Comparisons of diet and biochemical characteristics of stool and urine between Chinese populations with low and high colorectal cancer rates.

In an investigation of the roles of diet and stool biochemistry in human colorectal carcinogenesis, 24-hour food, urine, and stool samples were collected from randomly selected participants from two populations with a fourfold difference in colorectal cancer risk: Chinese in Sha Giao, People's Republic of China (low risk), and Chinese-Americans of similar ages in San Francisco County, Calif, in the United States (high risk). The findings supported the hypotheses that colorectal cancer risk is increased by the consumption of high-fat, high-protein, and low-carbohydrate diets and is associated with high levels of cholesterol in stool as well as increased daily outputs of 3-methyl-histidine and malonaldehyde in urine. However, risk does not increase with low stool bulk and low total stool fibers.

Hormonal risk factors in testicular cancer. A case-control study.

The authors interviewed 273 northern California testicular cancer cases aged 40 and under diagnosed between 1976 and 1981, their mothers, and matched peer controls and their mothers on prenatal hormone exposure and other variables. Included was a population-based substudy (1979-1981) of all interviewable cases reported to the San Francisco Bay Area Surveillance, Epidemiology, and End Results registry. They found odds ratios (OR) of from 8.3 (sons' report) to 4.5 (mothers' report) associated with cryptorchidism, but found no association with mothers' hormone exposure or diethylstilbestrol exposure in pregnancy. They also found a significant association with lower age at puberty (OR = 2.0); a marginally significant association with mothers' breast cancer (OR = 2.9, p = 0.054); and a significant protective effect of reported mononucleosis (OR = 0.6). These associations remained strong in the population-based substudy. When cases were divided by histology, strong and specific associations of earlier puberty (OR = 2.3) and mothers' breast cancer (OR = 4.4) with nonseminomatous cancer, and of reported mononucleosis (OR = 0.3) with seminomatous cancer, were found. These observations suggest that 1) prenatal exogenous hormone exposure does not account for a significant fraction of testicular cancer, 2) a cluster of "breast-cancer-like" risk factors are associated with nonseminomas, and 3) there is some genetic risk of nonseminomas.