A segregation analysis of colorectal cancer
Description
Segregation analysis was used to model the occurrence of colorectal cancer in 70 families ascertained through colorectal cancer probands. Families with hereditary colorectal cancer syndromes, including familial adenomatous polyposis (FAP) and hereditary non-polyposis colorectal cancer (HNPCC), were excluded from the data. A transmission probability model with variable age of onset (S.A.G.E. REGTL Model 1) with all regressive effects fixed equal to zero was used to test whether a genetic or purely environmental model best fitted the data. The hypotheses tested were: a one level of effect, purely environmental model in which colorectal cancer has one distribution of age of onset for the entire population, two and three level of effect, purely environmental models in which transmission is due to an unmeasured environmental factor that is not correlated within families; and two and three level of effect, genetic models in which transmission is due to a single locus, two allele, mendelian gene and additional environmental risk is assumed to be randomly distributed The genetic and purely environmental hypotheses were tested against two unrestricted models: the heterozygote unrestricted model and arbitrary transmission probability model. The environmental hypotheses with one, two or three levels of effect were strongly rejected against both models. The genetic hypotheses were not rejected when compared to the heterozygote unrestricted model. They were rejected at the p $<$.05 level but not at the p $<$.01 level when compared to the arbitrary transmission model. There were no significant differences between the likelihoods of the dominant, recessive and codominant Mendelian models