Purpose: The action spectrum of ultraviolet radiation mainly responsible for melanoma induction is unknown, but evidence suggests it could be ultraviolet A (UVA), which has a different geographic distribution than ultraviolet B (UVB). This study assessed whether melanoma mortality rates are more closely related to the global distribution of UVA or UVB.
Methods: UVA and UVB radiation and age-adjusted melanoma mortality rates were obtained for all 45 countries reporting cancer data to the World Health Organization. Stratospheric ozone data were obtained from NASA satellites. Average population skin pigmentation was obtained from skin reflectometry measurements.
Results: Paradoxically, melanoma mortality rates decreased with increasing UVB in men (r = -0.48, p < 0.001), and women (r = -0.57, p < 0.001), and with increasing UVA in both sexes. By contrast, rates were positively associated with increasing UVA/UVB ratio in men (r = + 0.49, p < 0.001) and women (r = + 0.55, p < 0.001). After multiple adjustment that included controlling for skin pigmentation, only UVA was associated with melanoma mortality rates in men (p < 0.02) with a suggestive but non-significant trend present in women (p = 0.12).
Conclusions: UVA radiation was associated with melanoma mortality rates after controlling for UVB and average pigmentation. The results require confirmation in observational studies.