The Vaginal Microbiota Composition and Genital Infections during and after Pregnancy among Women in Pemba Island, Tanzania

Abstract

We investigated the vaginal microbiota (VMB) composition, prevalence of genital pathogens and their association among pregnant and post-delivery women in Pemba Island, Tanzania. Vaginal swabs were collected from 90 women, at two time points during pregnancy (<20 weeks of gestational age [GA] and ≥20 weeks GA) and once after delivery, when possible. IS-pro assay was used for VMB characterization. Chlamydia trachomatis (CT), Neisseria gonorrhea (NG), Trichomonas vaginalis (TV), Mycoplasma genitalium (MG) and human papillomavirus (HPV) were detected by qPCRs. VMB were mostly Lactobacillus dominant during pregnancy and non-Lactobacillus dominant post-delivery. A significant decrease in VMB richness was observed during pregnancy among paired and unpaired samples. Shannon diversity was significantly lower during pregnancy than post-delivery among unpaired samples. Klebsiella species and Streptococcus anginosus were the most commonly identified pathobionts at all timepoints. A high abundance of pathobionts was mostly seen in women with non-Lactobacillus dominant VMB. At ≥20 weeks GA timepoint during pregnancy, 63.0% of the women carrying one or more genital pathogen (either HPV, CT, TV, or MG) had L. iners dominant VMB. NG was not detected pre-delivery. This study contributes evidence on VMB composition, its changes during pregnancy and post-delivery, and their association with pathobionts and genital pathogens.

Keywords: genital pathogens; pathobionts; vaginal microbiota.

Figure 1

Boxplot for the richness (A) and Shannon diversity index (B) at the species level for each collection point. Results of the <20 weeks GA pregnancy collection are in blue (n = 44), ≥20 weeks GA pregnancy collection in yellow (n = 82), and post-delivery in grey (n = 44). (A) Species richness is lower at ≥20 weeks GA pregnancy collection compared to the pregnancy collection at <20 weeks GA (p = 0.02). Between the other timepoints there was no significance difference. (B) The Shannon diversity index is higher at post-delivery compared to the index at ≥20 weeks GA pregnancy collection (p = 0.03), but not compared to the diversity index at <20 weeks GA pregnancy collection. During the pregnancy, the index did not differ significantly. NS = non significant. * p-value < 0.05.

Figure 2

Boxplots for species richness at each collection point for paired samples. Results of <20 weeks GA pregnancy collection are in blue, ≥20 weeks GA pregnancy collection in yellow, post-delivery in black and overall during pregnancy in red. (A) The richness was significantly lower at the ≥20 weeks GA pregnancy collection point than at <20 weeks GA pregnancy collection (p = 0.02) in matched samples from 38 women. (B) There was no significant difference in the richness between <20 weeks GA pregnancy collection and post-delivery matched samples from 6 women. (C) No significant difference was calculated in the richness between ≥20 weeks GA pregnancy collection and post-delivery matched samples from 38 women. (D) For 42 women that had samples collected at least once during pregnancy and post-delivery, no significant difference in the species richness was calculated. NS = non significant. * p-value < 0.05.

Figure 3

Boxplots for the Shannon diversity index at species level for paired samples. Results of <20 weeks GA pregnancy collection are in blue, ≥20 weeks GA pregnancy collection in yellow, post-delivery in black and pregnancy in red. (A) The Shannon diversity index was not significantly higher at the ≥20 weeks GA pregnancy collection point than at the <20 weeks GA pregnancy collection in matched samples from 38 women. (B) There was no significant difference in the Shannon diversity index between the <20 weeks GA pregnancy collection and post-delivery matched samples from 6 women. (C) No significant difference was calculated in the Shannon diversity between the ≥20 weeks GA pregnancy collection and post-delivery matched samples from 38 women. (D) For 42 women who had samples collected at least once during pregnancy and post-delivery, no significant difference in the Shannon diversity index was calculated. NS = non significant.

Figure 4

Frequency of five community state type (CST) identified each collection timepoint (n = 44 at <20 weeks GA and post-delivery timepoints; n = 82 at the ≥20 weeks GA timepoint). In the pie charts the CST cluster is given in colour: CST I, blue; CST II, green; CST III, red; CST IV, yellow; CST V, purple.

Figure 5

Schematic representation of the switch between community state types (CSTs) between two sampling timepoints. The numbers in the circles indicate how many vaginal samples were clustered in a certain CST by the time collection (during pregnancy). Arrows represent the direction of the switch. The numbers by the arrows represent the number of vaginal microbiota tested belonging to the same women who switched to a given CST at a later timepoint. (A) Thirty-eight women were clustered with a specific CST at timepoint <20 weeks GA and ≥20 weeks GA during pregnancy, of which 18 CST changed type during pregnancy. (B) Among the paired vaginal swabs tested at ≥20 weeks GA and post-delivery, 27 of the 38 CST changed. (C) The CST of three of the vaginal swabs from the same six women tested at <20 weeks GA and post-delivery changed.

Figure 6

Nine pathobionts observed in vaginal samples per collection point. Numbers of vaginal samples detected with pathobionts at <20 weeks GA pregnancy collection (n = 44), ≥20 weeks GA pregnancy collection (n = 82), and post-delivery (n = 44). Blue refers to the first timepoint during pregnancy, yellow to the second timepoint during pregnancy and grey to the post-delivery timepoint.