Significant Differences in Fecal Microbiota Are Associated with Various Stages of Glucose Tolerance in African-American Male Veterans

There is emerging evidence that intestinal microbiota is a contributor to the metabolic/glycemic phenotype. While changes in microbiota have been described in obesity and diabetes, little is known about microbiota composition in various dysglycemic states.

This study aimed to investigate the relationship between microbiota and changes in the glycemic control of prediabetic subjects.

Stool was collected from African American men participating in a randomized controlled trial of vitamin D Intervention at VA. Four groups (Gr) were characterized based on changes in OGTT between baseline (T0) and the study completion at 12 months (T12): Gr 1- stable normal glucose tolerance (GT); Gr 2- stable impaired fasting glucose or stable impaired GT; Gr 3 - worsened GT; and Gr 4 - improved GT. Microbiota DNA was extracted from stool collected at T12, analyzed using high-throughput next-generation sequencing of microbial rRNA genes and data processed using established bioinformatics pipelines. Microbiota (composition, alpha diversity, abundance) was analyzed in 116 subjects: Gr 1= 35, Gr 2 = 27, Gr 3 = 24, and Gr 4 = 29.

At Phylum level significant differences in bacterial composition were observed between Gr 1 and Gr 2 (p= 0.03) and a trend to significance for Gr1 vs Gr3 (p= 0.06), and Gr 1 vs Gr4 (p= 0.06). Bacteroidetes was higher, Firmicutes lower, and hence the Bacteroidetes/Firmicutes ratio (B/F) was lower with worsening glycemic control (B/F: Gr 1 vs Gr 2 = 1.9 vs 0.9, p= 0.01; and Gr 1 vs Gr 3 = 1.9 vs 1.1, p= 0.04). Proteobacteria decreased in Gr 2 and Gr 4 compared to Gr 1 (p=0.01 for both). Similarly, there were significant differences in microbiota at the Family and Genus levels. In Gr 2 vs  Gr1 there was less Prevotella (hence higher Bacteroides/Prevotella ratio, 5.6 vs 2.7, p= 0.05), less Enterobacteriacea (p=0.03), and more Ruminococcae (p= 0.01) and Veillonelacea (p= 0.02). Notably, Akkermansia was more abundant in Gr 4 vs Gr 1 (p=0.04).

We speculate that lower abundance of Prevotella may be associated with worsening glycemia, and conversely higher abundance of Akkermansia might be associated with improving glycemia, thus corroborating suggestions from previous studies. Ruminococcae might be associated with higher insulin level (seen in previous research), and therefore conducive to maintenance of stable glycemia. Observed association of Veillonelacea and glycemia was novel.

In conclusion, significant differences in microbiota (composition, alpha diversity, abundance) were observed in various GT states. Interesting most of the differences were seen between normoglycemic subjects and those who were remained prediabetics at the end of the study. These findings suggest that there may be a certain makeup of the gut microbiota associated with steady glycemic states. Further studies are needed to evaluate whether causative relationship exist between microbiota and changes in GT.