AUTHOR=Dunlop Anne L. , Satten Glen A. , Hu Yi-Juan , Knight Anna K. , Hill Cherie C. , Wright Michelle L. , Smith Alicia K. , Read Timothy D. , Pearce Bradley D. , Corwin Elizabeth J. 

TITLE=Vaginal Microbiome Composition in Early Pregnancy and Risk of Spontaneous Preterm and Early Term Birth Among African American Women

JOURNAL=Frontiers in Cellular and Infection Microbiology

VOLUME=Volume 11 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/cellular-and-infection-microbiology/articles/10.3389/fcimb.2021.641005

DOI=10.3389/fcimb.2021.641005

ISSN=2235-2988

ABSTRACT=Objective.  To evaluate the association between the early pregnancy vaginal microbiome and spontaneous preterm birth (sPTB) and early term birth (sETB) among African American women. 

Methods. Vaginal samples collected between 8-14 weeks’ gestation from 436 pregnant African American women underwent 16S rRNA gene sequencing of the V3-V4 region, taxonomic classification, and community state type (CST) assignment.  We compared vaginal CST and abundance of taxa for women whose pregnancy ended in sPTB (N = 44) or sETB (N= 84) to those who delivered full term (N = 231).  

Results. Nearly half had a vaginal microbiome classified as CST IV (Diverse CST), while one-third had CST III (L. iners dominated) and 16% had CST I, II, or V (non-iners Lactobacillus dominated). Compared to vaginal CST I, II, or V (non-iners Lactobacillus dominated), both CST III (L. iners dominated) and CST IV (Diverse) were associated with sPTB with an adjusted odds ratio (95% confidence interval) of 4.1 (1.1, infinity) and 7.7 (2.2, infinity), respectively. In contrast, no vaginal CST was associated with sETB. The linear decomposition model (LDM) based on amplicon sequence variant (ASV) relative abundance found a significant overall effect of the vaginal microbiome on sPTB (p=0.034) but not sETB (p=0.320), whereas the LDM based on presence/absence found no overall effect on sPTB (p=0.328) but a significant effect on sETB (p=0.030). The LDM found that no ASV was significantly associated with sPTB after controlling for multiple comparisons (FDR 10%), although in marginal analysis the relative abundance of Gardnerella vaginalis (p=0.011), non-iners Lactobacillus (p=0.016), and Mobiluncus curtisii (p=0.035) and the presence of Atopobium vaginae (p=0.049), BVAB2 (p=0.024), Dialister microaerophilis (p=0.011), and Prevotella amnii (p=0.044) were associated. The LDM identified the higher abundance of 7 ASVs and the presence of 13 ASVs, all commonly residents of the gut, as associated with sETB at FDR < 10%.  

Conclusions. In this cohort of African American women, an early pregnancy vaginal CST III or IV was associated with an increased risk of sPTB but not sETB. The relative abundance and presence of distinct taxa within the early pregnancy vaginal microbiome was associated with either sPTB or sETB.