Determinants of polyp size in patients undergoing screening colonoscopy

Pre-existing polyps, especially large polyps, are known to be the major source for colorectal cancer, but there is limited available information about factors that are associated with polyp size and polyp growth. We aim to determine factors associated with polyp size in different age groups.

Colonoscopy data were prospectively collected from 67 adult gastrointestinal practice sites in the United States between 2002 and 2007 using a computer-generated endoscopic report form. Data were transmitted to and stored in a central data repository, where all asymptomatic white (n = 78352) and black (n = 4289) patients who had a polyp finding on screening colonoscopy were identified. Univariate and multivariate analysis of age, gender, performance site, race, polyp location, number of polyps, and family history as risk factors associated with the size of the largest polyp detected at colonoscopy.

In both genders, size of the largest polyp increased progressively with age in all age groups (P < .0001). In subjects ≥ 80 years the relative risk was 1.55 (95% CI, 1.35-1.79) compared to subjects in the youngest age group. With the exception of family history, all study variables were significantly associated with polyp size (P < .0001), with multiple polyps (≥ 2 versus 1) having the strongest risk: 3.41 (95% CI, 3.29-3.54).

In both genders there is a significant increase in polyp size detected during screening colonoscopy with increasing age. Important additional risk factors associated with increasing polyp size are gender, race, polyp location, and number of polyps, with polyp multiplicity being the strongest risk factor. Previous family history of bowel cancer was not a risk factor.