Epidemiology of Polyps in the Rectum and Sigmoid Colon: Discriminant Analysis for Identification of Individuals at Risk of Developing Colorectal Neoplasia

Abstract

Hoff G, Larsen S. Epidemiology of polyps in the rectum and sigmoid colon. Discriminant analysis for identification of individuals at risk of developing colorectal neoplasia. Scand J Gastroenterol 1986, 21, 848-852

In an endoscopic population screening examination for detection of colorectal polyps among 324 men and women aged 50-59 years, adenomas ≥5 mm in diameter were found in 38 individuals, and 212 were polyp-free. Dietary registration was performed by 155 individuals on a case/control basis. At 2 years’ follow-up study polyps less than 5 mm in diameter were also removed. This rendered a total of 39 adenoma patients and 59 polyp-free individuals available for a combined analysis of both case history and dietary information. Application of a linear discriminant model on the combined clinical and dietary information enabled us to select a possible set of variables to characterize 88 out of 98 individuals (90%) correctly as adenoma patients or polyp-free individuals. However, when only case history information was used for the total material of 38 adenoma patients and 212 polyp-free individuals, only 189 out of 250 individuals (76%) were correctly classified. Although age, smoking habits, and a family history of colorectal cancer appeared as useful variables in the discriminant function, this study demonstrated the importance of nutritional factors in characterization of adenoma patients in the present age group of 50-59 years. In the present study we could not find any set of case history information which could indicate the usefulness of a questionnaire-based ‘pre-screening’ to guide recommendations for endoscopic screening examination.

Key Words:

• Adenomas
• case history
• colon/rectum
• diet
• discriminant analysis
• endoscopy
• population screening