. Hopefully, with more prospective data in the future, and some other databases, maybe like the All of Us database, we're able to get some better data to get to these answers.

We're also interested in body site data, so where someone might get a melanoma on their body, whether it's on their nails, their hands, their feet, or other areas that maybe are not in sun-exposed areas. It'd be interesting to look at how melanoma rates differ by body site compared with how much UV exposure someone gets. Right now, the NIH-AARP study does not have this data, but hopefully, other prospective studies will have this.