Prospective study of serum cholesterol and site-specific cancers.

From 1965 to 1968, 7716 Japanese-American men were examined and tested for serum cholesterol. After 22 years, 1380 incident cancer cases were identified. Of the site-specific cancers, only colon cancer cases had a significantly lower mean serum cholesterol value than that of noncases (213.0 mg/dl vs 219.0 mg/dl). When the study subjects were separated into either a low, middle or high group, based on their serum cholesterol values, there was a significant inverse trend for cases of oral/pharyngeal/esophageal cancer combined. The association was present for cases diagnosed within 10 years of examination (p = 0.012), but not for cases diagnosed after 10 years. This suggests that the inverse association is due to the metabolic effects of undiagnosed oral/pharyngeal/esophageal cancer upon serum cholesterol levels. These results are discussed in relation to other studies on serum cholesterol.