Longitudinal Associations Between the Microbiome and Glucose Homeostasis Over Time in The Microbiome and Insulin Longitudinal Evaluation Study (MILES)
Abstract Body: Type 2 diabetes (T2D) is associated with both relative, and absolute, changes in the microbes colonizing the digestive tract (the gut microbiome). However, T2D-related pathogenesis predates diagnosis by up to decade, and few studies have examined whether gut microbiome differences also relate to the various components of insulin and glucose homeostasis that contribute to onset. The goal of this study was to conduct such an investigation using repeated stool samples from 114 non-Hispanic White and 44 African American individuals without T2D participating in MILES (mean age 60, 67% female), leveraging MILES’ longitudinal design to minimize the influence of temporary microbiome fluctuations. Participants underwent three oral glucose tolerance test, spaced ~ one year apart, from which fasting glucose (FG) and insulin, and C-peptide were derived, and insulin sensitivity, insulin secretion, and disposition index (DI) calculated. Microbes with a prevalence >= 30% were included in multi-level linear models which examined associations of overall composition (diversity, evenness and richness), as well as the relative abundance of individual species, with each of the six glucose measures. Models used centered log-ratio transformed data, included a correction for bias due to compositional effects, and controled for sequencing depth, batch, demographics, and medication useage. P-values equivalent to a<.05 after an FDR correction for the number of microbes (N=130) were considered significant. FG was associated with lower (Shannon index; b=-0.19(0.08),P=.03) and evenness (Inverse Simpson; b=-0.02(0.01),P=.02), with a similar trend inverse association seen for DI (diversity: b=-0.19(0.08),P=.03; evenness: b=0.02(0.01),P=.02). These two indices were also associated with the relative abundance of seven microbes; both FG (logFC=0.46,P=9.9*10-4) and DI (logFC=-0.54,P=1.1*10-4) were associated with Ruthenibacterium lactatiformans, FG only with Anaerotruncus colihominis (logFC=1.31,P=1.22*10-4) and Flavonifractor plautii (logFC=0.68,P=8.3*10-4), and DI only with Adlercreutzia equolifaciens (logFC =-0.55, P=4.6*10-5), Asaccharobacter celatus (logFC=-0.38,P=6.6*10-4), Enterorhabdus caecimuris (lofFG=-0.82,P=8.1*10-4), Gordonibacter pamelaeae (logFC=-0.49,P=5.9*10-4), and Limnochorda Pilosa (logFC=-0.40,P=1.2*10-3). The roles of these microbes in lactate, equol, butyrate, and tryptophan production point to the gut microbiome as a mediator of diet-T2D associations.
Wood, Lekki
( Baylor College of Medicine
, Houston
, Texas
, United States
)
Hoffman, Kristi
( Baylor College of Medicine
, Houston
, Texas
, United States
)
Bertoni, Alain
( WAKE FOREST UNIV SCHOOL MED
, Pfafftown
, North Carolina
, United States
)
Rich, Stephen
( UNIVERSITY VIRGINIA
, Charlottesville
, Virginia
, United States
)
Rotter, Jerome
( The Lundquist Institute
, Torrance
, California
, United States
)
Chen, Yii-der Ida
( The Lundquist Institute
, Torrance
, California
, United States
)
Goodarzi, Mark
( Cedars-Sinai Medical Center
, West Hollywood
, California
, United States
)
Author Disclosures:
Lekki Wood:DO NOT have relevant financial relationships
| Kristi Hoffman:No Answer
| Alain Bertoni:DO NOT have relevant financial relationships
| Stephen Rich:DO NOT have relevant financial relationships
| Jerome Rotter:No Answer
| Yii-Der Ida Chen:No Answer
| Mark Goodarzi:DO have relevant financial relationships
;
Consultant:Organon:Past (completed)
; Speaker:Nestle Health Science:Past (completed)