Background
The human oral cavity harbors a complex ecosystem of hundreds of different microbes primarily as biofilms on the surfaces of the teeth [
1]. The healthy oral microbiota is in a symbiotic state with its host, but dysbiosis of the oral microbial community often results in development of dental disease [
2,
3]. Increasing evidence suggests that the oral microbiota is able to migrate to extra-oral habitats, causing inflammation [
4]. Epidemiological studies have linked periodontal disease to obesity [
5], insulin resistance [
6] and type 2 diabetes [
7]. In a rodent study, oral administration of
Porphyromonas gingivalis, a periodontal pathogen, induced systemic inflammation and insulin resistance [
8]. A meta-analysis found that periodontal treatment in type 2 diabetes patients improved the glycaemic regulation in a 3 months follow-up period [
9]. All together, the available literature provides suggestive evidence of a role of oral microbiota in the pathophysiology of metabolic disorders.
Gestational diabetes mellitus (GDM) is one of the most common obstetric complications, characterized by hyperglycaemia during pregnancy [
10]. Gestational diabetes is associated with adverse pregnancy outcomes [
11,
12] and has long-term consequences for both the women and their offspring [
11‐
15]. Like GDM, periodontal disease has been linked to adverse pregnancy outcomes such as preterm birth [
16], low birth weight [
16,
17] and pre-eclampsia [
18‐
20], suggesting a possible involvement of the oral microbiota in GDM development.
Interestingly, evidence suggest similarities between the placental and oral microbiota [
21] and recent reports indicate that women with GDM have an altered placental microbiota compared with normoglycaemic women [
22,
23]. In a murine model of translocation of oral bacterial species, human salivary microbiota injected into mice was identified in the placenta, with higher abundance compared with the original abundance in the oral cavity [
24]. If the salivary microbiota is influenced by GDM it might contribute to the future increased risk of disorders in the offspring by affecting the placental microbiota and may likewise be a potential causal factor for the increased risk of later metabolic disorders in women who have suffered from GDM.
In the present study we assessed whether GDM is linked to variation in the oral microbial community by examining the diversity and composition of the salivary microbiota in third trimester of pregnancy and about 9 months postpartum.
Discussion
Our study of pregnant women demonstrated that, at the time of diagnosis, GDM is only associated with a moderately altered composition of the salivary microbiota. However, 9 months after delivery, GDM during the preceding pregnancy was associated with an aberrant salivary microbiota composition.
We found that richness of the salivary microbiota decreased from pregnancy to 9 months after delivery. A previous study of salivary microbiota over the course of a healthy pregnancy found higher bacterial richness in early pregnancy compared with late pregnancy in the same pregnant women. When comparing the salivary bacterial richness to a matched non-pregnant control group of women the authors found richness in early pregnancy significantly higher than in non-pregnant women, but did not report on statistical significance when comparing richness in late pregnancy with non-pregnant women [
32]. Our finding is somewhat in line with previous observations of increased richness of salivary microbiota in pregnancy when comparing with non-pregnant adults, although our study represents two different time points of samples from the same individuals, and the previous study [
32] compared a group of pregnant women to a group of matched non-pregnant controls.
Compared to first trimester, late pregnancy is characterized by increased levels of proinflammatory cytokines and insulin resistance [
33] and in our study plasma glucose levels were, not surprisingly, higher during pregnancy than postpartum which could be expected to be mirrored in an increased salivary glucose level, as it has been described in non-pregnant adults [
34]. Increased levels of salivary glucose have been associated with decreased bacterial richness in adolescents and salivary bacterial count could predict high salivary glucose concentration (salivary glucose ≥1.0 mg/dL) more accurately than clinical characteristics such as BMI and fitness levels [
35]. These results are somewhat surprising as glucose is an energy source for many oral bacteria, however, the authors of the study argued that oral hyperglycaemia changed the oral microbiota environment by salivary acidification. In studies of the human gut microbiota, richness has been found to decrease with increasing proinflammatory state in the human host [
36,
37]. Both findings are opposite to our findings of declining richness from pregnancy to postpartum. Interestingly, we have previously shown that, in the same group of women studied here, gut microbiota richness was higher during pregnancy compared to 9 months postpartum [
25]. Increased microbiota richness in saliva and the gut in third trimester of pregnancy suggests that the proinflammatory state may have a different influence on microbiota composition during pregnancy than in non-pregnant adults and that the physiological changes during pregnancy influence the microbial community in the host. We hypothesize that increased bacterial richness in late pregnancy is of benefit for the fetus as it ensures that the child will encounter a more diverse bacterial ecosystem at birth. Comprehensive, longitudinal studies of women before, during and after birth, and their offspring are necessary to address this hypothesis.
So far, only one study has reported differences in the salivary microbiota composition during late pregnancy [
38], and no differences has been reported postpartum in women with GDM compared with normoglycaemic women. An aberrant oral microbiota composition during pregnancy has been associated with adverse pregnancy outcome [
16‐
20] and with type 2 diabetes in non-pregnant adults [
6,
7]. The compositional aberration of the salivary microbiota during pregnancy and postpartum in women with GDM in our study has similarities with salivary microbiota previously described in non-pregnant individuals with type 2 diabetes. Species of
Neisseria were depleted in women with GDM during pregnancy and postpartum.
Neisseria mucosa has been reported to decrease in relation to an increased salivary glucose level, though not statistically significant [
35]. In a mouse model of pregnancy, salivary
Neisseria was found in the placenta of all animals, indicating that this species is able to migrate from the oral cavity to distant, anatomically separate organs [
24]. This, together with our findings, indicates that decreased oral abundance of
Neisseria potentially is associated with an unhealthy metabolic state in the human host and might have an influence on the increased risk of later metabolic disorders in children born to mothers with GDM. Similarly, a decreased abundance of oral
Atopobium has been associated with an increased risk of diabetes [
39] and in line with this we found
Atopobium at the OTU-level to be depleted postpartum in women with previous GDM. The depletion of
Atopobium postpartum in women with previous GDM might contribute to the known increased risk of developing type 2 diabetes [
11].
Increased abundance of periodontal pathogens including
Porphyromonas gingivalis and
Treponema denticola has previously been reported in the oral microbiota of individuals with pre-diabetes [
6] and type 2 diabetes [
40‐
43].
P. gingivalis induces insulin resistance when administered orally in mice [
8] and
T. denticola and
T. socranskii are positively associated with increased salivary glucose levels [
35] and have been detected in increased abundance in the placenta of women with preeclampsia [
44]. Postpartum we identified one
Treponema OTU increased in women with previous GDM. Comparing non-pregnant and pregnant women, increased abundance of
Porphyromonas gingivalis–a species with known immunomodulatory capacities [
45,
46]–in subgingival plaque has been reported during pregnancy [
32]. Interestingly,
P. gingivalis has been detected in the amniotic fluid of women at risk of premature delivery [
47] and in the placenta of pregnant women with preeclampsia [
48,
49]. Based on these findings it could be speculated that
Porphyromonas is implicated in the increased risk of adverse pregnancy outcomes associated with GDM.
Although not differentially abundant between women with and without GDM, we identified another species of
Porphyromonas, P. endodontalis, which was positively associated with higher fasting and higher 2-h stimulated plasma glucose level and with lower disposition index. Like
P. gingivalis, P. endodontalis is considered a periodontal pathogen [
50], and we therefore speculate, that
P. endodontalis might have similar effects on glucose homeostasis as
P. gingivalis during pregnancy. Collectively, our findings suggest similarities between women with previous GDM and non-pregnant adults with periodontitis, pre-diabetes and type 2 diabetes. Likewise,
Streptococcus spp. were depleted during pregnancy and postpartum in women with GDM, and during pregnancy abundance of different
Streptococcus spp. were associated with lower fasting plasma glucose concentration, higher stimulated glucose and both lower and higher disposition index.
Streptococcus in saliva has previously been positively correlated with cardiovascular disease markers in patients with atherosclerosis [
51]. Our divergent results might reflect different strains of
Streptococcus, but the applied 16S rRNA gene sequencing method does not allow us to investigate this at a higher taxonomic resolution.
Obesity has been found to alter the oral microbiota in non-pregnant women [
52], and increased abundance of
Neisseria mucosa and
Prevotella has previously been reported in overweight non-pregnant adults compared with normal weight controls [
52,
53]. After adjustment for pre-pregnancy BMI we found
Prevotella enriched in pregnant women with GDM, and–without adjustment for BMI–
Prevotella spp. were associated with higher concentrations of fasting and stimulated plasma glucose levels. Oral
Prevotella has previously been associated with increased glucose concentration in women with GDM compared to healthy pregnant women at the time of delivery [
38]. The same study found that the most remarkably concordance in microbial composition was between amniotic fluid and the oral microbiota in women with GDM, and the authors indicated that the that the microbial composition and variation of both mother and newborn could be driven by the health status of the pregnant woman and furthermore that the effects of GDM on microbes in pregnancy might be vertically transmitted to the baby during pregnancy [
38].
Notably, we found that body mass index had a major influence on the salivary microbiota composition, as many of our results were abolished after adjustment for pre-pregnancy BMI. The strong confounding effect of BMI in our study suggests, that the variation in salivary microbiota composition associated with gestational diabetes and glycaemic traits during pregnancy, is related to the pre-pregnancy metabolic state of the host.
Uniquely, our study compares salivary microbiota composition in late pregnancy and 9 months postpartum in the same women making it possible to investigate pregnancy-associated changes in the salivary microbiota. However, the normoglycaemic control group might not be representative of pregnant women in general without any risk factors for development of GDM. Being at increased risk of developing GDM, our normoglycaemic women might have an aberrant salivary microbiota. Consequently, the difference between the two groups of women might not be as pronounced as if a group of healthy, pregnant women without risk factors had been studied. Another limitation of the study is that the periodontal status of the participants was not applicable in the study design, but would be of interest to investigate in a similar study.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.