In this study, we investigated the oral inflammatory status and the compositions of the oral microbiome in women with PTB and compared them to women with TB. The 2017 World Workshop Classification of Periodontal Diseases was used to set explicit and prospectively comparable disease definitions that were absent from previous studies, making them challenging to align and compare [39,53,54]. With a median age of 31 (IQR 27–34) years, the women in our cohort were below a maternal age of >35, which is considered a risk factor for preterm delivery [55]. The dental status represented by the DMFT index was similar in both groups. Notably, it was below representative values for the German population reported in The Fifth German Oral Health Study (DMS-V) and comparable to results in another study on pregnant women [56,57]. In a systematic review, PTB was not associated with higher DMFT [58]. While socioeconomic status has been reported to affect dental caries during pregnancy in our cohort [59], we did not observe a relationship between educational background and DMFT. 4.1. CAL, PPD, and PISAThe periodontal CAL and the relative number of PPD ≥ 4 mm were significantly higher in women with PTB. However, no significant associations were observed with GBI or BOP. Pathological periodontal pockets and tissue breakdown occur when immunological mechanisms fail to contain and resolve an existing gingival inflammation [60]. Consequently, dysbiosis, in combination with a florid immunological reaction, causes the increased breakdown of the periodontal connective tissue [61]. However, during pregnancy, increased periodontal pockets cannot result only from a periodontal breakdown but also from gingival enlargement [62], where increased pocket depths, but not increased CAL, are observed. Pregnant women are more prone to gingival inflammation, as the altered levels of sex steroids during pregnancy lead to higher vascularization of the gingiva and proliferation of fibroblasts [63,64]. In this cohort, the periodontal parameters CAL and PPD for women with PTB were aggravated, whereas BOP, PISA, and GBI were similar in women with TB and PTB. Both bleeding parameters were present in 41]. GBI and BOP are positively impacted by improved domestic oral hygiene procedures, which could arise due to a higher awareness of oral hygiene during pregnancy. Subgingival sites with >2 mm PPD are usually not accessible for domestic hygiene procedures [65,66]. Those inflamed areas need professional subgingival periodontal treatment. It can be assumed that the equal bleeding values of both groups are a result of a generally high awareness for domestic oral hygiene procedures. The number of sites with deep pockets and BOP was masked through the much higher number of sites with PPD 4.3. MicrobiomeA limited number of studies have used 16S rRNA gene sequencing technologies to investigate the oral microbiome of pregnant women [38]. To our best knowledge, this is the first study comparing the oral microbiome of women with PTB and TB. We explored changes in the supra- and sub-gingival microbiomes using 16S rRNA-gene sequencing of plaque samples. We found the α-diversity, dominance, and evenness of the microbiomes to be significantly affected by PTB. A significantly higher dominance and lower evenness within PTB microbiomes indicate a potential overrepresentation of certain bacteria and dysbiotic bacterial shift. Moreover, we found relationships between CAL, WOC, and the number of previous abortions and PTB, underscoring our other findings. Previous studies found high abundances of the phyla Firmicutes, Bacteroidetes, Fusobacteria, Proteobacteria, and Spirochaetes in pregnant women [69,70]. Here, Leptotrichia was more abundant in pregnant women with gingivitis compared to gingival health, while Veillonella was more abundant in women with gingivitis, and different operational taxonomic units (OTU) of the genus Prevotella were present in both groups [69]. We also found significant differences in the abundance of several RSVs between groups. Atopobium rimae has been associated with autoimmune diseases such as Systemic Lupus Erythematosus [71] as well as chronic tonsillitis, infection-related glomerulonephritis, and splenic abscesses [72,73,74] and with IL−1beta and IL8 in GCF [75]. Here, it was less abundant in the supragingival samples of our PTB group. Bifidobacteria spp associated with periodontal health [76] as well as caries [77] were also less abundant (−log2 fold changes of −7.90 and −7.75) in supragingival plaque of PTB. Lautropia spp and Lautropia mirabilis were associated with periodontal health [78]. Prevotella spp were also more abundant (−log2 fold changes of 9.33, 3.89, and 9.27) in the PTB group. Prevotella belongs to the phylum Bacteroidetes, a common periodontal pathogen of the orange complex [79], and is associated with bacterial vaginosis [80]. Prevotella and other oral species may impair neutrophil leukocyte function and produce collagenases and fibrinolysins, which may collectively contribute to the induction of preterm birth [81]. A connection between Prevotella and preterm birth has also been reported in a study investigating the vaginal microbiome and metabolome of women with preterm membrane rupture, which found Prevotella to be among the most abundant species within the preterm membrane rupture group [82].Moreover, in agreement with the findings of Miranda-Rius, we found a higher abundance of proteobacteria but not actinobacteria within the supragingival samples of the PTB group [83]. Proteobacteria are generally associated with periodontal health, particularly L. mirabilis, which was identified as part of the core microbiome [78,84]. Therefore, the overabundance of proteobacteria illustrates that the supragingival microbiomes in PTB women do not undergo a substantial pathogenic shift and still display a partially healthy profile. The pathogenic shift occurs in the periodontal pockets and is driven by the more pronounced subgingival inflammatory environment [85].In subgingival samples, three RSVs were differentially abundant in PTB and TB groups: (1) Corynebacterium matruchotii, which is known to form corncobs in human plaque and is part of the oral core-microbiome [86]; (2) Leptotrichia spp, which is also part of the core oral microbiome and, when overabundant, associated to halitosis and oral leukoplakia [87,88] was less abundant (−log2 fold changes of −1.99 and −5.19, respectively); and (3) Abiotrophia defectiva was more abundant (−log2 fold change = 3.91). An overall increase in proteobacteria was observed in both supragingival and subgingival samples from the PTB group.The microbial profile of placentas based on 16S rRNA sequencing was recently reported by Miranda-Rius et al. (2021). In this study, three groups were compared: (1) women with adverse pregnancy outcomes and periodontitis; (2) women with adverse pregnancy outcomes without periodontitis; and (3) periodontally healthy women with TB. The diversity of the placental microbiome was found to differ significantly between the groups, where periodontitis was a more discriminant variable than adverse pregnancy outcomes, and a principal coordinates analysis (PCoA) found significant differences in β-diversity (p83]. This observation is consistent with our findings, where aerococcaceae A. defectiva was associated with PTB in women with periodontitis. Abiotrophia spp is an apparent potent inducer of proinflammatory cytokines [89], and A. defectiva is reportedly associated with endocarditis [90], indicating it is a pathogen with high systemic inflammation-inducing activity.The major periodontal pathogen Porphyromonas gingivalis—a periodontal keystone pathogen [16]—was found in the amniotic fluid and the oral cavity of a subgroup of women with PTB and periodontitis [91]. In our study, there was no significant difference in the abundance of P. gingivalis between groups (−log2 fold change = 1.23, p = 0.999). Similarly, P. gingivalis, T. forsythensis, T. denticola, and F. nucleatum abundance did not differ significantly in women with PTB and TB in a study using Checkerboard DNA–DNA hybridization [92]. Since P. gingivalis is a widespread periodontal pathogen, its equal abundance in both our study groups is to be expected. Notably, it is a potent inducer of dysbiotic conditions and, with its subversive abilities regarding immunological defense mechanisms, a potent inducer of pathogenic conditions [93]. As subgingival dysbiosis progresses, the relative quantity of periodontal pathogens generally increases in periodontal pockets, and thus the chance of systemic dissemination increases [94]. The ulceration of the gingival epithelium and hyperpermeability of blood vessels favors the invasion of bacterial products, inflammatory mediators, and microorganisms into the bloodstream, elevating the risk of effects in distant body sites [25]. Therefore, systemic dissemination of P. gingivalis might occur in its presence, with even domestic mechanical plaque removal associated with systemic bacteriemia [2,95].

In summary, our sequencing results on the oral sub- and supra-gingival microbiome indicate a possible association of the microbial composition and specific genera and species with the occurrence of PTB. The exact role of specific species, their interactions, and the impact of their metabolic products on PTB represent urgent areas for future research. Future studies with larger numbers of participants should evaluate differences in the oral and placental microbiome, using not only 16srRNA but full metagenome sequencing and proteome and metabolome analyses.

Significant limitations of our study were the size of its cohort and that we were not able to include the full calculated sample size. Hence, some p-values need to be interpreted with caution. Moreover, it was not possible to conduct all examinations on all women. These factors were partly due to participants who had recently given birth being in a stressful situation in an unfamiliar clinical environment, where psychological strain and physiological burden negatively impacted their willingness to participate in this study. The long hospitalization period of some women prior to examination may have reduced oral hygiene, even though gingival inflammatory conditions were not elevated in our cohort. This factor might not have affected clinical attachment loss but could affect the pocket depth. The number of decayed teeth was assessed only clinically and without radiographs due the hospitalization of the patients. Nevertheless, a strength of our study is the detailed periodontal examination with measurements on six sites per tooth, including clinical attachment loss and the use of the classification of periodontal diseases recommended in 2017. These factors will improve the comparability of our findings in future studies. In addition, the use of 16S rRNA sequencing advanced our understanding of oral microbiome composition and increased the feasibility of future comparisons between the oral and placental microbiome. Our findings highlight the importance of professional oral care before and during pregnancy, especially in diagnosing and treating periodontal disease.