Recent studies implicate the gut microbiome in the etiology of obesity. However, the contribution of the microbiome remains uncertain due to conflicting results. Previous studies have examined different populations and ages, but most have focused on the fecal microbiome, while few studies have examined the role of microbial diversity at other body sites. To explore the association between obesity and the microbial diversity of the upper digestive tract, we studied in cross-section a cohort of 659 healthy Chinese adults (ages 34 to 67 years) with a measured body mass index range of 15.0 to 35.7. We characterized the upper gastrointestinal tract microbiome using the HOMIM DNA microarray, which tests for the presence of approximately 300 species/strains of bacteria. We characterized alpha-diversity using taxa counts, and we created beta-diversity metrics using an unweighted unifrac distance matrix of pairwise comparisons of the whole cohort. We performed unsupervised clustering in this matrix and found support for three clusters, and we generated principal coordinate vectors from the matrix. Body mass index was examined as a continuous variable on the log scale in regression models adjusted for age, sex, tobacco smoking, alcohol drinking, antibiotic use, sample collection device, and the number of decayed, missing, and filled teeth. We found that alpha-diversity (numbers of strains, species, genera, families, orders, classes, or phyla) was not associated with body mass index. The presence or absence of each individual species/strain of bacteria on the HOMIM DNA microarray was not associated with body mass index after Bonferroni correction for multiple comparisons. However, beta-diversity, as assessed by cluster membership (first cluster compared with third cluster, p=0.0121; second cluster compared with third cluster, p=0.0002) and principal coordinate vectors, scaled to their interquartile ranges (p-values for the third and fourth vectors were p=0.0132 and p=0.0280, respectively), were associated with body mass index. Moreover, in adjusted models with multiple microbial diversity metrics, cluster membership remained independently associated with body mass index, while the vector associations were attenuated. Individuals in the first (median body mass index 22.35, interquartile range 19.17-23.31) and second (median body mass index 22.55, interquartile range 20.50-24.38) clusters had lower median body mass indexes that fall within the optimum Asian BMI range than the median body mass index in the third cluster (23.59, interquartile range 21.50-25.32), which is considered to fall in the range of increased risk for obesity-related diseases in Asian populations. Our results demonstrate for the first time that beta-diversity in the upper digestive tract microbiome is associated with body mass index. Future longitudinal, prospective studies should address whether the upper digestive tract microbiome plays a causal role for obesity.

This proffered talk is also presented as Poster 56.

Citation Format: Shih-Wen Lin, Neal D. Freedman, Jianxin Shi, Mitchell H. Gail, Guoqin Yu, Vanja Klepac-Ceraj, Bruce J. Paster, Bruce A. Dye, Wen-Qiang Wei, Jin-Hu Fan, You-Lin Qiao, Sanford M. Dawsey, Christian C. Abnet. Beta-diversity metrics of the upper digestive tract microbiome are associated with body mass index. [abstract]. In: Proceedings of the AACR Special Conference on Post-GWAS Horizons in Molecular Epidemiology: Digging Deeper into the Environment; 2012 Nov 11-14; Hollywood, FL. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2012;21(11 Suppl):Abstract nr PR8.