Fig. 1

Decreased vaginal Lactobacillus spp. abundance occurs prior to PPROM and is further exacerbated by membrane rupture and erythromycin treatment. a Ward’s linkage hierarchical clustering analysis of vaginal bacterial species data from cervical vaginal fluid samples (n = 165) collected from women with uncomplicated term delivery, sampled at 28 weeks (n = 20), pre-PPROM (n = 15), following PPROM before erythromycin (n = 39), after 48 hours of erythromycin (n = 43), 48 hours to 1 week of erythromycin (n = 22) and >1 week of erythromycin treatment (n = 26). Vaginal bacterial communities were classified based on Lactobacillus spp. abundance into dominant, intermediate and depleted, and further into eight vaginal microbiota groups: VMG 1: L. iners dominant, VMG 2: L. iners high diversity, VMG 3: L. crispatus dominant, VMG 4: L. gasseri dominant, VMG 5: L. jensenii dominant, VMG 6: L. crispatus high diversity, VMG 7: Lactobacillus spp. depleted and low diversity, VMG 8: Lactobacillus spp. depleted and high diversity. b Relative Lactobacillus spp. abundance is significantly lower in the pre-PPROM and membrane rupture groups compared to gestation age matched and normal pregnancy controls (P = 0.011). Erythromycin treatment exacerbates Lactobacillus spp. depletion and expansion of dysbiotic vaginal communities (P = 0.001). Reduced Lactobacillus spp. abundance is accompanied by a reciprocal increase in c richness and d diversity. e Bacterial load is significantly higher pre-membrane rupture in comparison to post-membrane rupture (P = 9.37 × 10^-8) but remains stable thereafter, despite ongoing erythromycin treatment. PPROM preterm prelabour rupture of the fetal membranes