Recently, there has been compelling evidence from epidemiologic studies that higher caffeine intake is associated with lower risk of basal cell carcinoma (BCC) of the skin. However, the biological mechanisms by which caffeine protects against skin cancer are largely unknown. In the post-genomic era, it is now possible to screen the entire genome to uncover potential genetic interactions with caffeine. Examining gene-caffeine interactions could identify novel genes that act synergistically with caffeine and help uncover biological mechanisms underlying the observed inverse association between caffeine intake and BCC. Taking advantage of a large existing genome-wide association study (GWAS) of BCC, we propose to explore the effects of gene- caffeine interactions on BCC in the entire genome. In addition to the standard case-control approach, we will adopt the powerful new cocktail approach, which has demonstrated the greatest power under a wide range of interaction patterns and is particularly appealing in this study, given that the true interaction effects were not identified in BCC GWASs. Testing gene-caffeine interactions in BCC GWAS is also a promising approach to discover novel genetic susceptibility loci for BCC;genes influencing BCC through interactions with caffeine may be missed in traditional GWAS if their main effects are small. Of note, GWASs of caffeine/coffee consumption have newly identified a number of genetic loci associated with caffeine intake, offering strong candidates for BCC association via gene-caffeine interactions. We therefore proposed to test these loci for their associations with BCC and clarify the role of gene-caffeine interactions in their total genetic effects. First, we will evaluate their associations with BCC by testing their marginal effects, interactions with caffeine intake, and the joint effects of both marginal effects and gene-caffeine interactions. In our preliminary study, testing gene-caffeine interactions demonstrated greater power than the traditional test of SNPs'marginal effects for BCC associations. A genetic score of caffeine consumption will be calculated based on all coffee-related loci and used to represent the overall effect of these loci. Furthermore, we will apply a new approach of mediation analysis to decompose the total genetic effect of each coffee-related SNP into three components: the direct effect on BCC, the indirect effect through caffeine intake, and the mediated interactive effect with caffeine intake. This analysis will make clearer the role of gene-caffeine interactions Specifically, we plan to conduct a two-stage analysis using a large existing BCC GWAS of 2,277 BCC cases and 6,716 controls for discovery and an additional BCC case-control study of 1,000 BCC cases and 1,000 controls for validation. Our study would be the first well-positioned genome-wide gene-environment interaction (GWGEI) study on BCC. The statistical approaches tested and analytical experience gained in this proposal can be applied to GWGEI studies on other exposures and diseases as well.

Public Health Relevance

We will take advantage of a large existing genome-wide association study (GWAS) of BCC to discover novel genes that act synergistically with caffeine in BCC development and validate the findings in a replication set. This innovative work may uncover novel biological mechanisms underlying the observed protective effect of caffeine on BCC and suggest personalized BCC prevention strategies. With a focus on the genetic determinants of caffeine intake newly identified by GWAS, we will apply a new statistical approach that decomposes the total genetic effects into direct effects, indirect effects via caffeine intake and mediated gene-caffeine interactions, which makes clearer the role of their interactions with caffeine on BCC in mediation analyses.

National Institute of Health (NIH)
National Cancer Institute (NCI)
Small Research Grants (R03)
Project #
Application #
Study Section
Special Emphasis Panel (ZCA1-SRLB-Y (M3))
Program Officer
Mechanic, Leah E
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Indiana University-Purdue University at Indianapolis
Public Health & Prev Medicine
Schools of Public Health
United States
Zip Code