<emph type="ital">ABCA1</emph> Gene Mutations, HDL Cholesterol Levels, and Risk of Ischemic Heart Disease

Abstract

would not expect a change in our results. Regarding the use of self-reported data on family history, prior studies have found self-reported information on family history of cancer in first-degree relatives to be valid. Underreporting tends to be the most common source of error in family history reports. This misclassification would tend to attenuate any association of family history of colorectal cancer on patient survival. Moreover, the prevalence of family history of colorectal cancer among study participants was consistent with other studies. We agree that people who are aware of a family history of colorectal cancer might adopt a more health-conscious lifestyle and may be more likely to be diagnosed at earlier stages. However, we did not observe any significant differences in T stage (invasion through the bowel wall), N stage (number of positive lymph nodes), tumor differentiation, performance status, or postoperative serum carcinoembryonic antigen between patients with or without a family history. Furthermore, a significant association between a family history and disease-free survival persisted even after adjusting for predictors of cancer recurrence or death as well as in stratified analyses of patients according to T and N stage. Regarding preferential responsiveness to chemotherapy, all patients were randomized to standardized chemotherapy protocols, and there were no differences in survival according to treatment group. Because we did not assess family size or sex of the affected family member, we are unable to examine family history according to sex of the affected relative or personyears of exposure. However, we are unaware of data demonstrating that risk of developing colorectal cancer is influenced by the sex of an affected first-degree relative and have no reason to suspect that sex of the relative would influence survival. We did examine whether the effect of family history varied according to the age at which the firstdegree relative was diagnosed with colorectal cancer. As reported, we found no difference in the adjusted hazard ratio for disease-free survival among patients whose relative was diagnosed with colorectal cancer at an age younger than 50 years compared with 50 years or older.