Longitudinal Reduction in Diversity of Maternal Gut Microbiota During Pregnancy Is Observed in Multiple Low-Resource Settings: Results From the Women First Trial

Abstract

Objective: To characterize the changes in gut microbiota during pregnancy and determine the effects of nutritional intervention on gut microbiota in women from sub-Saharan Africa (the Democratic Republic of the Congo, DRC), South Asia (India and Pakistan), and Central America (Guatemala).

Methods: Pregnant women in the Women First (WF) Preconception Maternal Nutrition Trial were included in this analysis. Participants were randomized to receive a lipid-based micronutrient supplement either ≥3 months before pregnancy (Arm 1); started the same intervention late in the first trimester (Arm 2); or received no nutrition supplements besides those self-administered or prescribed through local health services (Arm 3). Stool and blood samples were collected during the first and third trimesters. Findings presented here include fecal 16S rRNA gene-based profiling and systemic and intestinal inflammatory biomarkers, including alpha (1)-acid glycoprotein (AGP), C-reactive protein (CRP), fecal myeloperoxidase (MPO), and calprotectin.

Results: Stool samples were collected from 640 women (DRC, n = 157; India, n = 102; Guatemala, n = 276; and Pakistan, n = 105). Gut microbial community structure did not differ by intervention arm but changed significantly during pregnancy. Richness, a measure of alpha-diversity, decreased over pregnancy. Community composition (beta-diversity) also showed a significant change from first to third trimester in all four sites. Of the top 10 most abundant genera, unclassified Lachnospiraceae significantly decreased in Guatemala and unclassified Ruminococcaceae significantly decreased in Guatemala and DRC. The change in the overall community structure at the genus level was associated with a decrease in the abundances of certain genera with low heterogeneity among the four sites. Intervention arms were not significantly associated with inflammatory biomarkers at 12 or 34 weeks. AGP significantly decreased from 12 to 34 weeks of pregnancy, whereas CRP, MPO, and calprotectin did not significantly change over time. None of these biomarkers were significantly associated with the gut microbiota diversity.

Conclusion: The longitudinal reduction of individual genera (both commensals and potential pathogens) and alpha-diversity among all sites were consistent and suggested that the effect of pregnancy on the maternal microbiota overrides other influencing factors, such as nutrition intervention, geographical location, diet, race, and other demographical variables.

Keywords: gut microbiota; inflammation; low middle income countries; pregnancy; small quantity lipid-based nutrient supplements (SQ-LNS).

FIGURE 1

Adjusted mean estimate for alpha-diversity values by site and time (12 vs. 34 weeks). Each panel represents one pairing of site and alpha-diversity measure [eveness, richness, and Shannon diversity (richness and eveness)]. Each value is the adjusted mean estimate from the GEE model (diamond point) and 95% CI (error bars) for the respective measure of alpha-diversity split by time point and supplement status. Supplement status “Yes” indicates Arm 1 at 12 weeks and Arms 1+2 at 34 weeks. Supplement status “No” indicates Arm 2 at 12 weeks and Arm 3 at 34 weeks. The corresponding p-values for tests of time and supplement status are overlayed in the top-right corner of each panel.

FIGURE 2

Bray-Curtis PCoA plot of all sites and timepoints colored by Shannon H index. PCoA1 shows a separation of observations based upon a measure of eveness. The percentage of variance explained by each axis is presented in parentheses within the figure legends. Together, the first two components explain just under 4% of the variation in the Bray-Curtis metric.