Abstract Diet remains a promising target with the potential to impact colorectal cancer risk via the balance of energy and body weight, as well via specific food components with known anti-carcinogenic properties. Growing evidence implicates the importance of the gut microbiome in the interaction of diet, energy balance, and inflammation—key mechanisms associated with colorectal cancer development. The tight coupling of diet and the gut microbiota may hold new answers for colorectal cancer prevention. Methods: We conducted a pilot epidemiologic study among 50 cancer-free MD Anderson colonoscopy patients with no underlying familial/genetic or chronic inflammatory bowel conditions. With the overall goal to identify dietary intervention targets in obese and lean individuals at varied risk of colorectal neoplasia, patients provided fasting blood at colonoscopy, stool samples via mail ~ 1 month post-polypectomy, and comprehensive dietary assessments. Total energy intake was derived from responses to both the Diet History (DH) and 24-hour dietary recalls (24HR) and diet quality was defined by the Healthy Eating Index (HEI). We conducted 16S rRNA gene sequencing of the fecal microbiota and measured circulating blood adipocytokines via Luminex multiplex assays (Millipore). Results: Majority of colonoscopy patients recruited into the study were female (67%) and diagnosed with precancerous colorectal polyps (57%), primarily tubular and sessile serrated adenomas. Patients with precancerous polyps removed at colonoscopy were more likely to be obese than patients without precancerous polyps, but otherwise similar by age, macronutrient composition, and HEI. Differences in the overall composition of the fecal microbiome via weighted and unweighted unifrac were observed by reason for colonoscopy (symptomatic vs asymptomatic screening), chronic medication use (yes vs. no), total energy and alcohol intake (tertiles); and effects of each these factors were considered in subsequent analyses. No differences were observed by other clinicopathologic features, demographics, or lifestyle factors. High total energy intake was significantly correlated with higher alpha diversity of the fecal microbiome. This linear relationship remained following adjustment for diet quality and was consistent across colonoscopy status, BMI, medication use and alcohol intake (FDR-adj P<0.05). Total energy intake was inversely correlated with the relative abundance of Bacteroides (r2=0.27; FDR-adj P<0.001), while diet quality was positively associated with multiple bacteria involved in short chain fatty acid metabolism, including Faecalibacterium, Subdoligranulum, and Pseudobutyrivibrio (all P<0.05). Many of these same bacteria were significantly correlated with blood levels of circulating adipocytokines, including C-peptide, TNF-alpha, lipocalin-2, and PAI-1. Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt) analysis further revealed that gut bacteria may mediate the relationship between diet quality and adipocytokine signaling pathways. Conclusion: Modifiable aspects of diet quality and quantity were more closely linked to the composition of the gut microbiome than obesity or a recent diagnosis of precancerous colorectal polyps. In general, low microbial diversity has been associated with states of disease/poor health, but in our case, eating more may mean exposure to a more varied diet, promoting proliferation of a number of distinct bacteria, both “good” and “bad”. We plan to further explore functional relationships and pathways between diet, the microbiome, and adipocytokines with ongoing metagenomic analysis of the fecal microbiome. Citation Format: Carrie R. Daniel, Kristi L. Hoffman, Akhil Sood, G.S. Raju, Samir M. Hanash, Xifeng Wu, Joseph F. Petrosino. Diet quality vs quantity in relation to the fecal microbiome among colonoscopy patients. [abstract]. In: Proceedings of the AACR Special Conference on Colorectal Cancer: From Initiation to Outcomes; 2016 Sep 17-20; Tampa, FL. Philadelphia (PA): AACR; Cancer Res 2017;77(3 Suppl):Abstract nr A26.
Read full abstract