Corder Elizabeth H, Hefler Lukas A
Center for Demographic Studies, Duke University, Durham, North Carolina 27708-0408, USA.
Rejuvenation Res. 2006 Spring;9(1):56-60. doi: 10.1089/rej.2006.9.56.
We investigated inherited polymorphic variation in genes spanning estrogen metabolism (10 SNPs [single nucleotide polymorphism]) to distinguish multilocus genotypes associated with breast cancer (n = 393), benign breast lesions (n = 154), and low risk (n = 1936). Three latent classification GoM extreme type groups represented the data: (a) fibroadenoma, and infrequent SRD5A2 and VDR alleles; (b) postmenopausal breast cancer, and infrequent CYP1A1-1 and CYP1A1-2 alleles (both over-represented infrequent alleles for CYP17, CYP19-3, and COMT); and (c) women at intrinsically low risk. Carriage of the respective multilocus genotypes increased risk 25-fold. We conclude that GoM latent classification may be useful to identify genetic risk sets and estimate risk for individuals.