Abstract
BACKGROUND: Cutaneous basal cell carcinoma (BCC) has long been associated with UV radiation (UVR) exposure, but data are limited on risks by anatomic site. METHODS: We followed 63,912 cancer-free White U.S. radiologic technologists from cohort entry (1983-1989/1994-1998) to exit (date first BCC via 2003-2005 questionnaire). We estimated associations between cumulative ambient UVR and relative/absolute risks of self-reported BCC by anatomic location via Poisson models. RESULTS: For incident first primary BCC in 2,124 subjects (mean follow-up, 16.9 years) log[excess relative risks] (ERR) of BCC per unit cumulative ambient UVR = 1.27/MJ cm(-2) [95% confidence interval (CI): 0.86-1.68; P (trend) < 0.001] did not vary by anatomic site (P = 0.153). However, excess absolute risks of BCC per unit cumulative ambient UVR were large for the head/neck = 5.46/MJ cm(-2)/10(4) person-year (95% CI: 2.92-7.36; P (trend) < 0.001), smaller for the trunk (2.56; 95% CI: 1.26-3.33; P (trend) = 0.003), with lesser increases elsewhere. There were lower relative risks, but higher absolute risks, for those with Gaelic ancestry (P < 0.001), also higher absolute risks among those with fair complexion, but relative and absolute risks were not generally modified by other constitutional, lifestyle or medical factors for any anatomic sites. Excess absolute and relative risk was concentrated 5-15 years before time of follow-up. CONCLUSIONS: BCC relative and absolute risk rose with increasing cumulative ambient UVR exposure, with absolute risk highest for the head/neck, to a lesser extent in the trunk. IMPACT: These associations should be evaluated in other White and other racial/ethnic populations along with assessment of possible modification by time outdoors, protective, and behavioral factors.