Gestational Diabetes Is Uniquely Associated With Altered Early Seeding of the Infant Gut Microbiota

Abstract

Gestational diabetes mellitus (GDM) is a worldwide public health problem affecting up to 27% of pregnancies with high predictive values for childhood obesity and inflammatory diseases. Compromised seeding of the infant gut microbiota is a risk factor for immunologic and metabolic diseases in the offspring; however, how GDM along with maternal obesity interact to alter colonization remains unknown. We hypothesized that GDM individually and in combination with maternal overweight/obesity would alter gut microbial composition, diversity, and short-chain fatty acid (SCFA) levels in neonates. We investigated 46 full-term neonates born to normal-weight or overweight/obese mothers with and without GDM, accounting for confounders including cesarean delivery, lack of breastfeeding, and exposure to antibiotics. Gut microbiota in 2-week-old neonates born to mothers with GDM exhibited differences in abundance of 26 microbial taxa; 14 of which showed persistent differential abundance after adjusting for pre-pregnancy BMI. Key pioneering gut taxa, including potentially important taxa for establishing neonatal immunity, were reduced. Lactobacillus, Flavonifractor, Erysipelotrichaceae, and unspecified families in Gammaproteobacteria were significantly reduced in neonates from mothers with GDM. GDM was associated with an increase in microbes involved in suppressing early immune cell function (Phascolarctobacterium). No differences in infant stool SCFA levels by maternal phenotype were noted; however, significant correlations were found between microbial abundances and SCFA levels in neonates. Our results suggest that GDM alone and together with maternal overweight/obesity uniquely influences seeding of specific infant microbiota in patterns that set the stage for future risk of inflammatory and metabolic disease.

Keywords: excess gestational weight gain; gestational diabetes; infant; maternal obesity; microbiota; short-chain fatty acids.

Figure 1

GDM and maternal overweight/obesity alter infant gut microbiota. Box plots of genus-level alpha diversities richness (Chao1), evenness (Shannon $\frac{H}{H_{\max }}$), and complexity (Shannon H) by (A) GDM and (B) GDM and weight group. Black dots represent outliers within strata. (C) Relative abundance of microbiota taxa for each infant at the genus level. Genus-level taxa shown are the 10 most abundant with the remaining classified as Other. Relative abundance estimated using negative binomial regression on 16S rRNA gene amplicon sequences from stool samples, shown as percent of total. GDM, gestational diabetes; NW, normal-weight; OW/OB, overweight/obese.

Figure 2

Summary of the impact of maternal GDM on infant gut microbial taxa in the interaction model. Box plots of centered log ratio (CLR) transformed (A) family- and (B) genus-level taxa counts by maternal GDM status and weight group. Taxa displayed are those whose relationship in infants from mothers with GDM differed significantly based on maternal weight group. GDM, gestational diabetes; NW, normal-weight; OW/OB, overweight/obese.

Figure 3

Heat maps of significant relationships between maternal phenotype and infant gut microbiota. Average taxa relative abundances are displayed by maternal phenotype with higher averages colored a deeper red. Cells with a shape inside indicate a significantly different (FDR adjusted p value <0.05) average relative abundance from the NW/No GDM group. Triangles indicate significant differences from the interaction models and circles indicate differences from the additive models for both (A) family- and (B) genus-level taxa.

Figure 4

Summary of the impact of maternal GDM on infant gut microbial taxa in the additive model. Box plots of centered log ratio (CLR) transformed (A) family- and (B) genus-level taxa counts by maternal GDM status. Taxa displayed are those whose abundance changed significantly in infants from mothers with GDM after controlling for maternal weight group. GDM, gestational diabetes.

Figure 5

Gut microbiota genus-level taxa correlate with infant stool SCFA levels. (A) Infant stool SCFA levels by maternal GDM status. Rocky Mountain Plots representing (B) acetate, (C) propionate, (D) butyrate, and (E) total SCFA measurements with a center log ratio transformation of genus-level counts. Correlations were derived using Spearman Correlation with r > |0.5| indicating a meaningful correlation. n = 23 for each measure.