Abstract Background Although epidemiological evidence is limited, experimental data indicated a tumor-suppressive effect of yogurt containing Bifidobacterium. We hypothesized that the association between yogurt intake and colorectal cancer incidence differed by Bifidobacterium status in tumor tissue. Methods We utilized data from two prospective cohort studies: the Nurses’ Health Study and the Health Professionals Follow-up Study. Inverse probability weighted multivariable Cox proportional hazards regression was used to assess differential associations of yogurt intake with the incidence of colorectal carcinomas subclassified by tumor Bifidobacterium status. Results During follow-up of 132,056 individuals, we documented 2,950 incident colorectal cancer cases, including 1,191 with available Bifidobacterium data. The association between yogurt intake and colorectal cancer incidence differed by Bifidobacterium status (Pheterogeneity = 0.03). This differential association became stronger in a subgroup analysis of proximal colon cancer (Pheterogeneity = 0.003). Multivariable hazard ratios in individuals who consumed ≥ 2 servings/week (vs. < 1 serving/month) of yogurt were 0.43 [95% confidence interval (CI), 0.22-0.85] for Bifidobacterium-positive proximal colon cancer, and 1.40 (95% CI, 0.95-2.05) for Bifidobacterium-negative proximal colon cancer. Conclusion Yogurt intake was inversely associated with the incidence of Bifidobacterium-positive (but not Bifidobacterium-negative) proximal colon cancer, suggesting an interactive influence of yogurt intake and intestinal microbiota on tumor development. Disclosures All Authors: No reported disclosures
Read full abstract