Prediction of the risk of skin cancer caused by UVB radiation exposure using a method of meta-analysis

Shin, D C; Lee, J T; Yang, J Y

Abstract

Under experimental conditions, UVB radiation, a type of ultra violet radiation, has shown to relate with the occurrence of skin erythema (sun-burn) in human and skin cancer in experimental animal. Cumulative exposure to UVB is also believed to be at least partly responsible for the "aging" process of the skin in human. It has also been observed to have an effect of altering DNA (deoxyribonucleic acid). UVB radiation is both an initiator and a promotor of non-melanoma skin cancer. Meta-analysis is a new discipline that critically reviews and statistically combines the results of previous researches. A recent review of meta-analysis in the field of public health emphasized its growing importance. Using a meta-analysis in this study, we explored more reliable dose-response relationships between UVB radiation and skin cancer incidence. We estimated skin cancer incidence using measured UVB radiation dose at a local area of Seoul (Shin chon-dong). The studies showing the dose-response relationships between UVB radiation and non-melanoma skin cancer incidence were searched and selected for a meta-analysis. The data for 7 reported epidemiological studies of three counties (USA, England, Australia) were pooled to estimated the risk. We estimated rate of incidence change of skin cancer using pooled data by meta-analysis method, and exponential and power models. Using either model, the regression coefficients for UVB did not differ significantly by gender and age. In each analysis of variance, non-melanoma skin cancer incidence after removing the gender and age and UVB effects was significant (p>0.01). The coefficients for UVB dose were estimated 2.07x10-6 by the exponential model and 2.49 by the power model. At a local area of Seoul (Shin chon-dong), BAF value were estimated 1.90 and 2.51 by the exponential and power model, respectively. The estimated BAF value were increased statistical power than that of primary studies that using a meta-analysis method.