Association between antibiotic consumption and colon and rectal cancer development in older individuals: A territory-wide study.

ABSTRACT

BACKGROUND: Antibiotics may alter colorectal cancer (CRC) risk due to gut dysbiosis. We aimed to study the specific and temporal effects of various antibiotics on CRC development in older individuals. METHODS: This was a territory-wide retrospective cohort study. Subjects aged 60 years and older who did not have CRC diagnosed on screening/diagnostic colonoscopy diagnosed between 2005 and 2013 were recruited. Exclusion criteria were history of CRC, colectomy, inflammatory bowel disease, and CRC diagnosed within 6 months of index colonoscopy. Exposure was use of any antibiotics up to 5 years before colonoscopy. The primary outcomes were CRC diagnosed >6 m after colonoscopy. Covariates were patient demographics, history of colonic polyps/polypectomy, concomitant medication use (NSAIDs, COX-2 inhibitors, aspirin, and statins), and performance of endoscopy centers (colonoscopy volume and polypectomy rate). Stratified analysis was conducted according to nature of antibiotics and location of cancer. RESULTS: Ninety seven thousand one hundred and sixty-two eligible subjects (with 1026 [1.0%] cases of CRC) were identified, 58,704 (60.4%) of whom were exposed to antibiotics before index colonoscopy. Use of antibiotics was associated with a lower risk of cancer in rectum (adjusted hazard ratio [aHR] 0.64, 95% CI 0.54-0.76), but a higher risk of cancer in proximal colon (aHR 1.63, 95%CI 1.15-2.32). These effects differed as regards the anti-anaerobic/anti-aerobic activity, narrow-/broad-spectrum, and administration route of antibiotics. CONCLUSIONS: Antibiotics had divergent effects on CRC development in older subjects, which varied according to the location of cancer, antibiotic class, and administration route.