PURPOSE
To construct a composite score representing modifiable lifestyle and environmental risk (e-score) and test for associations with colonoscopy findings among US Veteran participants of CSP #380.
BACKGROUND
Understanding environmental and genetic risks beyond self-reported family history is a way to develop personalized colorectal cancer (CRC) screening. The Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO) study examined CRC risk stratified by sex and included an e-score along with genetic risk scores, where higher scores indicated higher risk. Both e-scores and genetic risk scores were significantly associated with CRC risk and models that included these were more robust than models that only included family history.
METHODS
CSP #380 is a prospective study of outcomes during colonoscopy screening (1994-97) and follow- up (1994-2009) for 3,121 asymptomatic Veterans aged 50-75. The dichotomous outcome of interest was most significant colonoscopy findings (MSCF) of i) advanced neoplasia (AN: ≥10mm adenomas or advanced histology, or invasive CRC) vs. ii) non-advanced adenomas (<10mm with tubular histology) or no neoplasia. The independent variable, e-score, was weighted according to the GECCO male sample and comprised BMI, height, diabetes, NSAID use, education, alcohol intake, smoking, exercise, and diet.
DATA ANALYSIS
Logistic regression was used to test associations between MSCF and e-scores, controlling for age, family history and number of colonoscopies.
RESULTS
Among 2,846 participants with complete data, 33.3% were aged 50-59 at baseline, 97% were male, and 83.8% were White. Those with AN (n=405, 14.2%) compared to those without AN (n=2,441, 85.8%) had higher median e-scores (29.5, range:0-99.8 vs. 29.0, range:5.2-100), suggesting a difference. The logistic regression models showed older participants (aOR: 1.04, 95% CI: 1.03-1.06) and those with more colonoscopies (aOR: 1.19, 95% CI: 1.06-1.33) had higher odds for AN. However, e-scores and family history were not significantly associated with MCSF.
IMPLICATIONS
E-scores were not significantly associated with MSCF in this preliminary study. Developing escores among a larger, diverse sample (N~381,695) of US veterans in the Million Veterans Program study will allow for stratified models in investigations of environmental and genetic risk for CRC. Outcomes from those analyses will support advances in screening guidelines with tailored programs for long-term CRC prevention.