Nasopharyngeal and oral microbiota profiling in SARS-CoV-2 infected pregnant women

Categories: Disease & Virus

October 9, 2025

Study population and pregnancy outcomes

A total of 43 pregnant women were enrolled in the study, including 21 non-infected (healthy controls) and 22 SARS-CoV-2-infected participants (13 asymptomatic and 9 symptomatic; Table 1). Only in 37 of the 43 participants was it possible to perform both nasopharyngeal and oral microbiota analysis (18 non-infected, 11 asymptomatic, and 8 symptomatic SARS-CoV-2-infected women). Additionally, in a further 5 patients only the nasopharyngeal (N) swab specimens could be analysed with a total of 42 patients analysed (21 from healthy controls, 12 asymptomatic and 9 symptomatic SARS-CoV-2-infected women), and in only one other patient was it possible to analyse the oral (O) swab specimen, with a total of 38 (18 from healthy controls, 12 asymptomatic and 8 symptomatic SARS-CoV-2-infected women) (Figure S1).

Table 1 Demographic and clinical variables of the study population subdivided in non-infected (healthy), asymptomatic (SCoV-2_A) and symptomatic (SCoV-2_S) SARS-CoV-2 infected pregnant women in nasopharyngeal (N) and oral (O) swab.

The characteristics of the study population, along with laboratory test results and pregnancy complications, are detailed in Table 1. The prevalence of pregnancy complications was comparable between infected and non-infected groups. Among the 22 infected women, 9 were symptomatic; of these, 4 (44%) experienced severe symptoms (dyspnea), while 5 (56%) had mild to moderate symptoms (cough, fever). None required admission to the intensive care unit. Notably, the control group exhibited a significantly higher white blood cell count compared to the symptomatic patients.

Perinatal outcomes, summarized in Table 2, did not significantly differ between the groups.

Table 2 Demographic and clinical variables of newborns of non-infected (healthy), asymptomatic (SCoV-2_A) and symptomatic (SCoV-2_S) SARS-CoV-2 infected pregnant women. Data are presented as median (interquartile range) or n (%) as appropriate.

Nasopharyngeal and oral microbiota in the overall study population

The meta-taxonomic analysis of the nasopharyngeal and oral microbiota in asymptomatic and symptomatic SARS-CoV-2 positive and negative pregnant women revealed variations in the microbial community structure consistent with the type of swab, reflecting the biogeography of the human respiratory microbiota^25,26,27.

In the studied population, bacterial variability in the nasopharyngeal microbiota was greater than in the oral microbiota. The most abundant genera in the nasopharyngeal microbiota were: Lachnospiraceae, Corynebacterium, Porphyromonadaceae, Dolosigranulum, Lactobacillus, Staphylococcus, Prevotella, Veillonella, and Streptococcus; while in the oral microbiota, the most abundant genera were primarily Streptococcus followed by Prevotella, Veillonella, and Haemophilus (Fig. 1A and Table S4).

Alpha and beta-diversity analyses demonstrated differences in microbial community structure between nasopharyngeal and oral microbiota (Fig. 1B–D), although no significant differences in alpha-diversity were observed. Comparisons of alpha-diversity between nasopharyngeal and oral microbiota were performed using the Wilcoxon rank-sum test (Healthy, P = 0.11; SCoV-2_A, P = 0.091; SCoV-2_S, P = 0.29).

Beta-diversity analysis between nasopharyngeal and oral microbiota was performed using PERMANOVA for the Unweighted UniFrac distance (Healthy, P = 0.001; SCoV-2_A, P = 0.001; SCoV-2_S, P = 0.003), for the Weighted UniFrac distance (Healthy, P = 0.001; SCoV-2_A, P = 0.001; SCoV-2_S, P = 0.001) and the Bray-Curtis dissimilarity index (Healthy, P = 0.001, SCoV-2_A, P = 0.001; SCoV-2_S, P = 0.001).

Changes in the nasopharyngeal microbiota in SARS-CoV-2 infected pregnant women

The nasopharyngeal microbiota of SARS-CoV-2 infected pregnant women (both asymptomatic and symptomatic) and healthy control women was studied. The alpha and beta-diversity analyses showed no significant differences among the groups, although the results were close to statistical significance (alpha diversity: Wilcoxon rank-sum test, p > 0.05; beta diversity: PERMANOVA, p = 0.056 for the Unweighted UniFrac distance and p = 0.056 for the Bray-Curtis dissimilarity index) (Fig. 2A).

In the nasopharyngeal microbiota, the analysis of microbial community structure according to clinical data shows that the alpha-diversity (i.e. the within sample diversity) showed no significant differences according to the different clinical variables taken into account (i.e. ethnicity, BMI, pathologies, past surgeries, drug usage, allergies, smoke, dental caries last 2 years, comorbidities) (pairwise Wilcoxon rank-sum test, p > 0.05).

The analysis of beta-diversity (i.e. the between samples diversity) showed a significant difference in the microbial community structure according to smoking and allergies by using the Unweighted UniFrac distance (PERMANOVA, p p

Overall, these data suggest that the nasopharyngeal microbial community structure is not significantly affected by the health status of the patients, showing only minor differences close to statistical significance.

Further analysis revealed variations in microbial taxa between the nasopharyngeal microbiota of asymptomatic, symptomatic, and healthy pregnant women. Both asymptomatic and symptomatic women showed an enrichment of pathogens and pathobionts such as Corynebacterium, Fusobacterium, Neisseria, Streptococcus, Haemophilus, Mycobacterium and Porphyromonas compared to the control group.

Asymptomatic pregnant women showed an enrichment of pathobionts such as Neisseria, Streptococcus, Haemophilus and Porphyromonas compared to the control group (Fig. 2B,C, and Table S5).

Symptomatic patients were enriched in pathobionts such as Campylobacter, Fusobacterium, Mycobacterium, Neisseria and Porphyromonas compared to the control group (Fig. 2D,E, and Table S6). Compared to asymptomatic women, symptomatic patients showed enrichment in pathobionts such as Campylobacter, Corynebacterium, Fusobacterium, Haemophilus, Mycobacterium, Neisseria and Porphyromonas, Streptococcus (Fig. 2F,G, and Table S7).

Changes in the oral microbiota in SARS-CoV-2 infected pregnant women

The alpha-diversity analysis showed no significant differences among groups using different ecological indices (pairwise Wilcoxon rank-sum test, p > 0.05). The analysis of beta-diversity based on patients’ health status showed a significant difference in microbial community structure using the Unweighted UniFrac distance (PERMANOVA, p = 0.043), while no differences were observed using the Weighted UniFrac and Bray-Curtis metrics (PERMANOVA p > 0.05) (Fig. 3A).

In the oral microbiota, the analysis of alpha-diversity (i.e. the within sample diversity) showed no significant differences according to the different clinical variables taken into account (i.e. ethnicity, BMI, pathologies, past surgeries, drug usage, allergies, smoke, dental caries last 2 years, comorbidities) (pairwise Wilcoxon rank-sum test, p > 0.05).

Additionally, significant differences in microbial community structure were observed according to the ethnic origin of patients using the Weighted UniFrac distance (PERMANOVA, p

A variation in microbial taxa was observed between the oral microbiota of the three study groups (asymptomatic and symptomatic SARS-CoV-2 infected pregnant women and healthy group): both asymptomatic and symptomatic pregnant women showed an enrichment of pathobionts such as Neisseria, Fusobacterium, and Streptococcus. Asymptomatic women showed enrichment of pathobionts such as Neisseria and Streptococcus compared to the control group (Fig. 3B,C, and Table S8). Symptomatic patients showed enrichment of pathobionts such as Fusobacterium and Streptococcus compared to the control group (Fig. 3D-E, and Table S9).

Regarding the differences between asymptomatic and symptomatic women, many pathobionts such as Streptococcus, Fusobacterium and Porphyromonas were enriched in both groups, while other potential pathogens such as Haemophilus and Neisseria were more abundant only in asymptomatic women (Fig. 3F,G, and Table S10).

Best predictor of SARS-CoV-2 infection between nasopharyngeal and oral microbiota in pregnant women

To better understand which sampling method would better predict the infection status of the patients, we used a random-forest classifier on metagenomics data from nasopharyngeal and oral swabs. With the AUROC of 0.78 and an out-of-bag (OOB) error rate of 28.6%, the classifier showed a significant predictive power (permutation tests p-value = 0.003, accuracy = 76%, kappa = 52%) for the nasopharyngeal microbiota (Fig. 4).

The most important features selected by the classifier were Gulbenkiania, Burkholderia, and Actinomyces, among others, all taxa were significantly enriched in control group compared to SARS-CoV2 infected pregnant women. In contrast, the oral microbiota was a poorer predictor of patients infection status (OOB = 41%, permutation tests p-value = 0.117, accuracy = 57%, kappa = 15%).

However, when trained to classify patients as symptomatic COVID, asymptomatic COVID, or healthy controls, the classifier performed poorly in predicting infection severity showing an OOB error rate of 52.63% (p = 0.083) for the oral site and 55.81% (p = 0.05) for the nasopharyngeal site.

The 10 most important genera for classifying SARS-CoV-2 infection vs. controls in nasopharyngeal microbiota are: Gulbenkiania, Burkholderia, Actinomyces, Enterococcus, Fusobacterium, Oscillibate, Streptococcus, Moraxella, Thermus, Neisseria, and Kocuria.

Correlation of the nasopharyngeal and oral microbiota with pregnant women’s clinical parameters

In addition, we correlated the relative abundance of the most prevalent genera (with a mean relative abundance > 0.1%) identified in the nasopharyngeal and oral microbiota with the different clinical parameters of pregnant women. Dolosigranulum and unclassified Neisseriaceae genera correlated positively with increased body weight in pregnancy.

On the other hand, we found a negative correlation of unclassified Prevotellaceae with BMI and body weight gain. Lastly, Treponema, Peptosteptococcus and an unclassified genera of Bacteroidales positively correlated with blood procalcitonin (PCT) levels (p 5A,B).