More specifically, the researchers found that the correlations between UV-related variables and melanoma were inconsistent and weak while the correlations between diagnostic scrutiny variables and melanoma were consistently positive and stronger.
“As physicians, despite our best intentions, we may be the source of the epidemic of melanoma diagnosis,” first author Adewole S. Adamson, MD, an assistant professor with the Department of Internal Medicine at Dell Medical School, The University of Texas at Austin, told Medscape Medical News. “The current pattern of melanoma incidence in the US is less associated with UV radiation exposure, and more associated with medical practice.”
Cutaneous melanoma, once a rare cancer, has become the third most diagnosed cancer in the US in recent years, with an incidence rate now six times higher than it was 40 years ago, the authors noted.
Adamson and colleagues questioned whether this increase might be due to greater UV radiation exposure or overdiagnosis — the result of increased diagnostic scrutiny.
Consistent with overdiagnosis patterns observed with other cancers, there has been no corresponding increase in mortality from melanoma.
To investigate, Adamson and his team conducted a cross-sectional ecological study, published this month in JAMA Internal Medicine, exploring Surveillance, Epidemiology, and End Results (SEER) data from 727 US counties between January 2020 and July 2022.
The entire dataset included 10 variables related to UV exposure and eight variables related to diagnostic scrutiny, but the team presented three for each category in the main analysis. The three proxies for UV radiation were UV daily dose, variability in cloudiness, and variability in temperature across the counties, and the three for diagnostic scrutiny were median household income as well as concentrations of dermatologists and primary care physicians.
With a total of 235,333 melanomas diagnosed during the study period, the results showed no correlation between the average UV daily dose and the incidence of melanoma (P = .42).
However, melanoma incidence did significantly correlate with the three proxies of diagnostic scrutiny. The incidence of melanoma did significantly correlate with median household income (P < .001). In addition, counties with the highest concentrations of dermatologists and primary care physicians had the highest incidence rates of melanoma, despite having lower mean daily UV exposure, while counties with shortages of dermatologists and primary care physicians had the lowest incidence of melanoma.
In fact, as Adamson explained in a tweet, “because of reviewer request we even took it a step further and added another analysis in the supplement to show that counties that gained dermatologists had HIGHER increases in melanoma incidence over time.”
The observed trends can be seen in two communities. Orange County, California, and Puget Sound, Washington, had nearly identical melanoma incidence rates — 73 and 72 per 100,000, respectively; however, the two communities sit on opposing extremes of UV radiation exposure with Orange County at the top decile and Puget Sound at the bottom.
The authors also pointed to southeastern New Mexico, which has among the highest UV daily doses in the country, but among the lowest melanoma incidence.
Another factor pointing to overdiagnosis: The authors found only a small association between melanoma incidence and melanoma mortality.
For context, the authors noted that the prevalence of smoking was highly correlated with lung cancer incidence and mortality in the same counties.
While previous studies have shown conflicting results in terms of UV exposure and melanoma incidence, “this is the first study that examined proxies for UV exposure and proxies for diagnostic scrutiny side by side,” Adamson said.
Among factors that could be viewed as confounders, Adamson and colleagues pointed to the possibility that those living in higher-income counties might have greater UV exposure from traveling to warmer weather destinations, particularly during the winter; however, that theory would not explain similar patterns across entire county populations, the authors noted.
In addition, higher income counties also have higher incidence rates of breast, prostate, and thyroid cancer, suggesting that “unless one argues that these cancers are also caused by vacation travel, the more parsimonious explanation is that all four cancers are sensitive to diagnostic scrutiny,” the authors said.
Given the evidence, some of the more exaggerated public messaging on sunlight exposure should be toned down. For instance, the idea that 95% of melanomas are caused by the sun is “particularly troubling,” they wrote.
This suggestion would mean that the “sun is a carcinogen more tied to melanoma than smoking is to lung cancer,” Adamson noted. “That defies common sense.”
“Advising people to be sun safe is important, particularly for children, adolescents, and young adults,” he added. “However, phobia of the sun should not get in the way of people engaging in the many beneficial aspects of leading an active, social lifestyle outside in the sun.”
Douglas Grossman, MD, PhD, who was not involved in the study, noted that “the proxies used for UV exposure and diagnostic scrutiny are indirect, and weaker than more direct indicators not evaluated in the analysis — such as individual sun exposure data for UV exposure and screening visits and biopsies performed for diagnostic scrutiny.”
Due to the indirect proxies, “we need to be careful in drawing conclusions of causality from the study,” said Grossman, a professor of dermatology and co-leader of the HCI Melanoma Center at the University of Utah, in Salt Lake City.
However, even if causality is confirmed, trends behind overdiagnosis may be hard to change, Grossman added. “As long as there are financial incentives to perform biopsies, and continued risk of medical liability for missing a melanoma, it is likely that screening practices will continue,” he explained.
Adamson is an associate and web editor at JAMA Dermatology and reported no other disclosures. Grossman reported no disclosures.
JAMA Internal Medicine. Published Oct. 3, 2022. Abstract.