Spatial Associations and Potential Explanations between UV Exposure and Melanoma Incidence in Populations

Melanoma is one of the most lethal forms of skin cancer, and its incidence is strongly linked to ultraviolet (UV) radiation exposure. Although prior studies have examined this relationship, most have relied on linear models that overlook spatial heterogeneity and nonlinear effects. To address these gaps, we analyzed 386,701 primary cutaneous melanoma cases from 1,038 U.S. counties, covering up to 48 % of the U.S. population, by using a combination of Cox regression and geographically weighted regression (GWR). Our results revealed that a one-unit increase in population-weighted UV radiation (PUR) was associated with a 53.1 % increase in melanoma risk (P<0.01), while spline analyses uncovered a nonlinear relationship between PUR and melanoma incidence across geospatial regions. By incorporating local covariates such as socioeconomic status, healthcare access, and racial composition into the GWR model, we found that PUR remained a significant predictor of melanoma incidence (regression coefficient =–5.87, p<0.001), and that local variations in healthcare resources and economic conditions modulate this association. These findings offer valuable insights for designing spatially targeted melanoma prevention strategies.